It's been almost 12 years since I first started a thread "introducing" a company called xDuoo to HeadFi. At the time the company was focused on portable amp/DAC units, and their offerings seemed quite well sorted for that era. They had extensive functionality and very competitive sound for what I thought was fair pricing, and I enjoyed the time I spent with a few of their demo units. Since portable gear wasn't really a big focus of mine, I kind of lost track of what xDuoo might be up to from there, despite my enjoyment of their UA-05 device in particular.

All these years later, xDuoo is still around, and seems to have expanded its offerings to include both portable and desktop oriented devices. Their gear looks to be a mixture of both tube and solid state headphone amplifiers, dedicated DACs, and some integrated DAC/headphone amp combo units. xDuoo North American distributor Apos Audio was kind enough to send over an xDuoo TA-20 Plus which gives me a chance to see how things have evolved over the past decade. They also sent their Apos Ray 12AU7 tubes for some tube rolling action, so I'll have some thoughts on those later as well. I've been using the TA-20 Plus for a few months with a variety of headphones and source components, and have grown to really appreciate what it has to offer.


Design
The TA-20 Plus is a hybrid desktop headphone amplifier selling for $499. I'm assuming there was an original TA-20 model (before the Plus) but I'm not familiar with that at all, so I went into this with a totally clean slate (unless we count the units I heard a dozen years ago). xDuoo tells us the TA-20 Plus is a fully balanced hybrid design using a 5RZ4 tube for rectification, a pair of 12AU7 tubes for the gain stage, and a discrete class A buffer stage on the output. This should theoretically give a nice mixture of strengths - beautiful valve tone combined with solid-state technicalities, plus the ability to roll tubes to taste.

Of course, we've seen many hybrid headphone amps in the past, and some deliver on this promise more than others. The key distinction here may be the tube rectification, as that isn't so commonly seen in other hybrid designs. Or it may be the discrete (rather than opamp-based) output stage capable of delivering up to 2,600mW - enough to comfortably drive most headphones out there. Using the single-ended output lowers that number to 2,000mW which is still healthy enough to do justice to most headphones with just a select few (very difficult) outliers.



The company makes a point to note the size of the chassis whose profile is smaller than a standard A4 sized sheet of paper. Those desktop-friendly measurements conceal a fairly comprehensive feature set: two sets of RCA inputs plus a pair of balanced inputs in both XLR and 4.4mm formats (the latter handy for use with DAPs which increasingly feature 4.4mm balanced line-out connectivity). Outputs include the usual 1/4" unbalanced jack plus balanced connections by way of 4.4mm and 4-pin XLR. There's also an RCA line-out for aid in system integration, which would come in handy for those looking to work the TA-20 Plus into an existing speaker-based setup.

I really like the design of this amplifier, both from an aesthetic and a functional perspective. The left front panel is dominated by the previously mentioned trio of headphone outputs. In the middle of the panel is the (relatively large) VU meter, and I will never tire of watching that hypnotic dance as I do my listening. To the right of that is the volume display composed of 25 red LED "dots" per digit, which can show any volume level from 00 to 99. Next we get small indicators showing which input is currently active. Lastly, on the far right side we find the multi-function knob, which is tastefully done in red anodized aluminum to both match the red volume display as well as contrast the otherwise monotone dark gray of the device. Spinning the knob adjusts volume whilst pressing it cycles through the various inputs. It's all quite intuitive and I really can't think of anything I would want done differently.



Speaking of volume adjustment, that aspect is handled by the Texas Instruments PGA2311 digitally controlled analog volume chip, or more accurately a pair of them since this is a fully balanced design. That results in excellent matching between channels even at very low volume settings - avoiding the imbalance problems often seen in this price class when using traditional potentiometers. While the TA-20 Plus does not have multiple gain settings, the range seems large enough to accommodate everything from sensitive IEMs to power-hungry planars, as I'll discuss shortly.



Looks are obviously subjective but I personally find the TA-20 Plus quite fetching. The fit and finish are about as nicely done as anything I've seen in the sub-$1k price bracket, with tight tolerances and consistent finish indicative of a finely-tuned assembly process. The tube cage is effective and actually looks good (neither of which is always true with tube amps) and I actually leave it on much of the time. I was not sure how I'd like the dark gray color as it doesn't quite match with the black gear in my setup. But after living with it for some time I've grown fond of it - I'm reminded of my "silver" Pass Labs HPA-1 which is distinctly darker than my other silver gear yet somehow looks amazing when assembled together. The xDuoo amp has a similar effect when paired with my black equipment. Matching gear is great but sometimes a deliberate variation keeps things interesting without being over the top.




Listening
I plugged the xDuoo TA-20 Plus into various setups ranging from budget to modest to fairly extreme, just to get a sense of how it performed in different situations. I started with the stock tubes and later explored a few different upgrade options. As usual, listening was mainly done via Roon, and the gear was wired with Audio Art cabling all around. Balanced power was provided by an Equi=Core 1800 and a really wide range of headphones were used, the specifics of which I'll get into later.

In stock form, the TA-20 Plus gives a mostly neutral, universally pleasing sound that should have something to offer for almost any listener. I say "mostly" neutral as there is a bit of bloom and warmth, plus a very mild softening of treble energy as well as transient snap which collectively makes it a somewhat relaxed presentation overall. I would not call it overly dark though - there's ample response in the highest regions, it just doesn't have as much bite or sparkle as we might get from a truly neutral amplifier.

That slightly relaxed feel helps us focus on the crucial midrange presentation, which is beautifully done on the TA-20 Plus. Playing everything from audiophile classics (Pink Floyd, Fleetwood Mac, John Coltrane) to a variety of generalized rock (The Mother Hips, Murder By Death, Dengue Fever), highly technical metal (Carnosus, Beyond Creation, First Fragment), to other music I find harder to categorize (Zoe Keating, OK Goodnight, Nils Petter Molvaer), the xDuoo amp seems especially focused on midrange fullness, convincing vocal projection, a wide soundstage, and just a slight but welcome tube bloom. Low-end extension is deep enough to be convincing if not quite as authoritative as we might find in a top solid-state design priced in this same ballpark. It's a bit of a trade-off for that engaging midrange, and I'd say more often than not it's a worthwhile sacrifice.



The overall feeling brings to mind listening to a good single-driver speaker design, where midrange coherence, purity and natural tonality are prioritized over having the most extension up top and the most impact down below, as compared to a comparably priced 3-way multi-driver design. But again, this is not a massive coloration, more like a subtle bit of character that I notice after logging enough hours across different musical genres and headphones.

Note that pairing the TA-20 Plus with a somewhat lean sounding DAC is generally not a great idea in my experience. I don't think I'd personally choose a Benchmark or RME DAC (as great as those can be in other circumstances) as the result would not be as dynamically involved as I might like. This can be somewhat mitigated by tube rolling but even so, I think this amp pairs better with sources giving a somewhat meatier presentation with robust low-end impact.

The most synergistic pairing I found was probably the Musician Audio Aquarius DAC which is perhaps a bit much in terms of pricing compared to the xDuoo. But I imagine the more affordable Musician Audio Pegasus or even the Draco would also make great matches. One surprising combo that I ended up liking far more than expected was the Eversolo DMP-A6 streaming DAC. As a fairly neutral source, I would have theoretically placed it closer to the previously-mentioned Benchmark/RME camp in terms of perhaps not being an ideal match. Yet with the right choice of digital filter plus tube rolling (more on this later) the versatile Eversolo actually did a great job as long as we avoid certain headphones (HD800 for example). The DMP-A6 with the TA-20 Plus - along with some upgraded tubes - ends up costing roughly $1,500. While not an insubstantial sum, that seems like a pretty fair price for the level of SQ it delivers.



Headphones
When it comes to headphone pairings, I found quite a few that I enjoyed along with several that didn't seem ideal.

The Good
ZMF Atrium Closed: the 300 ohm ZMF designs always seem to thrive when driven by tube gear, and the TA-20 Plus is no different. Despite being a hybrid design, it still gives enough of that welcome liquidity and beautiful tube tone to make a great pairing. I particularly enjoy this combo with stuff like Pearl Jam, Hendrix, Soundgarden, The New Amsterdams, and other similar rock, due to the lovely way this combo renders guitars and vocals.

ZMF Caldera: I was unsure how this would go but the result is very enjoyable. It steers the sound away from the clinical side which can sometimes be the case with Caldera depending on the amplifier involved. This has just enough euphonic coloration to make it enjoyable for all sorts of music, whilst retaining strong technicalities. Listening to jazz or classical of all sorts? This might be my favorite combo of the bunch for those genres.

Audeze LCD-5: Similar to the Caldera, but even more accentuated. The LCD-5 has the potential to sound amazing but can also be very sterile, so system matching is key. The TA-20 Plus can be a great partner depending on source and tube choice. Stock tubes don't quite do it for me, but others move the needle in a pleasing direction with increased low-end impact and even more engaging midrange bloom. I realize this is something of an unlikely duo (based on pricing) but I just wanted to note that with the right tubes this combo is surprisingly enjoyable.

AKG K812: What a great pairing! I sometimes forget about the old K812, but using it with the TA-20 Plus reminds me of how good this thing still sounds. It's a great "monitor" style presentation that hits just the right combo of accuracy and resolution without going too far. The result is a punchy, fast, clean sound that feels accurate but not overbearing whilst sounding great with everything from Crooked Still to The Civil Wars to Chthonic. Anyone who loved the HD800/HD800S on a theoretical level, but couldn't bring themselves to actually enjoy those in the real world, would probably have a great time with this setup.

Focal Elex: Another superb pairing, with the added benefit of being much more price appropriate compared to the more expensive headphones listed above. This is just a great all around performer, suitable for all genres and pleasing whether listening at low or high levels. This was one of the few headphones where I spent time switching from single-ended to balanced output and while the differences were not drastic, I would recommend going balanced if you have the capability. It just seemed to give "more" of everything I enjoyed, from transient snap to soundstage width, with no downside other than potentially the cost of a balanced cable. I'd call it maybe a 10% increase in SQ, where single-ended listeners still get an enjoyable experience, but balanced is the best way to take full advantage of the amp.

Meze Liric 2: The Liric 2 is sort of marketed as a high end "portable" headphone but I just ignore that altogether and enjoy it at home. I haven't had it as long as the above headphones but so far it sits right up there with the Elex, Atrium Closed, and K812 in terms of being an attractive match for the TA-20 Plus. In fact I actually enjoy it more than its twice-as-expensive sibling, as I'll discuss shortly. Note that this one really does benefit significantly from balanced operation, and driving it single-ended seems rather flat and dull in comparison.

IEMs: I don't listen to IEMs as much lately but I did try some out and had fairly good results across the board. My AME Radioso was the best match, with inky black backgrounds and a saturated, punchy, and refined sound where the tube harmonics really shined through. The Ultimate Ears Reference Remastered was similarly excellent, with the focus shifted a bit towards speed and clarity. My 64 Audio A18t was also very enjoyable but did have a minor background hiss - not very intrusive, and actually quieter than the noise inherent in many older jazz recordings, but still noticeable at times. I suspect that would be more of a problem with certain other hiss-prone models like Empire Ears, but I didn't actually verify that yet. Still, the takeaway is that at least some IEMs do a great job here, despite that not being an expected strength for a fairly powerful tube-based desktop amplifier. Note that single-ended operation is totally acceptable here, with the percentage of improvement gained by going balanced ranging in the low single digits.


The Bad
Meze Elite: The Elite is one of my all time favorite headphones, but this combo just doesn't do it for me. It sounds a bit dull and dynamically compressed, lacking the full-bodied sound I know it is capable of delivering. There's no energy there, and the whole thing seems a bit "gray" for lack of a better word. Even with the ideal mix of source, tubes, and pads, the result topped out at merely "ok", which to my ears makes this something of a dead end. I'm not sure how many people would pair a $500 amplifier with $4k headphones, but as far as mismatched pricing goes, you'd be much better off with a Caldera or LCD-5.

Campfire Audio Cascade: This one actually didn't surprise me much. The Cascade is tons of fun but demands an iron-grip to control its fun but admittedly somewhat tubby low-end, and the TA-20 Plus isn't really up to the task. While I do think some aspects of the midrange presentation are great, the overall experience doesn't fit all that well.

Raal Requisite CA-1a: This is one of my new favorites and I'm still working on familiarizing myself with its character. It obviously requires the use of Raal's (included) TI-1b transformer box between the amplifier and headphone, and the specs indicate it needs at least 2 full watts minimum to perform well. The TA-20 Plus does meet that spec but feels like it needs a bit more juice to really make the ribbon drivers come alive. Either that or it's just not an ideal match in terms of sound signatures. It's certainly not the worst combo I've tried, but neither does go far enough in showcasing the capabilities of this unique headphone. I end up preferring the AKG K812 which gives a generally similar yet slightly better presentation, for a lower price, and without the complexity of the transformer box. Note that with the right amplification the CA-1a is a major step up from the K812 (as much as I still love that headphone) but the TA-20 Plus is just not the right match to make that happen.



Tubes
Apos also sent over their Ray 12AU7 tubes which sell for $180 for a matched pair. I'll do a separate writeup on those going into more detail, but the short version is that they pair extremely well with the TA-20 Plus, transforming it from a solid little amplifier to a borderline great one. The Ray tubes seem to unlock better articulation, more soundstage layering, and increased treble air over the stock tubes. The general nature doesn't change drastically but rather feels like an evolution, just being a more capable amplifier all around.

I've got some NOS 12AU7s that do certain things better than the Ray tubes, but that often comes at the expense of some other aspect. Not to mention the crazy pricing of NOS tubes these days. So no, the Ray tubes don't quite compete with my vintage Amperex Holland Bugle Boys, but that's expected considering the prices involved.

The stock tubes are not bad at all and I don't think people need to immediately upgrade in order to enjoy this amplifier. But down the road, I certainly think the Apos Rays are worth the expense for bringing the sound to another level without completely changing the overall presentation.


Comparisons
At the moment I tend to have a bunch of higher priced amplifiers on hand, so it's not really fair to directly compare most of them. I really wish I had the latest Schiit Lyr here as that seems like an intriguing competitor, but unfortunately I haven't heard that one yet. The best I can do are the following which are not terribly different from the TA-20 Plus in terms of pricing.

Musician Audio Andromeda: This solid-state amp sells for $869 which is less than $200 off from the TA-20 Plus if we factor in the Apos Ray tube upgrade. It offers a drastically different presentation, being very fast with extreme clarity and absolutely no bloom or coloration whatsoever. This gives a completely unique feeling when listening back to back with the xDuoo amp - Andromeda digs far deeper into the smallest details with laser-like precision, and is correspondingly ruthlessly revealing. Whether that's a good thing or a bad thing depends on your source/music/headphones.

Meanwhile the TA-20 Plus is much more accommodating and gives a relatively smoother, more laid back experience all around. While both models are fully balanced designs, the Andromeda takes a major hit when used via single-ended inputs or outputs. Meanwhile the xDuoo, while sounding its best in balanced mode, remains very enjoyable even when fed via RCA or using the 1/4" unbalanced headphone connection. The end result is the tube hybrid xDuoo being far less picky, focusing on musical enjoyment even at the cost of some accuracy, whilst the solid-state Andromeda is a more "serious" and demanding amplifier.

I can find situations where either amp is clearly superior, and there is not much overlap between them. This means folks would likely have no trouble choosing one amp over the other, though I can't predict which one that may be. Or, to look at it from another perspective - it would actually be quite useful to own both amplifiers, as they are vastly different to the point where each gives unique insight, and they could very happily coexist in a system.


Cayin HA-1A mk2: The Cayin has been around for a while by now so plenty of people have owned or auditioned it at some point. The original price was $999 but I see it go for significantly less on a fairly regular basis, to the point where it becomes generally cost competitive with the TA-20 Plus with the upgraded Apos Ray tubes. To contrast with the tube hybrid xDuoo design, this is what I'd call a more traditional transformer-coupled tube amplifier - although the aesthetic design is anything but traditional with its tall, boxy shape and rosewood enclosure.

In terms of sound the Cayin offers a more obvious tube bloom with excellent solidity and fluidity plus superior bass extension. But the xDuoo is more technically accomplished with a lower noise floor leading to superior dynamics and thus better compatibility with sensitive headphones or in-ear monitors. It also feels faster and more accurate overall, where the Cayin is somewhat obviously euphonic regardless of headphone choice.

Those differences can vary based on tube rolling which is another significant point of divergence. The TA-20 Plus gets moderately better or worse (or sometimes just different) when swapping tubes, but always retains largely the same core character. Contrast that with the HA-1A mk2 which completely transforms based on choice of tubes, to the point where it really does feel like a bit of a chameleon at times. There's definitely more potential inherent in the Cayin but of course extracting it can get rather pricey. If we focus on matching prices with a TA-20 Plus running the Apos Ray tubes compared to a (sale priced) HA-1A mk2 on stock tubes, I think I'd have a hard time not choosing the xDuoo for its cleaner and more linear presentation which still retains some great tube flavoring.


Conclusion
As stated above, I regret not having access to several more popular amplifiers in this price range, but I think my point stands - the TA-20 Plus competes nicely with two somewhat more expensive amps that I regard highly. The specifics of those comparisons may not matter as much as that general idea.

As for headphones, again the specifics may not matter quite as much as the overall numbers. I had great success with 6 different headphones and 3 CIEMs, whilst only 3 headphones didn't pair all that well for one reason or another. I consider that a very respectable ratio, particularly for a tube amp with a bit of character to it. I have a suspicion that if I were to incorporate more "sanely" priced headphones into the mix, the ratio would be even more favorable.

Ultimately the TA-20 Plus is an impressive amplifier which deserves attention for anyone shopping in this price range. I'm pleased to see xDuoo still around all these years after I first gave them a spotlight on this website, and I look forward to seeing/hearing what else they can come up with.

Matrix Audio has released a good number of products over the years, and nearly every one I've tried has been quite competitive in its price range. But the one thing they had yet to offer - despite me telling them they should, over and over again - was a dedicated transport. They did end up adding streaming audio playback functionality to all their DACs, some of which also had pre-amp capabilities and/or headphone outputs, but still never gave us a device to handle that function on its own.

Finally, the team at Matrix offers just that, in the new Element S music streamer.




The Matrix Element S ($1,499) is a dedicated transport with Ethernet and built-in WiFi capabilities. It does Roon, Tidal Connect, Spotify Connect, DLNA/UPnP, and a bunch of other options for playback. The front panel display is a touch screen and shows album art, sample rate, and track information. Basically if you've ever seen or used one of the more recent Element products such as the Element X2, this device is very similar - just focused on providing clean digital outputs rather than integrated D/A conversion.

The Element S also has a lot in common with the recent Matrix X-SPDIF 3 which is known as one of the best digital-to-digital (aka USB to SPDIF) converters around. Basically take the internal circuitry of the X-SPDIF 3, put it in the enclosure from the Element series with the usual streaming/touchscreen/etc, then add a high quality internal PSU specifically made for this application, and you've got yourself the Element S. Critical components of the design include the pair of ultra low phase noise Accusilicone "femto" clocks along with proprietary FPGA data marshalling for the lowest possible jitter and noise. That should give us extremely clean digital outputs in USB, AES, optical, coaxial, or electrically isolated I2S (via HDMI) which is configurable in software to work with various DACs. The Element S is capable of playing up to DSD512 or 768kHz PCM sample rates, obviously depending on the capabilities of the DAC it pairs with.

You can also feed the Element S via USB input, which initially seems weird because why not just get an X-SPDIF 3 if we aren't using the integrated transport features? But Matrix points out that the Element S works well using a USB output from phones and tablets, which I admit could come in handy from time to time. Still, my primary use case would mostly be direct streaming, using Roon for the files stored on my NAS as well as Tidal and Qobuz. I have also had good results using the Matrix MA Remote app, so Roon isn't the only game in town (though it is my preferred player of choice).



This write-up is nearly a year in the making, not because I'm that far behind on my reviews (not quite at least...) but due to some circumstances beyond my control. You see, Matrix launched the Element S in early 2023. I got my review unit in May and got to listening. Results were extremely promising from a sonic perspective but there were some bugs and issues to work out, including the pending Roon certification. Long story short, it took many months for those tasks to be completed, and since I'm primarily a Roon user I decided to move on to other projects until that was ready. I know some folks who skipped the Element S altogether, and some who actually bought it and sold it due to the Roon issue along with other bugs. So overall I would call this a pretty rocky product launch for Matrix.

That said, the Element S is now fully functional, Roon ready, and runs about as flawless/bug free as any other streaming device I've used. Which means it's finally ready for a full review.



In the interest of getting this published in a more timely fashion, I'm going to get right to the point. The Matrix Element S is a superb transport. When it comes to audio quality, I would put it up against any of the entry level or midrange units that I've heard from brands like Aurender, Innuos, Lumin, Auralic, etc. True, many of those offer more features, larger displays, integrated storage, and in some cases more brand recognition. But in my experience you have to move very far upfield in terms of pricing to get a notable improvement over the Element S. The trend in this segment seems to lean towards multi-function devices (think onboard DAC, headphone amplifier, preamplifier functionality, etc) far more frequently than dedicated transports. So although there are quite a few models out there from those various brands, there actually aren't many options with a similar focus as the Element S. Or at least not in the same price range.

Now that Roon is working and the vast majority of the other bugs have been squashed, I have to say the user experience is satisfying these days. Wifi connectivity is fast and stable with no discernable SQ difference between that and Ethernet. Obviously your network chain plays a huge part here so I can't guarantee anything, but that's been my experience. The touch screen - though a tad on the small side compared to what I'm used to seeing on my Eversolo DMP-A6 - looks very sharp, and feels responsive while in use. I initially thought I would miss having at least a few hardware buttons or knobs but in this case it actually works quite well without them. Even the power/standby function is handled by what I'd call a "virtual button" on the left side of the panel area, which sits near the barely noticeable IR sensor. Overall it's a very sleek, modern design that looks great in an audio rack.



The Element S can play music directly from a USB C source, such as an external SSD or thumb drive. It can also pull music from a NAS or other device via UPnP. Those, combined with Tidal Connect, Spotify Connect, Airplay 2, etc accessed via the MA Remote app mean that the Element S can be your one stop shop for music playback. As is almost universally the case with devices like this, I find that it does a pretty good job with smaller libraries but is not really intended as a server managing a massive collection. The processing power is just not up to the task. So plugging in a 256GB thumb drive full of music is totally acceptable but adding a 4TB SSD will not give the greatest experience. For libraries of that size nothing can touch Roon's smooth interface, but of course that requires additional hardware so it's not a fair comparison. I suspect most folks will primarily use streaming services rather than locally file storage and I have no complaints about how the Element S handles that aspect.



One thing worth noting is that the UI really isn't the same full Android experience as seen on devices like the Eversolo or the Fiio desktop players. For example, those let us choose Tidal, Qobuz, or whichever service we favor, browse through various artists, pick our favorite, and then select an album for playback - all from the front panel display. The Element S isn't quite a stand alone unit in that same way. We still need our smartphone or tablet to run the MA Remote or Roon apps in order to access that full browsing sequence. It's only once the music is playing that we can then use transport controls on the Element S itself.

Some users may find this limiting and I can totally understand that. But for me, even while using the Eversolo DMP-A6, I typically end up relying on my phone via the screen mirroring option. It's just more comfortable to operate that way. And since I actually use Roon most of the time, it ends up being of lesser significance since Roon works via this same principle. Bottom line: I think it's really important for potential users to examine their needs and understand what the various devices on the market can actually do (with or without a supporting PC/tablet/phone/etc), in order to make the most informed choice.


Sound
When it comes to listening, I have used the Element S with a wide variety of gear and always achieved excellent results. I started with comparisons to a Microsoft Surface Laptop running Roon, acting as a sort of baseline general purpose computing device to represent the kind of thing many people use when they first get into streaming. The most noteworthy improvements came from DACs which tend to be highly sensitive to transport quality. In my current collection, that would be the Wyred4Sound 10th Anniversary DAC (with the jitter reduction option disabled), the Cayin CS100-DAC, and an old Sonic Frontiers SFD-1 mkII that I've been borrowing from a friend. In those cases, using amps such as the Cen.Grand Silver Fox, Cayin HA-6A, or Niimbus US4+, I perceived what I would call fairly significant gains in multiple areas. With careful listening of high quality recordings, I heard improved soundstage size and accuracy, deeper bass extension, more engaging midrange with less treble glare, a lack of grain in the upper midrange, and more focused transients. Using great headphones like the Audeze LCD-5, Meze Elite, ZMF Caldera, or 64 Audio A18t CIEMs, these changes were not difficult to spot. Obviously switching to less resolving headphones or playing something other than reference caliber recordings will reduce this impact by some degree, but I would say it's still an appreciable upgrade in most cases.



Less substantial but still very much worthwhile improvements came when using gear which is somewhat less picky about transport quality. For me that would be the Musician Audio Aquarius R2R DAC or the Violectric V590 integrated DAC/headphone amplifier. Those did not transform quite so drastically yet still represented what I would call noticeable gains, on the level of upgrading amplification or perhaps adding a good power conditioner to the system.

As a counterpoint, I also have some devices here which generally don't seem to care as much about transports. The Yulong DA-1 barely registered any improvement at all, with just the tiniest hint of more air and soundstage depth. That designer made what amounts to heroic efforts building a proprietary digital interface to maximize the quality of incoming signals, so this result is not at all surprising.



My Cen.Grand DSDAC1.0 Deluxe also has proprietary things happening on the input side to help reduce transport impact, and it does a very good job as well. But in this case the difference between Surface and Element S was larger than I expected... particularly when using an AES/EBU connection from the Matrix device. This tells me that despite Cen.Grand's significant attention to that aspect, the Element S is so good that it does break through and give us a reasonable upgrade.



One somewhat unusual combo I ended up trying - and absolutely loving - involved the T+A Solitaire T. While the Solitaire T is chiefly marketed as a wireless headphone, it can also connect via USB which results in lossless playback using the integrated DAC and headphone amp sections built in to each earcup. It's a surprisingly impressive headphone although for $1,700 I suppose it should be. In any case, using the Element S as a transport sounded clearly better than the Surface Pro when used in this fashion. Bigger, bolder sound with improved nuance and delicacy, the Element S/Solitaire T combo made a fantastic pairing for a simplified setup. I ended up using it way more than anticipated.



During this testing, I obviously had to use the Surface Laptop with USB - it doesn't give us anything else to work with. When using the Element S I could choose between USB, AES/EBU, Toslink, coaxial, or I2S, depending on the particular DAC in question and its feature set. This was eye opening and served as a great reminder that as much as we like to think USB is well-sorted, in some cases it will still lag significantly behind the other inputs in terms of performance. Yet with some DACs it is equal or even the best option to use, so it's worth having multiple options to choose from.


Comparisons
Having established the Element S as being significantly more capable than a basic consumer-grade laptop, it was time to move on to some higher quality dedicated audio devices. My next comparison involved the popular Eversolo DMP-A6. At $899 it represents an excellent value, particularly if one intends to take advantage of its all-in-one capabilities.

Comparing the A6 directly to the Element S is actually tough though, since each machine has a very different focus. The Eversolo gets major points for having the surprisingly capable balanced DAC as well as the larger touchscreen display and more versatile Android-based OS. But for pure transport use the Element S takes a commanding lead in sonics. I especially liked having the options to use I2S or AES, which are often the preferred connection types (obviously depending on the DAC). But even when using the same output - coaxial for example - I felt like most systems went a bit further with the Matrix than the Eversolo. Again, it's tough to make blanket conclusions since each DAC is different, but in many cases the Element S proved to have the upper hand.



I also compared the Element S to my Stack Audio Link II with the matching Stack Audio Volt PSU. The cost of this combo is fairly similar to the Element S, so I was interested in seeing which one came out ahead.

Turns out this comparison is sort of the inverse of the Eversolo vs Matrix showdown. This time around it is the Element S which has the much larger feature set, whilst the competition is a highly focused device that offers better SQ in a certain context - that context being USB which is the only output available on the Stack Audio device. It sounded more fluid and dynamic overall, with just a bit of that "last word" realism which can be so elusive. If we can live without things like a dedicated playback app (the Link II mainly relies on Roon), or front panel display, and if we use a DAC which can sound its best via USB, then the Stack Audio Link II would be the better transport choice. If, however, you might need I2S or SPDIF outputs, or any of those various other creature comforts, the Element S gets the win. That's a big deal because despite not being all that well known, the Stack Audio Link II is among my very favorite transport devices, regardless of price. For the Element S to play in that same realm is extremely impressive.

Lastly, I cobbled together a system consisting of the Surface Pro feeding the previous generation Matrix X-SPDIF 2, which I powered by a Keces P8 linear power supply. Note that the cost of this DDC/PSU combo is roughly equal to that of the Element S. It also unlocks the same connectivity as far as AES, I2S, coaxial, and Toslink, to take full advantage of whichever type a particular DAC likes best. Long story short - despite the parity in these areas, and the heroic quality of the tanklike Keces PSU, the newer Matrix device still took the win by a reasonable margin. It sounded more focused, with superior bass definition and more accurate imaging as well as a sweeter midrange. I would have liked to reproduce this same experiment using the X-SPDIF 3 since that's essentially what we find inside the Element S. But reviewers can't have every piece of gear ever made, so we work with what we have on hand.

Getting into more general terms - I have owned or extensively auditioned most of the big models from Innuos, Aurender, Lumin, SOtM, Auralic, etc and have a great respect for most of them. Yet based on my needs, and with a focus on SQ over features, I would happily choose the Element S over the majority of them, even models costing significantly more. I'm not saying the Element S beats an $18k Innuos Statement, but I would certainly put it up against their $3.6k Zen mk3, or the $5k Lumin U2, or even the $6k Auralic Aries G2.2, and be confident of a very tough fight - again in terms of SQ only. The Matrix Element S really is that good!


I2S Odds and Ends
As I've mentioned, I2S is often the best choice for maximum performance, but it obviously depends on the DAC being used. I never want to make a blanket statement and say "X interface is always superior" because it really comes down to the combination of transport and DAC working together. I generally find that when DAC makers bother to integrate I2S, it does tend to be the preferred option. But that isn't always the case, so definitely try everything and see which sounds best in your particular setup.

That said, I also appreciate the fact that Matrix gives us 4 different software adjustable configurations for I2S compatibility. The I2S format is still a mess in terms of standards - some brands use RJ45 while many others use HDMI, but the pinouts are often different so you really have to be careful. I've been able to use the Element S with a Wired4Sound Anniversary DAC, Cayin CS-100DAC, Denafrips Pontus II, and Musician Audio Aquarius, all with excellent results. But I can't guarantee that it will work perfectly with every device out there, even if they do use a "standard" HDMI connection. Just something to be aware of for those planning on going the I2S route.



Conclusion
The Matrix Element S is one of the best sounding transports I've heard in quite some time. And I don't mean "in this price range" either. Remember, many of the more expensive competitors cost more due to having integrated DACs, large internal hard drives, headphone outputs, preamp functionality, or just massive (expensive) enclosures. The Element S is a pure transport and focuses all of its resources towards that singular goal. If you are looking for an extremely high quality way to feed your DAC - particularly via I2S, AES, or coaxial SPDIF - the Element S is absolutely worth checking out. It had a bit of a rocky start but now that everything is sorted, I can happily recommend it.



Bonus Section: MA Remote
I also want to throw in some pics of the MA Remote app being used, just to give you a general idea of what the experience looks like. I find it perfectly usable though not anywhere near as engaging as Roon (but I'm a huge Roon fan so keep that in mind). As you can see, it looks nice and has most everything the average user could want, even if it isn't the best for interacting with massive collections of music.

All in all I'd call it roughly on par with the competition - with the caveat that everyone's needs and preferences are different. I can't predict if you'll like this more or less than Aurender's Conductor, Auralic's Lightning DS, etc but I can say it generally plays on that same level.

(Note that these are just screenshots from my Pixel 8, using a tablet would probably offer a better experience)





Obligatory Roon shots as well:

The Background
I recently completed a roughly two-year journey to find a new reference DAC. The project involved listening to dozens of highly-regarded options, comparing and contrasting them in endless combinations, and attempting to judge them all on features, value, and sonic performance. In the end I chose the Cen.Grand DSDAC 1.0 Deluxe as my new reference, which resulted in one of the longest write-ups I've ever done. You can read that HERE if you've got a decent amount of spare time on your hands.

My approach to reviewing the Cen.Grand Silver Fox headphone amplifier will go in the opposite direction. Rather than a sprawling comparison, I hope to keep things focused and concise, getting to the core of what makes the amplifier tick and who it may - or may not - appeal to.



The Amp
My DSDAC review covered the history of Cen.Grand so I won't rehash any of that here. As for the Silver Fox, the full name is technically the 9i-906, in keeping with the naming conventions the company has used for years. I personally find those hard to identify with so I will stick with calling it the Silver Fox, whilst recognizing how that might feel a bit odd for users of the black version (yes, the Silver Fox can also be had in a very fetching black enclosure).

The Silver Fox sells for just under USD$4,200 which puts it somewhat on the affordable side when it comes to all-out, cost-no-object headphone amplifiers. You can certainly get a great amplifier for a lot less money, but you can also spend double, triple, or even more for a headphone amplifier if you have money to burn. I never thought I'd be complimenting a $4k+ headphone amplifier for its relative affordability but such is the market in 2023, so here we are.



Fit and finish is exemplary. The enclosure has various fine details which keeps it from becoming yet another anonymous silver (or black, if you chose that version) box in the audio rack. I particularly like the side panel venting details as well as the little angled recess on the upper middle portion of the faceplate which proclaims "music is my soul" - a common design theme among many Cen.Grand devices. Silver Fox tips the scales at roughly 26 pounds, comes equipped with fancy footers for vibration control, and generally feels like a truly high-end piece of kit.





Note that this is purely a headphone amplifier: no preamp capabilities, and no digital inputs. Multi-function devices are quite popular these days but the Silver Fox sticks with a singular task and does it extremely well. The matching DSDAC 1.0 devices do have preamplifier capabilities (including analog inputs on the Deluxe model) which helps explain why Cen.Grand didn't want to duplicate functionality here. Instead we get 4 inputs to work with - 2 each of XLR and RCA. I don't know how useful that might be for the average listener but for a reviewer comparing DACs it is quite handy.



The Silver Fox handles nearly any headphone connection I can think of. There's a 1/4" jack for standard headphones, and then 3 sets of balanced outputs in the form of 4-pin XLR, dual 3-pin XLR, and 4.4mm. The only thing technically missing is a 1/8" jack which is easily achieved by using the 1/4" option with the usual adapter. A reasonably large front panel display lets us cycle between inputs, choose from 4 impedance settings, and also select between 4 different operation modes - Normal, BTL, Parallel, and Active Ground. These each take a different approach to amplification in terms of output current and grounding method, with each giving a slightly different flavor. The amp is capable of putting out a massive 20W into 30 ohm loads which means it is able to drive even the most difficult headphone loads with authority.



Internally, the amp is centered around an array of 8 Exicon lateral mosfets. Clever readers may recognize those same transistors as being at the heart of the highly-regarded Enleum AMP-23R (USD$5k), with the distinction being that Cen.Grand offers twice as many. Those numbers make sense as this is a fully balanced design, thus 2 per channel times 4 channels (the Enleum is not balanced so 2x2 is adequate). Beyond that we get substantial power capacitance, a large custom toroidal transformer, and top quality parts throughout. Those Exicon mosfets use the massive chassis as heatsink which helps distribute the thermal load - the device gets a little warm but nothing unmanageable in the least.






For further technical detail, I'm actually going to do something I almost never do, which is point you to another review. That would be the in-depth exploration over at 6moons which is definitely worth the read. I know some folks find Srajan's writing style and sonic descriptions a bit difficult to follow, but I think we can all agree he does a great job digging into the technical aspects of the design as well as the company history and motivation behind the product. If nothing else, it says a lot when a reviewer with access to vast quantities of HiFi gear chooses a relatively unknown brand as a reference... and note that he's done it twice now with the Cen.Grand DSDAC as well as the Silver Fox.

One thing I do want to quickly point out is that the Silver Fox now ships in either black or pure silver in color. My earlier production unit (as well as the one seen at 6moons) has a few gold accents, including a gold volume knob, but that is no longer the case with the latest production models. Not a massive difference but I just want to be clear about the pictures seen here versus what someone will actually receive if they purchase this amp in the silver option. At some point I'll swap mine out so my pictures reflect the actual look of the current product.


The System
For the majority of my evaluation, I listened via my main system fed with balanced power from an Equi=Core 1800 power conditioner. AC and signal cabling was all from Audio Art, with the exception of an iFi Gemini3.0 USB cable and various headphone cable upgrades from Effect Audio. My source was usually a Euphony Summus music server fed by a Keces P8 linear PSU, which then streamed to a Stack Audio Link II Ethernet bridge with the matching Stack Audio VOLT power supply. I also used a Matrix Element S as well as an Eversolo DMP-A6, both streaming via Roon. On DAC duty, the obvious pairing was the DSDAC 1.0 Deluxe, but I also spent lots of time using a Musician Audio Aquarius, Yulong DA1, Wyred4Sound Anniversary, Cayin CS-100DAC, and various others which were listed in my earlier DSDAC article. Headphones included the ZMF Caldera and Atrium Closed, Meze Elite, Audeze LCD-5 and LCD-24 Limited, Kennerton Thekk, Sendy Peacock, Campfire Cascade, HiFiMAN HE-6, Sennheiser HD800 and HD660S, and more. So I think it's fair to say I've given the Silver Fox a thorough workout.








The Sound
In keeping with my goal to keep this write-up concise, I'll explain the sonic performance of the silver Fox using bullet points. Feel free to ask any follow up questions if I end up not exploring some aspects to your liking.


+The general signature of the Silver Fox is big, bold, and authoritative. I'd still call it a neutral sound overall but with an emphasis towards a more engaging "live" feeling rather than a stereotypically clinical studio reproduction.

+Detail retrieval is top-tier, yet the rich tonality keeps it from ever feeling the least bit sterile. This is not an analytical amp in the negative sense of the word, but it does excavate deep into the music to seemingly unearth everything there is to be found in a recording. This aspect can be emphasized or minimized based on your choice of settings.

+Despite the level of transparency and that "live music" excitement I mentioned earlier, I hear zero trace of any objectionable treble zing. In fact this is some of the most well balanced, articulate, convincing treble out there. But it seems to be presented as "quality over quantity", meaning no artificial boost or attempt to fool the listener with elevated treble energy.

+There's a sense of tonal richness that likely exceeds any other headphone amp I've experienced to date. The balance is perfect for my taste - I'm not saying this is the thickest sound I've heard, but rather the most satisfyingly rich tone without going overboard.

+It drives any headphone you throw at it with ease, and has enough subtlety and nuance to pair well with tricky designs like the HD800 and Kennerton Thekk.

+The volume control uses a Muses resistor ladder solution from New Japan Radio Corporation, which cycles through the 70dB range in .5dB steps. Which I very much appreciate. Contrast that with my reference tube amp which has a custom stepped attenuator with only 24 steps, and that simply isn't enough precision for certain circumstances.

+Output tops out at 20W into 30 ohm loads which is absolutely massive. I notice lots of amps these days which claim to do 5W or 10W etc, but then specify that rating as being for 16 ohm loads - meaning it really does roughly half or even a quarter of that number into the more useful 30-60 ohm range where a lot of headphones actually live. Silver Fox will have zero issues driving any headphones with full authority and having plenty in reserve for dynamic swells.

+Partially due to the potent output, but also just the general signature, the Silver Fox presentation feels totally effortless. No treble etch or extraneous grit, supreme bass control, fluid midrange expression. I can crank volume up indefinitely without the amplifier becoming the weak link. Lots of amps can sound unobjectionable at lower volumes, yet begin to lose the plot as volume increases. Silver Fox has no such limitations, so be careful. I'm generally not a very loud listener to begin with but with this amp I find myself boosting the volume here and there, almost without noticing, since everything sounds so composed and unforced. Eventually I realize I am playing things far louder than normal, and remind myself to back off a bit. Thankfully when I do, I still hear a rich dynamic presentation. Some amps only really come alive at higher volumes and tend to sound flat/boring until you reach those levels. But not this amp.

+Soundstage is superbly well defined. This perhaps relates to the effortless nature of the amp as previously mentioned, but it often feels like I'm bumping up against the limits of the recording itself rather than being constrained by the amp. This assumes a suitably high caliber of DAC, transport, headphones, and ancillary gear, but when all those pieces fall into place, the result is magical. There's a sense of dimensionality and space which is very rarely achieved. Have I heard better? Yes, my experience is that a select few top-tier valve amplifiers using vintage (aka expensive) glass can sound even more holographic than this. But the delta is not huge and the cost difference is fairly extreme. Not to mention there are other things such as impact and immediacy which I don't think any tube amp I've heard can match. So it really comes down to priorities in the end. Suffice to say the Silver Fox resides at the pinnacle as far as solid state designs go.

+I already mentioned the macro scale and dynamics of this amp, but I'm also constantly impressed by its inner detail and subtleties. Listening to three different versions of Yo-Yo Ma playing the Bach cello suites - his original 1983 recording, followed by his 1997 release, and finally his latest 2017 version - the Silver Fox allows me to hear every variation in pitch, inflection, pace, and rhythmic intensity. And that's just the performance itself. There's also the difference in recording and mastering which, although all are very well done in their own ways, reveal changes in focus from one release to the next. The list of amplifiers capable of showcasing these differences, to this extent, is vanishingly small.

+Likewise the sense of speed and control is mindblowing, particularly when playing more complex material. For my musical diet that often leans toward various metal from around the globe: the absurdly complex technical death metal of German band Obscura, atmospheric black metal with folk elements from Romanian group Dordeduh, densely layered Danish prog metal from Vola, Melbourne-based Ne Obliviscaris with their classical inspired extreme progressive style, the virtuosic skill of instrumental supergroup Blotted Science from the USA, or the elaborate symphonic death metal of Italy's Fleshgod Apocalypse. To the uninitiated, these all might just sound like aggressive noise, but to me they are wonderfully diverse and totally unique. Many systems cannot do them justice, which can often be blamed on the recordings or even the genre in general. But with an excellent system featuring the Silver Fox and a suitably accomplished headphone (ZMF Caldera and LCD-5 are my favorites for these genres at the moment) I am blown away at the layering, the rhythmic interplay, and just the overall technical capabilities in general. Would I love all of them to be pristinely recorded/mastered by an audiophile label such as Chesky or Reference Recordings? Sure. But even so, there is plenty of quality to be extracted if your system is up to the task.

+Well recorded vocals sound spooky real with the Silver Fox. From Corinne Bailey Rae to Eric Bibb, Sufjan Stevens to Aoife O'Donovan, this amp captures the unique nuances of each voice and reproduces them in a completely lifelike manner. If you're reading this, it's a safe bet you probably enjoy music as often as possible. But consider for a moment how often we hear real voices in our daily lives. It's a sort of benchmark that we probably don't even think about. Yet we are all likely much more well versed in evaluating the realism of voices than any other instrument. So when a chain does something even slightly wrong in reproducing vocals, it really stands out, even if our brains can't really pinpoint where the deficiency might be. The Silver Fox will absolutely not be the weak link in accomplishing this task.




Settings
As far as the different impedance options go, the naming is actually a little confusing. The "low impedance" settings actually involve added resistance to the output - low 1 adds 50 ohms, low 2 adds 100 ohm, and low 3 adds 150 ohms of output impedance. Meanwhile the so-called high impedance setting is actually the original signal with less than 1 ohm output impedance. So what they really seem to mean is those "low" settings are intended for more sensitive headphones which show background noise and/or just aren't suited for the full power output. Likewise, BTL mode is said to give us the full output voltage, while the other three modes are reduced - but still plenty potent for the majority of headphones. That means using "high" with the BTL setting results in a full power output whilst the other options all mix and match for some type of reduced (but again still extremely powerful) output.

The "low" settings having higher output impedances may be an issue for certain headphones, particularly those with dynamic or balanced armature drivers and interesting impedance swings. That said there are certain situations where people actually like having a higher output impedance available, so this is not always a bad thing. Meanwhile planar magnetic designs, with their generally flat impedance curves, don't seem to care much either way. Again, it's nice to have different options to play with, even if the naming scheme is somewhat confusing.

Cycling through the four modes seems to make a larger impact on sound, though it can be unpredictable as far as how it will interact with each headphone. This makes it sort of an adventure which can bring interesting surprises. Some headphones react quite strongly to the different modes while others show hardly any difference at all. In the most extreme cases there will be a significant change in gain, a shift or reordering in soundstage width and focus, and just an overall different feel to the presentation.

If I have to generalize, I'd say BTL mode consistently works best with difficult to drive headphones, sounding very full bodied and authoritative. But it can be used even with easier loads and the sound may prove worthwhile. Active Ground mode can help with sensitive headphones which need assistance squashing slight noise-floor issues, and sometimes this mode feels like the most nuanced, detail-oriented option. Normal mode is well-balanced and realistically a safe bet the majority of the time. Parallel mode tends to have the most open, expansive signature, and sounds a little like a mixture of normal and BTL.

But I'm making it seem like these are consistent results, which they totally aren't - this is a vague pattern gleaned after months of listening, which is accurate about as often as not. So I won't bother trying to describe the modes further as they are so unpredictable. It really is nice to have different options to mess with though - again, you never know which one might be the best fit for a particular headphone.



Pairings
As mentioned earlier, the Silver Fox plays well with every headphone I've thrown at it. I don't have a HiFiMAN Susvara here any longer but based on my experience using it to drive an original HE-6, it should have no trouble at all making big, bold, well controlled sound from the HiFiMAN flagship. Ditto other power-hungry cans such as the HEDDphone, AKG K1000, and the Abyss models. Sensitive models from Grado or the always-fun Campfire Audio Cascade also pair extremely well, proving the Silver Fox has more to it than just brute force.

A massive powerhouse like this is not typically my first choice to use with ultra-sensitive in-ear monitors. These things need power in the milliwatt range while this amp has thousands of times that much. Despite that, the Silver Fox does surprisingly well, at least in many circumstances and depending on the model. The best match I can find is my Radioso tri-brid custom IEM from (apparently defunct) Korean firm AME Custom. These are relatively inefficient as IEMs go, which may be why they pair exceptionally well with the big Cen.Grand amp. Whatever the cause, the sense of scale, authority, and dynamic swing are top notch with this combo. The wide volume range leaves me with plenty of room for adjustment which is not always the case when using full-size gear with miniature in-ear monitors. The Ultimate Ears Reference Remastered CIEM also pairs extremely well, as does the 64 Audio A18t, both of which give results easily on par with some of the best full sized headphones out there. I do hear a bit of background hiss when using my most sensitive models though, so I imagine there will be some pairings which won't be ideal. But that's not unexpected considering the absurdity of the power levels involved.

Naturally, the matching Cen.Grand DSDAC1.0 Deluxe is a perfect partner for the Silver Fox. But it also pairs beautifully with a variety of other sources. Notable matches include the Cayin CS100DAC which gives a liquid, flowing, delicate performance with the tube output stage engaged, and a more snappy, visceral sound when using the solid-state output path. Meanwhile the Yulong DA1 brings out more of a sweet midrange feel, making music sound intimate and highly involving. The Wyred4Sound Anniversary DAC is bold and dynamic, whilst the Soul Note D2 is wonderfully gentle, delicate, and nuanced. The Silver Fox is ideal for comparing and contrasting sources.

Rather than continue discussing the details of every DAC combination I've tried, I'll just sum it up by saying the Silver Fox is simultaneously ultra revealing yet complementary to any upstream gear. It is revealing in terms of being deeply transparent to showcase the capabilities and limitations of the associated DAC and transport. But it also feels somewhat complementary in being so potent, rich, dynamic, and lively that it can make a boring DAC sound more exciting than it otherwise should. I'd say the Silver Fox is thus excellent whether listening for enjoyment or for evaluation purposes, as long as one remembers to take that complementary aspect into consideration.








Comparison
Although I have many excellent headphone amplifiers here at the moment, there's really only one that feels worthy of comparison - that would be the mighty Niimbus US4+. Everything else currently in the house from Pass Labs, Cayin, Violectric, SPL, Musician Audio, iFi, Feliks Audio - all amps that I greatly enjoy - are just not in the same league. For those unfamiliar, Niimbus is the summit-fi arm of Germany's Lake People brand, who also makes the highly-regarded Violectric line of gear. Niimbus later rebranded this amp as the US5 Pro which currently sells for $6,599 with the only difference being the addition of a 4.4mm balanced jack. I've used this amplifier for years as a reference and I'm still thoroughly impressed with it during each and every listening session.

Both of these amps are incredibly powerful, dynamic, clean, and focused, with superb detail retrieval and a lifelike sense of projection. Both have massive soundstage presentations, precise image localization, and thunderous bass response. Their main point of divergence is overall tonality where the Silver Fox comes across as thicker and a touch more laid back, making the Niimbus seem comparably energetic and wiry. Note weight is thicker on the Cen.Grand unit, and there's a bit more emphasis on decay trails where the German amp has a greater focus on transient attack and speed. Yet in absolute terms the Silver Fox is just as resolving of fine details, despite the difference in the way those details are presented.

In ultra-simplistic terms we might say the Silver Fox is "warmer" and the US4+ is "brighter". But as usual with simplifications, that doesn't really capture the full truth of the situation. It's not as if the Niimbus isn't stunning with the lively HD800, nor will the Silver Fox push an HD650 into the oblivion of darkness. They are both extremely competent amplifiers, with each having its own character and ultimately its own unique presentation.

I can happily listen to either one of them all day long without thinking I might be missing out by not using the other option. But I do notice myself using the Silver Fox more often these days, both for business and for pleasure. I realize that "new toy syndrome" is very real, but I've actually had the Cen.Grand amp for roughly 10 months by now, which by my standards is a rather long time (and certainly enough to get well-settled). If I'm honest, the Cen.Grand Silver Fox does seem to offer slightly more performance for significantly less money - roughly $2400 separates the two models, in favor of the Chinese option. So as much as I still love the Niimbus, I have to admit that the Silver Fox displaces it and takes the top spot in my headphone amp hierarchy.




Conclusion
It's been a big year for my audio system. Not only did I find myself a new reference DAC, but also a new reference headphone amplifier. The prior choices had been in place for years and I remained absolutely thrilled with their performance, so this was an unexpected development. Even more surprising was the fact that both replacement choices came from the same brand, which is something that has never happened before. I can think of multiple companies who offer superb DACs plus excellent speaker amplifiers but that list dwindles severely when we switch from speaker to headphone amplification. And to have both new references come from a company I had never heard of before? Unprecedented.

The Cen.Grand Silver Fox is a beast of an amp. In sound, design, build quality, and appearance, it is up there with the absolute pinnacle of the market, in a place where very few competitors dwell. The fact that it does this for a touch under $4,200, when others are charging thousands more, is extremely impressive. I can very enthusiastically recommend the Cen.Grand 9i-906 Silver Fox to anyone in the market for a top-tier headphone amplifier.

Or, to frame this another way: I've gotten in on the ground floor with quite a few brands over the years. I was among the very first, and in some cases actually had the exclusive world premier, on plenty which later went on to become "household names" (as far as that goes in the audiophile world). Examples include brands like Aurender, Noble Audio, Auralic, Empire Ears, Matrix Audio, Unique Melody, Violectric, 64 Audio, and certainly others that I can't recall at the moment, which are all staples in their respective categories by now. I'm not at all claiming I can take any credit for their accomplishments, but I might be justified in saying "I told you so" to the massive numbers of enthusiasts and fellow reviewers who went on to hear the same thing I heard.

Cen.Grand feels like another one that is destined to become a well-known and highly-regarded name in the audio scene. While they have been around for years in their own country, this seems like the beginning of their breakout on the world stage, and I would not be surprised to see Cen.Grand become far more mainstream in the coming years. And when that happens, I'll be happy to have been here from the start, pointing out how excellent their gear is.



Epilogue
With the main review finished, I figured it would possibly be illuminating to share a bit more about Cen.Grand founder/chief designer JianHui Deng. He's a fascinating guy with an interesting history - which is about to become even more unique in the coming months if all goes according to plan (unfortunately I can't give details on that yet). His philosophy seems to be that in order to fully understand a brand, where it comes from and where it may be going, you need to understand the core person/people behind the scenes. I agree with that sentiment. I've also noticed that in the audio world we do tend to learn all about the tiny, quirky boutique manufacturers and what makes them tick. Yet oftentimes we hear very little about the folks behind the products at slightly larger companies. There are of course exceptions but this got me thinking that I really would like to know much more about the designers or teams at all of my favorite audio brands. So this is my effort towards making that happen.

JianHui Deng seems to have a strong basis of music appreciation as one of his defining traits. He started at a young age playing clarinet and then added saxophone, guitar, and drums to his skillset. He played with various bands throughout his school career ranging from his early teens through his university years. Graduating university in the 1980s, he took a brief detour to work at a railway company, but soon moved on in order to start his own company designing and manufacturing audio equipment. This was in the 1990s when the Chinese HiFi scene was beginning to take off, and some of his amplifier products picked up critical acclaim and won industry awards. I was fascinated to learn about this as I feel like I have a decent knowledge of HiFi history as it pertains to my own region... but I am reminded that many areas have their own unique scenes, with their own history, and they can be very passionate and thriving in their own ways.

Moving forward to the 21st century, JianHui continued in his audio design pursuits. Realizing that the digital market was growing more rapidly than analog, he turned his attention to digital media players, which he felt had significant room for improvement on the audio side. He worked closely with US firm Marvell Technology who at the time was the premier supplier of high-end processing solutions for the then-new Blu-ray market. His goal was to improve audio performance by implementing an "independant dual clock architecture" - Marvell agreed, enabled "external clock mode" on their chips, and JianHui's solution became the standard for future digital audio players.

In 2011, JianHui Deng founded Cen.Grand, with the goal of pushing even digital audio beyond anything that had come before. His focus moved towards DSD and after being turned down in his attempts to license DSD tech from Playback Designs, he set out to develop his own solution. 5 years later, after extensive research, he and his team successfully launched their DSDAC1.0 featuring proprietary DSD1024 upsampling. The follow-up DSDAC1.0 Deluxe came two years later, which beat out dozens of worthy (and often more expensive) competitors to become my reference DAC. With the inclusion of several other proprietary aspects ("clock blocking" and "synchronous direct clock" technologies) JianHui Deng humbly suggests that he has taken DSD beyond even what its creators envisioned in terms of realized potential.

Moving back to his analog roots, JianHui launched the 9i-906 Silver Fox headphone amplifier, which is an extremely complex design made up of over 1100 individual components. His philosophy is that well-designed simple circuits can hit the desired target in terms of tuning, but it takes a much more complicated design to transcend that level of performance and achieve something truly special. Again, based on what I'm experiencing from the device, he appears to be on the right track here.

JianHui's ambitions for the future involve revisiting the media player aspect to launch a world-class digital transport. Like many others in the high-end space, JianHui is not necessarily a huge fan of USB as it pertains to digital audio transmission. He certainly makes it work quite well in his DAC designs, but argues that it is ultimately a flawed solution for audio applications, and he thus intends to develop a better alternative. To do this he is working with Intel on a new product called the GLS1.0 - I don't have a lot of details yet but I know it received a patent in his home country and supposedly he is the first person to work with Intel on this level, much like his partnership with Marvell over a decade ago. At one point he was planning on implementing a proprietary format using a twin-cable system with Toslink and BNC connectors, and later he mentioned using PCI Express - I'm not clear if those are two separate ideas or if they are both describing different aspects of the same technology. Either way the claim is that it achieves 100% lossless, synchronous data transmission, capable of handling extreme data rates and even multi-channel DSD. The clock and data streams are transmitted separately for maximum integrity and then perfectly reconstructed on the receiving end. The GLS1.0 streamer/server device will likely be the next big launch from Cen.Grand and I look forward to learning more about the unique connection with Intel - it could be a game changer.

As for JianHui Deng, he sent me a few pictures which show part of his audio journey over the decades. I post these here with permission - we all have our own journeys with music and I find each one unique and fascinating. I also learned that his son was recently hired by the prestigious Duke University as a math professor. Which goes to show how the ability to handle complex equations runs in the family. I wish I could share the rest of the news I alluded to earlier but suffice to say I think we'll continue to see innovation and fresh ideas from Cen.Grand for years to come.






(on drums)


(now, in the Cen.Grand demo room full of great equipment)

History
Long story short: I've been using a Resonessence Labs Invicta DAC as my reference for over a decade. Thankfully Invicta was more of a platform than an individual DAC (think Schiit Yggdrasil) and thus did not remain stagnant during those years. It started with the original ($3999) and evolved multiple times, culminating with the Mirus Pro Signature Edition at nearly $8K. I absolutely loved that device and despite comparisons to others costing more than double the price, I never found anything that spoke to me quite on that same level. I was fully prepared to continue using it as my reference well into the future.

Unfortunately that plan didn't work out. Resonessence Labs abruptly closed up shop, leaving customers and dealers alike stranded with no support or information. I decided it was no longer appropriate to use one of their products as a reference, and embarked on a long journey to find a worthy replacement.

Roughly two years later, I'm here to report on those results, with a focus on the DAC I ended up choosing as my new reference - the Cen.Grand DSDAC1.0 Deluxe. It's not a DAC many people have heard of quite yet, but I took a chance and ended up enjoying it the most out of many worthy contenders.



A New Reference
So what is a "reference" component anyway? There's not exactly an entry in the dictionary that covers this particular context, and it differs depending on who we might ask. Sometimes, particularly when it comes to speakers, people just mean "the most flat sounding". Which is reasonable in a studio setting but perhaps a bit too simplistic anywhere else. Many reviewers tend to use "reference" to describe the best sounding, most resolving, and overall most musically enjoyable component they can find, even if it isn't strictly defined by measurements or specs. That component then becomes their gold standard against which other components are compared. Jeff Fritz describes it well in this article at SoundStage.

I'd just add that it makes sense to have specific references for certain price classes, in addition to our "absolute" reference which implies price is no object. It's not fair or useful to evaluate a $200 product and declare it inferior to our reference device selling for 10-30 times as much - an unfortunate cliche that I still see done in many "pro" reviews these days. So having a few different references at various prices seems reasonable.

I also feel - and this is a big one - that a reference component should be a currently available device, rather than custom built or only available on the second-hand market. If I review a pair of speakers and explain how they don't measure up to some DIY transducers that were custom built just for me, or a set of vintage speakers that are now basically unobtanium... that doesn't seem like a very useful conclusion. A possible exception to that rule would be something extremely popular and widely used. The original and now-discontinued Sennheiser HD800 is a good example, as nearly everyone has had a chance to own it or at least hear it at some point. But not many high-end DACs fall into that category. An original Benchmark DAC1 might, but the Invicta Mirus Pro Signature Edition certainly does not. Ultimately I suppose a reference doesn't necessarily have to be all that well known but it should at least be something a reader could actually purchase or demo if they wanted to.




Cen.Grand
As mentioned earlier, after nearly two years of extensive listening to a wide variety of excellent offerings, I ended up choosing the Cen.Grand DSDAC1.0 Deluxe as my reference DAC in the price-no-object category. Cen.Grand is a company that's been around in China for over a decade but is just now working to expand to other regions. During that time they have become well regarded for their various DACs, media players, and amplifiers. The company lays out their history here complete with many examples of interesting prior models - it really helps you get a feel for where the company has been and thus where they might be headed.

Company founder and chief designer JianHui Deng is an interesting character. He's a multi-instrumentalist, playing clarinet and saxophone and guitar in various settings throughout his life. He's also a talented designer and a very passionate audiophile, with these aspects combining to make Cen.Grand a compelling brand in the crowded world of high-end audio.

We discussed the history of HiFi in China and it was illuminating to hear all about it from an insider perspective. My takeaway is that JianHui has big plans for his company as well as a soft spot for the personalities in the industry, beyond just the gear itself. Think about names like Ken Ishiwata of Marantz, speaker designer Andrew Jones, Ed Meitner of EMM Labs, Paul McGowan of PS Audio, IEM gurus like Jerry Harvey and John Moulton, digital wizards like Bruno Putzeys and Andreas Koch, or John Curl and Rupert Neve in the analog realm. The list could go on but you get the point - these folks are responsible for a number of individually excellent audio products, but have also furthered the art in general, becoming somewhat interesting figures along the way. I believe JianHui Deng intends to follow in their footsteps rather than remaining a somewhat anonymous "man behind the curtain" - not that there's anything wrong with folks who take that approach.

Cen.Grand has some very big and rather unique plans for the future as well. I can't go into further detail at the moment but it involves some very bold steps that should better position them in the hi-fi landscape for years to come. Once these plans are announced, I think everyone will agree it is an interesting development.



The DAC
Anyway, the DSDAC1.0 Deluxe is the highest model in a series of three versions Cen.Grand offers. There's the baseline option and then an upgraded "Super Clock" edition, both of which look identical from the outside and use enclosures very similar to the one used by Cen.Grand's Silver Fox headphone amplifier (which has quickly become one of my favorite headphone amplifiers, but that's a story for another day). The Deluxe DAC sits at the top of the heap and offers additional upgrades including an even more extravagant enclosure, significantly upgraded power supply with complete isolation from the rest of the circuit, plus separate transformers for analog and digital sections, more complete pre-amp capabilities including RCA and XLR inputs, and various other internal upgrades which the company doesn't explain in full detail. The Deluxe version sells for $6,219 and is available in North America from exclusive distributor Power Holdings.

Unlike many of the more compact D/A converters we often see in headphone rigs, the DSDAC1.0 Deluxe is a rather beefy full-size design that weighs in at over 25 pounds. The chassis design and build is immaculate, and the device would not at all look out of place in a system using top-caliber gear from Esoteric, dCS, EMM Labs, etc. I find it elegantly understated, and yet accents like the structurally-complex side panels (which some may recognize as representing a graph of the DSD quantization process) make it uniquely identifiable.



At its core, the DSDAC1.0 is a completely original design which handles D/A conversion in a fairly unique way. After working with off-the-shelf chips from Wolfson and AKM for years, JianHui Deng attempted to move up in the world by licensing technology from highly-respected audio firm Playback Designs. While Playback has occasionally licensed their DAC solutions over the years (see Nagra's HD DAC for one example) they ended up declining JiamHui's request for one reason or another. My guess is that Cen.Grand may have wanted access to the DSD-based technology itself rather than just using a pre-built module like Nagra, but that's purely speculation on my part.

Whatever the case, JianHui hit the books and worked on the problem for 5 years before arriving at his own proprietary solution. I can't say I completely understand the intricacies, due to a variety of factors including a slight language barrier, my own lack of electrical engineering degree, and likely a desire for JianHui to safeguard his proprietary intellectual property. But from what I can gather, the answer involves a combination of custom code running on a powerful FPGA chip, plus resistor arrays and shift registers, converting all incoming signals to high-rate DSD which is user selectable in several increments from DSD128 to DSD1024. The trick, as far as I can make out, is that the process is so tightly integrated. It's common for devices to initially apply upsampling, then send that signal over to a DAC chip for actual conversion, in a two-stage process. The upsampling can often be defeated if the user desires. It seems the Cen.Grand handles it in a more integrated fashion where the DSD conversion naturally and inseparably leads to the discrete 1-bit D-to-A stage - indeed the DSDAC1.0 cannot natively process PCM at all.

Also integral to the design is more proprietary tech in the form of Cen.Grand's "synchronous direct clock technology", which involves a femtosecond clock feeding directly into the shift register array itself. Cen.Grand explains that most other designs, whether using an on-board femtosecond oscillator or even an expensive external clock, require the use of a frequency divider which adds jitter to the equation. Their solution side steps that issue altogether by once again taking a more unified approach.

Speaking of timing - yet another bit of proprietary tech is their "clock blocking" circuit which completely strips away all timing information from incoming signals. This makes sense when combined with their synchronous clocking mentioned above. As a result, Cen.Grand claims transport quality is far less critical than it otherwise would be with a DAC of this caliber. For the most part I find this to be effective but I'll discuss it in more detail shortly.

As far as user interactions go, the DSDAC1.0 lets us cycle through our choice of various digital filters, which (as is almost always the case) make an extremely small impact on sound quality. More important is the ability to select the rate of DSD used in the (necessary) conversion, which causes more noticeable changes to the presentation. The device can be set to fixed volume or variable output mode, which leverages the same switched-resistor solution used by Pass Labs in several of their expensive preamplifiers. Those with turntables (or other analog gear) in their rigs can take advantage of the high caliber preamp capabilities which the Deluxe has to offer, either via XLR or RCA inputs, with no digital processing involved. In those cases the display helpfully switches to larger text showing the volume level and selected input. A nicely built bundled remote also makes this a more viable preamp than those DACs which only work via manual adjustment.

On the digital side, the Deluxe covers all the usual bases including coaxial, optical, BNC, AES, and USB. Glancing at that input section, one might notice the small rectangular blank panel taking up space for no apparent reason. That's actually the spot for future upgrade modules. JianHui has an idea for a proprietary twin-cable connection system that will be featured on his upcoming media player transport device. I've also discussed the option of I2S over HDMI and while he initially seemed resistant (based in part on his own twin-cable plans as well as the existing clock blocking technology elevating the viability of the existing inputs) I think he warmed to the idea enough to potentially offer that option in the future. Unfortunately it seems like the updates will require a trip back to Cen.Grand HQ. So it would seem to not be a truly modular design like those from Sonnet or Schiit, where the end user just plugs in the new card. But I could end up being wrong on that.





Evaluation Gear
A quick word about the equipment I used in evaluating this DAC. You can already see much of it in the pictures, but in case anyone is curious: transports ranged from a Surface Pro 8 to a Euphony Summus music server with Keces P8 linear PSU, along with a Stack Audio Link II Ethernet bridge with matching Stack Audio Volt PSU. Amplification was the Cen.Grand Silver Fox, Niimbus US4+, Cayin HA-6A, Pass Labs HPA-1, Musician Audio Andromeda, Cayin HA-1A mkII, and a custom hotrod KGSSHV electrostatic amplifier. Headphones used included the Audeze CRBN, Audeze LCD-5, Audeze LCD-24 Limited, ZMF Caldera, ZMF Atrium Closed, Meze Elite, Focal Elex, Sendy Peacock, 64 Audio A18t, Ultimate Ears Reference Remastered, and the Radioso from AME Customs. The system is fed balanced power from an Equi=Core 1800 conditioner, and all cabling is from Audio Art with the exception of some headphone cables from Effect Audio and an iFi Gemini3.0 USB cable.

The Sound
Simply put, the Cen.Grand DSDAC1.0 Deluxe is the overall best DAC I've ever experienced. It actually sounds fairly similar to my beloved Resonessence Labs Mirus Pro Signature Edition in many ways, which helped make for a quick transition as I moved from one reference to another. But I also hear improvements in key areas such as timbral accuracy, liquidity, soundstage depth, instrument localization, and tonal weight, which collectively elevate the experience to new heights. Anyone who has heard me rave about the Invicta platform and its various iterations over the past decade will understand how significant that statement is. I've heard and reviewed many, many DACs during that time and nothing quite offered enough to pull me away from the Resonessence Labs sound... until now.

I consider the DSDAC1.0 Deluxe to have a very neutral and natural sound, with the caveat that people have a hard time finding consensus on what exactly "neutral" sounds like when it comes to DACs. Most would agree that it falls somewhere between the extremes of a brightly etched Benchmark DAC1 and an exceedingly warm BorderPatrol SE - beyond that I find opinions vary pretty wildly. The Cen.Grand, like my previous reference from Resonessence Labs, presents right down the middle and does not feel obviously tilted in either direction. This generally tends to be the case with most of the quality DACs I've auditioned lately, though exceptions do exist as I'll discuss later.

Having said that, the standout traits, the overarching character of the Cen.Grand DAC centers upon being extremely smooth, organic, fluid, and effortless. At first listen I think most people would not immediately point out its superb clarity or massive soundstage - though both aspects are certainly true. Instead I think they would notice how easily everything flows, which is quite unlike many other DACs on the market. Only after having worked that out would most folks then make the connection that the levels of detail and openness are also extremely impressive. Which is part of the charm, as usually something has to be sacrificed in order to produce such high levels of grace and fluidity.



The sonic signature of the Cen.Grand DSDAC1.0 Deluxe is thus a study in contrasts. Smooth and unforced yet extremely resolving. Wonderfully full-bodied yet not bloated or overdone. Superbly nuanced yet bold and dynamic. It paints vast sonic landscapes with lifelike spaciousness, but can also zoom in on microscopic minutiae at will. It's just an incredibly capable DAC which really sets a system up for success - supremely resolving and yet somehow not as picky as certain other high-end models I've owned/auditioned over the years.

This leaves the listener free to meander through their library uninhibited by worries over recording quality. Top-caliber releases from Reference Recordings, DCC Compact Classics, Three Blind Mice, or Blue Note? They sound vibrant and lifelike, believably placing the performers in the room with us as we listen. But how about less-than-pristine material from your favorite obscure artist? Cen.Grand says bring those on as well. This is not the DAC that tip-toes around your collection, only working well with the absolute best of the best. With the Deluxe DAC I heard album after album sound better than it ever has in my system, from high quality releases to the mediocre, and even some downright bad ones too.

Of course, the Cen.Grand can't make a loudness-war-casualty like Metallica or Red Hot Chilli Peppers sounds like a hi-res Coltrane or Pink Floyd release. But what I'm saying is that it makes the most of whatever it plays. For example, at this moment I'm listening with a Euphony Summus music server feeding a Stack Audio Link II Ethernet bridge, the DSDAC1.0 Deluxe, the Cen.Grand Silver Fox headphone amp, and an Audeze LCD-5. That's an extremely resolving chain which might theoretically discourage one from selecting anything less than pristine, audio-show-worthy music. And yet here I am enjoying all sorts of "non-audiophile-approved" material from the likes of Chthonic, Streetlight Manifesto, The Mother Hips, Solar Fields, Pinback, Aesop Rock, and Sufjan Stevens. My Mirus Pro Signature Edition would have made it very clear how inferior some of these choices were, yet the Deluxe DAC somehow doesn't object as much, while at the same time not holding back when scaling up with better recordings.

My first reaction was that perhaps the Resonessence Labs DAC was in fact the more discerning of the two. But after careful back to back listening over many hours, I now feel the opposite is true. The Cen.Grand seems to work with the music instead of judging it harshly, if that makes any sense. So while I do hear the deficiencies when they exist, it doesn't pull me out of the listening experience as easily as it can with other DACs. I'm able to just enjoy the music for what it is rather than focus on what it isn't. That sense of ease is hard to reconcile with the level of insight I hear being excavated by the Deluxe DAC - I've never quite heard anything like it. Maximum detail is unleashed, but presented in such a liquid and beautifully flowing way that it never offends the senses. It really is a unique and stunning achievement.

Another aspect which I love: the soundstage and imaging of the DSDAC1.0 Deluxe are incredibly open and lifelike. Certainly among the absolute best I've heard at any price. If the associated amplifier, headphones, and choice of music is up to the task, the listener will be treated to layered, holographic staging and superb localization. There have been multiple occasions where I was home alone listening on headphones, and could have sworn someone had walked into the room with me. I quickly remove my headphones and turn around - nope, it was just in the recording. This is the type of experience that may sound silly or contrived to read about, until you experience it for yourself.




Cen.Grand lets users choose their level of DSD upsampling, starting from the lowest possible setting of DSD128 and topping out at DSD1024. This makes incremental but appreciable differences to the resulting sound, and to my ears gets continually better as you move up to the higher rates. As I cycle from low to high, I hear the previously discussed sense of ease getting more pronounced with each step. To be clear, it doesn't seem to be darker, or more rolled off, or anything of that sort. It rather makes the whole affair seem more liquid and serene, with no corresponding reduction in low-level detail retrieval. The differences do seem to diminish as we go higher, with the jump from DSD128 to DSD256 being more significant than the final jump from DSD512 to DSD1024. I see no reason to use anything other than DSD1024 at all times since it delivers benefits with no drawbacks as far as I'm concerned. I suppose it might be possible to assemble an extremely laid-back system which lacks immediacy to the point where it does better with the lower rates, though I have not encountered such an animal during my travels thus far. Again, the lower settings are not at all brighter, more crisp, or more extended, nor are the higher settings muffled or dark in the least. It's more a difference in fluidity or the natural "flow" with which the music hits my ears - the higher DSD settings just feel more organic and supple.






Going back to the "clock blocking circuit", I generally agree that transport quality matters significantly less here than with many other DACs. I don't notice any drastic differences between the various input options, which is quite handy as some other DACs really only shine via specific inputs like AES or USB or I2S. The DSDAC1.0 leaves it open for equally great sound, even from "lesser" options such as Toslink or coaxial. I also like having BNC available as that seems less common these days, and I generally appreciate the physically robust connection it brings.

Thanks to that clock blocking technology, I also achieve very enjoyable sound running Roon via USB straight from a pedestrian source like a Surface Pro 8 - which can be pretty hit or miss on competing devices. Using my Euphony Summus music server directly, or the previously mentioned Stack Audio Link II will give modest improvements, but you have to listen pretty closely to spot the minor uptick in high-frequency resolution, imaging, and spatial accuracy. Ultimately I'd probably still recommend getting your hands on something better than a basic PC or Mac for transport duty, just to squeeze out the last bits of potential from the Deluxe DAC. But I would make sure to dial in all the other aspects of the system prior to fussing with that part.

As far as system pairing, the Deluxe proves a friendly partner with all manner of supporting gear. I used it most with the matching Cen.Grand Silver Fox headphone amplifier, which is the obvious match JianHui had in mind when designing the device. But I also found excellent synergy with the Pass Labs HPA-1, Niimbus US4+, Cayin HA-6A, and my custom built KGSSHV electrostatic amplifier. The DSDAC1.0 Deluxe seemed capable of unlocking each amplifier's full sonic potential despite the wide range of sonic voicing involved. It is definitely not a "specialist" DAC that works best only with certain amps/headphones/music types.



I also had a great time breaking out some lower priced amps which wouldn't normally be paired with sources this good. The Musician Audio Andromeda and the Cayin HA-1A mkII were both stunning performers, sounding better than they had any right to for their sub-$1k price levels. I don't believe I had ever heard either model elevated to such a high level as I did with the Deluxe DAC. Two very different signatures - the Musician is extremely fast and precise whilst the Cayin captures the essence of tube bloom and full-bodied warmth - with both being equally enjoyable in its own way. Again, pairing a $6k+ DAC with a sub $1k headphone amp might seem a bit silly to HeadFi folks, but I imagine someone who primarily listens via speakers could use either amp to add highly capable headphone playback to their setup without investing a fortune.


When you get to this level of performance, it really can be tough to break down the exact sonic character of a DAC. True, there are some high-end models which I could more easily summarize with terms like "smooth" or "fast" or "dark". But for the best DACs out there, much more complex and nuanced language is often required. So rather than drone on and on about the DSDAC1.0 Deluxe and how much I love it, I figure it might be better to talk about some of the other excellent DACs I've tried lately, and where they sit in relation to my new reference.



The Quest
Again, this was a journey I embarked on long before discovering Cen.Grand. My goal was to spend as much time as possible with a wide variety of high-end DACs that interested me for one reason or another. Anything that seemed like a potential contender to my prior reference, and that I could actually gain access to, was included. Some of these ended up being very well known while others were so obscure that I had never heard of them prior to this project. Note that I didn't let that factor into my ultimate decision one way or another.

Reflecting on all the reviews I've done over the years, many of the ones I tend to be most proud of are those which involved extensive comparisons. For example my Niimbus US4+ and Cayin HA-300 have impressions on many different amplifiers and headphones, to the point where they are hopefully useful to a broader audience - even if the reader wasn't all that interested in the actual Niimbus or Cayin devices themselves. I aspired to follow that same format in hopes this project will also be a good resource.

In order to demo all the DACs I was interested in, I took advantage of any opportunity I could, even when that didn't end up being an ideal listening situation. Thus some of these involved the DAC being in my own system for a decent amount of time, while others simply had me listening at a friend's house or even an audio dealer until they kicked me out. As far as budget goes, I avoided those with price tags approaching new car levels (or beyond). That means no dCS Vivaldi, Esoteric Grandioso, MSB Select, or others of that nature. The bulk of my interest ended up being DACs around the $3k-8k range with some models escalating a decent amount higher. I did treat price as a factor; not in terms of "more expensive is likely better" but rather "the models selling for significantly more must do something substantial to justify that increase". So any contender priced far above the rest had a steep but not insurmountable hill to climb.

Before diving in, keep in mind that this is not intended as a definitive ranking or full evaluation of each DAC. Given the differences in listening scenarios for many of these devices, it's certainly possible that some which I found mediocre would have performed much better with different gear, or when used for a longer period in my own system. I'm not trying to step on anyone's toes, nor am I trying to tear down your favorite product/brand. I'm simply reporting my experiences and ultimately explaining why I chose the Cen.Grand over all these worthy competitors. There are a few in the mix which I did do a formal review on, and I'll link to those whenever appropriate. Prices listed were accurate at time of writing, and reflect any options involved in the unit I tried, but expect some fluctuation there.

Anyone who doesn't care to read my long-winded ramblings can just skim through, taking a look at the devices involved and noting that I prefer the Cen.Grand to all of them for one reason or another. For anyone who reads this entire thing, I appload your patience.

That said, let's get to the DACs, in no particular order.



Chord Dave ($14,000)
I've gone back and forth with my feelings on Chord DACs ever since the original Hugo launched some years ago. They tend to strike me as impressive technical accomplishments which ultimately don't quite fit what I want out of my DACs. Some are better than others and I'm sure a search of my HeadFi posts will uncover moments of whiny complaints as well as favorable compliments, probably in equal measure.

The Dave DAC, I must admit, is an impressive sounding device. It gets deep into the mix and unravels complexities with ease. It tends to be more revealing than is useful at times, but with good recordings it can be a revelation. It can also slam hard when called to do so, which might surprise as the signature can also sound thin at times - but I'm convinced Dave is mostly just showing what's in the mix, for better or for worse.

My biggest sonic complaint is that Dave doesn't usually portray timbre in a convincing enough way. And unfortunately that's a deal breaker for me. The Cen.Grand can't quite match it for sheer resolution (though it trails much closer than I anticipated) but it sounds more natural and lifelike, whilst Dave often strikes me as somewhat plastic or artificial. And I need instruments to have that realistic tonality, or else I don't really care about their other technical aspects.

I'll also add that the Cen.Grand liquidity I rambled on about earlier is pretty obvious when compared to Dave. The Chord has technical excellence but somehow it tends to feel like a bunch of puzzle pieces trying to properly fit together in order to portray a sonic picture. Meanwhile the DSDAC1.0 Deluxe is so coherent, everything just naturally lines up.

Beyond that I find the Chord pricing way overdone, the ergonomics quirky, and the physical styling not to my taste. Those are all personal things but they help cement my opinion that Dave could never be my reference DAC, when others sound and look more appealing in various ways - and for significantly less money.

I've been told over and over that I need to add the Chord M-Scaler to get the full tonal impact Dave offers. All I can say is that if a $14k DAC needs another $5k+ add-on just to sound convincing, mistakes were made somewhere. Especially since Dave existed as Chord's top offering for several years before M-Scaler came along. If both pieces are truly required for the proper experience, just bundle them together and sell a single $20k device.



Matrix Audio X Sabre 3 ($3,000)
A beautifully constructed DAC with very well done integrated streaming on board. Generally neutral sound with the default settings though it can feel a little bright at times. Using its unique SYNC mode and feeding it a steady diet of upsampled DSD material via Roon/HQplayer brings out a more relaxed and natural tonality. While not quite up to the same level as my old Resonessence Labs reference or the Cen.Grand Deluxe - not a surprise given it sells for less than half the price - this is nonetheless a very enjoyable DAC which offers gobs of bang for the buck and could legitimately be an "end-game" DAC for most sane people. My full review is here.



Wavelength Callisto ($10,000)
My time with this was unfortunately more limited than I would have liked, but from what I did experience this is a truly unique and exceptional DAC. Vibrant and rich, the sound reminded me a bit of my old Wavelength Cosine from many years ago, cranked up to an entirely new level. Both had beautiful midrange sweetness and a level of "purity" to the presentation which remains elusive in most other DACs. The old Cosine was sometimes overly dark and I did get just a small hint of that here as well, which is not really shocking considering the NOS design with triode output stage. Still, the directness and tonal purity of the sound is possibly the very best I've heard, particularly with jazz and classical music. The Callisto can also totally rock when called upon to do so, with extremely satisfying dynamics and drive. Due to the somewhat forgiving NOS nature and tube bloom, this is perhaps not the ideal DAC for evaluation purposes, with its priorities being more like that of an Audeze LCD-4 than an LCD-5 if that makes sense. But the sound is so unique and enjoyable that I could definitely still make it work for my needs.

My list of downsides really is vanishingly small from a sonic perspective, and I would happily use this as a reference even with the $10k price tag being a bit higher than I would like. Unfortunately I later found out the example I heard was one of the last ones in production. Wavelength Audio founder Gordon Rankin is a genius at building superbly engineered DACs and amplifiers, but they are all hand built in somewhat limited runs. So I had to take the Callisto out of contention based purely on my previously identified criteria of "current products" which readers could actually acquire if they wanted to. If not for that, I would love to get one of these in my system for direct comparisons to the Cen.Grand Deluxe. I feel like it would be a close competition with the DSCAC1.0 likely being more expressive and the Callisto a bit more euphonically charged - both extremely capable in their own very enjoyable ways.

I haven't seen much (or any?) discussion or reviews of this DAC in the wild. If you ever get a chance to hear it, I strongly encourage you to give it a go. This is a world-class DAC as far as I'm concerned, and a possible contender for my reference if not for the limited production run.



DiDiT DAC212SE ($4,500)
For some reason I thought this Dutch DAC was a bit larger. In reality it is an absurdly compact and well made little device, which brings to mind Nagra or Weiss Engineering in terms of its precision build, but much smaller than anything they offer. The sound is actually in that same camp as well - very fast, incisive, clear and uncolored, it is highly revealing of any flaws in the chain. So much so that I found it veering towards being slightly analytical and dry at times, depending on the music and choice of headphones of course. I've certainly heard worse offenders though, and overall this is still an impressive DAC.

With sharply defined leading edges and very accurate imaging, its technicalities approach some of the best I've heard, even if it does sometimes leave me wanting as far as tonal density is concerned. This can throw off timbral accuracy with certain instruments, which results in this DAC feeling a little genre dependent. Electronic music and metal sound great, whilst things like classical or jazz make the timbre issue more obvious. The Cen.Grand Deluxe scratches that itch for richness whilst compromising nothing in terms of speed or detail, which is why I would place it on a higher tier than the otherwise excellent DiDiT DAC. I thought perhaps the DSDAC1.0 would feel a bit dark after switching from the Dutch DAC but that never ended up being the case. Instead, it felt "just right", and only reinforced my opinion that the DiDiT DAC was a little on the bright side.

I wasn't looking for a device with integrated headphone amplification but since the DiDiT has that functionality I went ahead and tried it. Surprisingly, I found it highly capable, and potent as well! I didn't expect much from such a tiny integrated unit but this thing paired exceptionally well with the Audeze LCD-4, Meze Elite, and Focal Stellia, which were the only headphones available to me while trying out this device. Again the sound was quick, precise, and illuminating, with a bit of top-down perspective. I imagine it isn't a universally appealing sound, as it likely wouldn't synergize very well with brighter headphones like LCD-5 or Utopia. But the ones I did have on hand were deliberately curated to make optimal pairings and the result was impressive.

Like the Matrix X Sabre 3 mentioned earlier, I don't think this DAC is quite up to the level of the Resonessence Labs or Cen.Grand devices. But it is nonetheless a compelling option for those seeking an all in one solution in a very small form factor, and I would choose it over the Chord Hugo TT2 which seems like a logical competitor.



Musician Audio Aquarius ($3,200)
The Aquarius is a somewhat unique sounding R2R DAC. It's from the same folks who brought us the Denafrips devices, but has a distinct signature that is entirely separate from the established Denafrips house sound. I'd call it better balanced, more nimble, and generally more technically accomplished, whilst dialing down (but not completely eliminating) the warmth and smoothness of its better-known Denafrips cousins. That said, it is still richer and more full-bodied than the DiDiT DAC mentioned above. It's just a very capable source that works well in almost any scenario, with a gentle focus towards helping bring out the best rather than pointing out flaws.

The Cen.Grand DSDAC1.0 Deluxe shares a similar sense of ease to the presentation, but achieves it through different (and more impressive) means. When listening to Aquarius directly after the Deluxe, I am always aware of a slight treble shade, a subtle darkening to the sound of the Aquarius which takes the edge off difficult recordings - a generally welcome thing - but also limits better recordings from truly sounding their best. There's a bit of rounding to leading edges which helps tone down unruly material, again at the expense of the details which I do want to hear. This initially became apparent in direct comparisons, but once I heard it I could not unhear it. The Aquarius seemed to actively be working to make everything sound a bit better than it should, at the expense of some realism/honesty. The DSDAC1.0 is not so obvious, and manages superior extension and transient impact without losing its nuanced touch. It's hard to explain but I guess the Aquarius has more of a smooth/liquid "sound" whilst the Cen.Grand has more of a smooth/liquid "feel" if you catch my drift.

This is another one of those "value" devices that would make an excellent choice for most people - seriously, this is more than enough DAC for almost any sane music lover. It does pretty much everything right and is only overshadowed by a few particularly worthy opponents, all of which cost significantly more money. See my review for more info.



Lab12 DAC1 Reference ($3,290)
With the word "reference" baked right into the name, one could be forgiven for expecting a signature in the same camp as Benchmark, exaSound, or RME. Instead this thing reminded me of vintage Luxman gear - a warmish, soft, forgiving, romantic glow with very obvious euphonic coloration. Which makes total sense once we learn about this being a non-oversampling design based around 8 of the old TDA1543 chips plus a tube output. I can see why some people are drawn to this thing though, as it certainly is a unique sound that will stand out in a crowd - it would not be easily confused for any of the other DACs on this list. I could build a superb audio system with this device at the heart of it, but I could also assemble some pretty bad matches too, and it wouldn't be meaningful to evaluate a headphone amp based solely on how well it pairs with this unique (and at times admittedly gorgeous sounding) device.

Comparisons to the Cen.Grand Deluxe DAC don't even really make sense here. The Lab12 is significantly darker and more rolled off on the top end, lacking the same sparkle, precise imaging and borderless soundstage. It also comes across as being a bit slow and plodding, mainly in direct comparison of course. The midrange is very engaging in its way yet sometimes feels out of balance with the rest of the spectrum, and I would have to carefully choose matching components to take full advantage of the strengths whilst avoiding the shortcomings brought on by the deliberate tuning. The Cen.Grand is far more universally appealing. Again, the Lab12 is an interesting DAC for what it is, but in my mind not suitable at all for a reference component - which is fine, as that's not really what everyone needs anyway. People who enjoy quirky signatures and perhaps don't care about the usual metrics of ultimate frequency extension or inky black backgrounds etc may want to give this thing a spin.



MHDT Oolong ($2200)
Another old-school NOS design with tube buffer in the output stage, this time from MHDT Audio - a brand which was fairly popular with HeadFi users roughly 15 years ago. Back then $400-500 would get you an MHDT tube DAC with a rich, meaty presentation that sounded at times beautiful if not really all that accurate.

Fast forward to present day and we find that some things have changed while others have remained more or less the same. The Oolong is vastly more attractive and solidly built than those old models, and has a more competitive selection of inputs and output from which to choose. The low end is both deeper and more authoritative, where the old models always felt somewhat mushy to my ears. And the treble, while rolled off to some degree, is more tastefully done than any of those classic designs from way back when. Overall this is a far nicer DAC than its predecessors. As it should be considering the price increase involved.

Still, some of the same things I disliked back then are still present on the new model. While it does give a solid, punchy low end, I still feel dynamics are lacking overall, with a lack of contrast and an almost monotone feel to it at times. Soundstage is reasonably large but not particularly accurate. Tonality is warm and relaxed to the point of going beyond the Lab12 - which is already a fairly laid-back sounding DAC.

Frankly I'm surprised I enjoyed this thing at all given the amount of complaints I have. But it does have a certain charm to it under the right circumstances. The person who owned this particular unit actually had two of them: the first I heard was in his (otherwise very nice) headphone rig, and I didn't really care for the resulting sound at all. I've heard all his other components individually and know what they are capable of, so I can only conclude the MHDT was the weak link bringing everything down a few notches. It sounded closed in, lacked articulation, and felt like it crushed dynamics. I don't understand how he could possibly choose this DAC for his headphone setup, but he seemed happy with it. I won't even bother with comparisons to the Cen.Grand as there is no point - to my ears it does everything substantially better.

However, the other MHDT unit was in a speaker-based system paired with the superb Feliks Audio Arioso 300B integrated, driving a massive set of custom built horn speakers which looked somewhat similar in design and execution to the Tannoy Westminsters. That result was a much better pairing, sounding extremely tight and meticulously tuned to his room. His budget clearly allows for almost any other DAC he might want, and indeed he went through many before landing on the MHDT Oolong and sticking with it for the long term (thus far anyway). So despite being underwhelming in his headphone setup, I have to admit this DAC seemed perfectly matched in his speaker rig.

I guess the takeaway here is that no matter how poor a device might sound in one context, there may still be a home for it in some other situation. I am of the opinion that this DAC would not be to my liking in the majority of systems out there but I can't deny the fact that in at least one very specific case, it did a phenomenal job.



Atoll Audio ST300 Signature ($4900)
I haven't heard gear from French firm Atoll Electronique in ages. And when I did, it always seemed like generally solid yet unremarkable gear to me. So when a dealer suggested I try out the Atoll ST300 Signature, I was pretty skeptical. This dealer had already let me borrow their Merason DAC1 for a while, so I figured I would listen to the Atoll demo unit as well, out of politeness if nothing else.

I ended up being surprised and impressed by this thing. Yes, it's really intended more as an all-in-one streamer, DAC, and preamplifier, but it does not in any way feel like a compromise as so many multi-function devices can. In fact I liked this much more than the Atoll DAC300 which is their dedicated DAC (no streamer) using a completely different design. The ST300 Signature has a lively, exciting presentation which (mostly) steers clear of information overload. It sounds open, clear, and sparkly, yet also captures a good amount of body and solid note weight, making for a nicely balanced sound overall. Backgrounds are jet black, and imaging is very focused and precise. This thing feels like it would fit well in most systems regardless of what other gear may be involved.

I can really only find two things to complain about. First, I would say it can become fatiguing after a while, depending on the rest of the audio chain. Some care must be taken in system matching to avoid making this otherwise minor issue into a major one. It isn't as analytical as the DiDiT DAC212SE but does still lean slightly in that direction at times. Second, macrodynamics, or the "puchiness factor", are merely good but not great. Which is not a huge deal but considering the price is nearly $5k, I would have liked to have seen a stronger showing here. Especially considering how well some of the competition performs (Schiit Yggdrasil for example) at a lower price. Had it done a little better here the Atoll would have been an even stronger contender and likely a finalist on my list. (Note: I suspect the exchange rate is working against me here, and if the Atoll was priced around $3.5 to $4k it would be far more compelling. International markets can be tricky.)

Thankfully the DSDAC1.0 Deluxe is able to meet or exceed everything the Atoll does. I hear just as much speed and detail (and at times even more) without drawing attention to those attributes, and the sound is punchier, meatier, more tonally rich. This makes timbre come across as more accurate and convincing, yet the Cen.Grand remains exceptionally well balanced overall. In most systems the Deluxe DAC would have the clear advantage, though I could see a slower/more syrupy chain perhaps synergizing better with the Atoll, at least with certain music.

If someone loved the fast and ultra-resolving sound of a Rockna Wavedream but wanted to spend a bit less and pick up integrated streaming in the process, this Atoll ST300 Signature could be a great choice. I have not heard Rockna's Wavelight which sells for roughly the same price as the Atoll (but again lacks integrated streaming) so I'm not sure how that comparison would go. My preference remains with the Cen.Grand but Atoll Electronique definitely earned my respect with this experience.



Meitner MA3 ($10,500)
This is a comparison I had really looked forward to given the general similarities of the approach - both DACs use custom FPGA processing for proprietary 16x DSD upsampling followed by discrete 1-bit conversion, and both are built like absolute tanks. I had spent time with an MA3 when it first launched and was very impressed, to the point where I felt it almost beat my Resonessence Mirus Pro Signature. Almost, but not quite.

Fast forward a couple years and I figured it might be wise to revisit this device for a second opinion. I'm glad I did because this time around it was even more impressive. I'm not sure if it improved based on firmware updates or just synergized better with the gear I was using, but this time I would call it roughly on par with my previous reference from Resonessence Labs. Again, I remain thoroughly satisfied with that device in terms of sound, so the Meitner MA3 being on the same level means it could definitely qualify as a replacement. And the integrated streaming via Roon is a nice bonus too.

Compared directly to the Cen.Grand Deluxe, I hear the MA3 as being a bit softer in the bass region, which detracts from the sense of articulation on fast material like metal or trance. It's not as noticeable with other genres but still there if I listen closely. The general impression is that the Cen.Grand is a bit more full bodied and dynamic, though again it's not obvious until I listen carefully for it. Both DACs have superb clarity in the midrange and exceptional top-end extension, with the Cen.Grand going very slightly beyond the Meitner in treble clarity whilst remaining totally free of artificial grain.

Both devices are extremely fluid and organic, and both paint wide open sonic pictures, though the Cen.Grand is noticeably more well-defined when used with the best recordings and most spacious/accurate headphones. Apart from that scenario, the soundstage differences are small enough to be inconsequential.

There were some instances where the MA3 felt just slightly artificial when reproducing things like woodwind instruments, acoustic strings, vocals, or other sounds where every aspect must be just right to sound convincingly lifelike. I can't quite put my finger on what exactly went wrong, but I suspect it had something to do with leading edges combined with tone colors that didn't quite add up to my ears. Again, not a huge complaint, and not applicable with many types of music. But it still came up multiple times with the Meitner, yet not once did I notice it with the Cen.Grand.

Ultimately the biggest difference, resulting from all the small ones described above, is that the Deluxe DAC ends up feeling simultaneously more engaging and better refined. The choice then is obvious to me: moderately superior sound at significantly lower pricing? Yep, I'll take that option every time.

I did quite like the Meitner MA3 though, and definitely considered it a top finalist in my DAC search. Of all the devices I tried, it is perhaps the most similar in overall character to the Cen.Grand, which may be helpful for folks trying to understand what the DSDAC1.0 is like in general.

I have spent a little time with the significantly more expensive EMM Labs DA2 (in the latest V2 form) and I would say the DSDAC1.0 Deluxe reminds me very much of that unit. At least based on my admittedly limited impressions. My opinion at the time when trying the DA2 was that of a supercharged MA3 - similar tone but generally superior in various small ways - and that's the same vibe I get from the Cen.Grand Deluxe. But I did not get enough time with DA2 to draw any solid conclusions, nor to include it in this article, so I only mention this as a general aside.



dCS Bartok (pre-Apex version at $16,000)
I had high hopes for the Bartok but have tried it out multiple times and never been impressed. Especially considering the price involved. After yet another audition, this time in my own system, my opinion has not changed one bit. I do think the dCS folks are brilliant engineers who obviously go all out with their craft, but unfortunately the end result in this case just doesn't resonate with me at all. I find the sound entirely too "safe" and boring, lacking the engagement factor which many other DACs on this list possess. Which is strange, as the various dCS DACs I've tried over the years all had a distinct house sound which was very different from this. They were not perfect, and didn't always suit my sonic preferences, but I appreciated their unique viewpoint for what it was. The Bartok almost seems as if it came from an entirely different company altogether.

In direct comparison to the DSDAC1.0 Deluxe, the dCS Bartok feels somewhat gray and undefined, lacking in both authority and "bite". Transients come across as soft and indistinct. Resolution is mediocre. Focus is smeared. Instrument localization is decent but does not stand out. The Cen.Grand has significantly better midrange projection, superior microdetail retrieval, a much more defined sense of the performance space, and a thicker, more palpable tonality which equates to more believable timbre. There really is no competition as far as I'm concerned.

Note that the Bartok is not a "bad" sounding DAC in absolute terms. It doesn't do anything terribly wrong, per se. In fact its biggest strength seems to be a lack of truly egregious weaknesses, and if that's what dCS was shooting for, I guess they succeeded. Again, I'd call it a very "safe" tuning which does not offend but also doesn't stand out in any particular attribute. In this field of very capable devices, the Bartok just doesn't do anything to elevate itself anywhere near the top of the pack from a sonic perspective. Factor in the high price and you can see why I'm not very enthusiastic about it, apart from perhaps a build quality standpoint.

I actually think there's a great analogy to be had just by physically examining the Bartok. The construction and materials are undeniably high end, a fact which is evident to anyone who interacts with the device. In this area, it is not embarrassed by any competitor, regardless of price. But from an aesthetic standpoint it does seem a bit unexciting if not downright generic. See other dCS models like the Rossini, Debussy, Vivaldi, Elgar, or Puccini and note the unique design language each device had. Love them or hate them, they were not easily confused with any other brand. Now take a look at Bartok again. It's certainly not ugly, but neither is it visually striking or even very recognizable in terms of design theme. It's just very safe. It actually reminds me of my old Esoteric SA-10 from around 15 years ago - another very well built device with fairly anonymous styling which could have belonged to any number of brands. I know looks are very subjective but when I compare Bartok to the equally well built DSDAC1.0 Deluxe, and note the far more interesting aesthetic detail on the Cen.Grand, it really stands out to me. But of course so do the pricing and significant sonic advantages.

Note that dCS has since launched their "Apex" upgrade, but I did my evaluation before that existed. It brings the price up to over $20K. I won't comment on it as I have not yet heard it, but do I hope it transforms the DAC into something more competitive.



ModWright Oppo 205 (originally $3,795 but worth lots more now)
Modwright Instruments takes the much-loved Oppo UDP-205 universal player and totally upgrades the entire signal path for the audio side. There are tubes involved, and Lundahl transformers, and a massive outboard power supply, and many other details such as custom ModWright capacitors, all of which results in the already great sounding UDP-205 transforming into what I consider a top-tier performer. Rather than borrowing a review loaner, I actually purchased one for myself and wrote about it for Darko Audio, and I still enjoy it on a regular basis. I also have a small fortune invested in tube upgrades which helped further tune the sound perfectly to my liking.

I've always known this device to be a giant-killer, so it was interesting to contrast it with the Cen.Grand Deluxe. Both have a vaguely similar sound overall - smooth, rich, unforced, and organic, yet with excellent detail, imaging, speed, transient response, and other technicalities. My first impression was that of the pair sounding more alike than different, with both being extremely enjoyable.

As I logged more hours with the Cen.Grand, I began to notice certain things which it does in fact handle better than the ModWright. I'd say the main improvement involves even more fluidity, with a greater sense of ease to the presentation. The Cen.Grand feels more confident reproducing absurdly busy passages or virtuoso performances as if it was no big deal. Picture a world class powerlifter bench pressing what is, to their capabilities, the relatively small amount of 400 pounds. While being a very impressive lift even for dedicated gym rats, this amount is actually hundreds of pounds below our powerlifter's maximum. Thus the feat is made to look easy - almost effortless - making us forget what a huge accomplishment it actually is. The ModWright Oppo still nails the lift but you can sense it bumping up against the limits of its capabilities relative to the more accomplished Cen.Grand.

The other key difference is the treble clarity of the DSDAC1.0, which is subtly but meaningfully superior. Brush strokes, trailing edges of cymbal strikes, and various overtones of all types sound more delicate and lifelike with the Cen.Grand in the chain. This leads to a more convincing portrayal of "they are here" performers being live in the space with me. The ModWright remains exceptional but I think we are again bumping up against the limitations of what it can accomplish here. No shame in it though, the ModWright still remains competitive or downright superior to almost every other competitor in this roundup.

On the flip side, the ModWright sounds more rhythmically bouncy, punchy and bombastic. It's got more energy, more drive, whilst the Cen.Grand feels comparably reserved and buttoned-down - despite its supreme capabilities. This part is largely related to the deliberate tuning I've settled on via extensive tube rolling. After spending an embarrassing amount of money over the years, I arrived at a pair of vintage Amperex 7308 gold pin tubes in the output stage, and a Philips 5R4GYS rectifier, to dial in the ModWright just so. In this case the Cen.Grand has a more dignified presentation, as if to say it could do what the ModWright does - if it wanted to. It's got enough meat on the bones to avoid sounding the least bit thin even in direct comparison, but can't (or deliberately doesn't) match the impact of the ModWright.

Unfortunately as many readers know, Oppo Digital got out of the audio/video business a few years back. So this device is not suitable to be my reference, just based on my self-imposed criteria of availability. It would certainly be a finalist if not for that issue. Interestingly, the demand for Oppo 205 units has only increased since the company exited the market. Factor in the steady increase in tube prices lately, and this is one of the few pieces of audio equipment I own which has actually significantly appreciated in value beyond the original purchase price. To acquire a mint new UDP-205 these days (if one can even find such a thing), then do the ModWright treatment (it still appears available on their website but that could be an oversight), and then add NOS tubes of the highest caliber, would very likely cost more than double what I paid a few years back.



Esoteric N-05XD ($11,000)
I loved the old Esoteric gear from way back. I still have very fond memories of the P-30/D-30 combo from the mid 1990s, and the evolutionary P-70/D-70 that followed. The X-series of the early 2000s was similarly exceptional. After that point their sound started deviating from my preferences, and I lost interest despite still having huge respect for the brand. The build quality and appearance of the subsequent K and D series models was superb, but the sound never really clicked with me.

The N-05XD is part of a new generation of Esoteric products which starts fresh using their new "Master Discrete" DAC implementation rather than the prior AKM chips. And to my ears this is a vastly better sound. It's much more full bodied and well rounded than the previous models which always felt a bit thin and brittle to me. N-05XD also gets integrated streaming, preamp capabilities, and even a balanced headphone output, though I mainly focused on the device as a dedicated DAC.

I'd say the N-05DX does qualify as a top-tier DAC, though I'm not in love with that five-figure price tag. It has a big, bold presentation that retains enough grace and refinement to sound universally appealing. Those who loved the old Esoteric sound might find it overly brash but to me it seems far more lifelike and realistic, and I think most listeners these days would agree. Even the old flagship K-01, which sold for around $22k a decade ago, feels dry and papery in comparison to the N-05DX. Anyone who was turned off by the prior direction Esoteric went should consider revisiting if they get a chance - it's a completely different experience.

Despite those improvements, the Cen.Grand is even more punchy and flowing, with greater dynamic contrast. The Esoteric sounds very slightly rounded at times, almost in deliberate backlash against their former clinical approach. This means less transient snap than the Cen.Grand and an almost "mellow" feel to the presentation. The Cen.Grand is liquid smooth yet manages to be more insightful, with a deeper view into the mix which somehow remains fatigue-free. The only time I might prefer the N-05XD is when using an otherwise overcooked system with a sharp amplifier and bright headphones - there the greater damping of the Esoteric would come off as superior control and probably do the system a favor. In any other context I prefer the Deluxe DAC by a reasonable margin.

Still, the Esoteric N-05XD is a superb DAC overall. It's everything I feel the dCS Bartok should have been, at a far lower cost, and with a much more striking appearance. My complaints about the value proposition might be abated by using the N-05XD as an all-in-one, since the headphone output seems quite well done from my limited use. I believe the streaming aspect was co-developed by Korean firm Lumin, which means it should be of a very high caliber as well. So while it was not my first choice, this thing ranks fairly high on the list overall.



Lampizator (various prices)
I was able to demo three different Lampizator models over the course of this project: the Baltic 3 (~$6,500), a balanced Atlantic TRP (~$10k), and a Big Seven mk2 (~$10k). That represents something like the entry level to mid range of the current lineup. I believe there is at least one slightly less expensive model available, and several higher end offerings beyond the Big Seven, priced into the stratosphere. I'm also aware that Lampizator is constantly updating things and releasing newer iterations. I'm not sure if the models I heard are considered "current" at this point, but since they were all sold new within the past year or so, my impressions should still be relevant.

I came away from my Lampizator experience with mixed feelings. On the one hand, I heard a common thread between all three units, something like a "house sound" that I believe Lampizator was shooting for when building these devices. And it certainly has some charm. Warm, rich, flowing, and nicely detailed, I absolutely get why so many people love their gear. Despite trying three models, with each one being meticulously tuned with the owner's choice of specific tubes, they all ended up sounding remarkably similar, with just subtle improvements as I moved up the line to the more expensive models. I suspect the single-ended Atlantic TRP (which shaves thousands of dollars off the balanced version) may be the best value of the bunch.

That said, despite the three different models and their various tube rolling selections, to my ears all of them seemed a bit overly dull on the top end. Detail retrieval was excellent in general (particularly with the Big Seven), yet treble was somewhat pushed down into the mix, or given lower priority than I'd like. Transients felt a little rounded which made the sound less expressive overall, and the soundstage, while spacious, was less distinct than the Cen.Grand presentation. Lastly, there seemed to be a mild timbre-related issue, moreso on the Baltic but still present in the other models, where instrument tonality felt a bit off. I can't quite put my finger on what exactly caused this but I did not hear the same issue on the Cen.Grand, nor the Meitner, Esoteric, Wavelength, or ModWright units. So ultimately while I could certainly build a nice rig out of any one of these three DACs, I didn't consider them particularly compelling for my reference use.

I can already hear the disagreement now. A legion of Lampi-lovers calling me deaf or accusing my associated gear of not being up to the task. When a brand makes so many rabid followers, they must be doing something right... and yet in this case that "something" just doesn't seem to line up with my needs or preferences.



Soul Note D-2 ($9,000)
Impeccably built and uniquely styled, Soul Note DACs are highly regarded in Japan but have little exposure here in the USA. This thing has some interesting features like quad ES9038Pro chips and the ability to switch to mono mode - allowing users to run a pair of them in monoblock mode just like the flagship Esoteric Grandioso DAC. It's a seriously impressive device - fully balanced, dual mono, massive chassis, very complex internals, with attention paid to the finest details such as vibration control and damping. This thing seems a natural competitor for fellow Japanese brands such as Luxman, Esoteric, and Accuphase, at least in terms of design and build.

After extensive listening, my impression of the D-2 is that it does have what I've heard referred to as a very "Japanese" tuning. That means emphasis on airy, delicate treble, beautiful tone, and a somewhat polite, reserved tonality. It's a highly technical and complex signature which manages to reproduce challenging sounds like trumpets with exceptional clarity and bite, yet without sounding overly harsh. Separation, or the space between instruments and the ability to pick them out while listening, is extremely well done. This is everything I wished my old Esoteric D-07x could have been, but wasn't.

The DSDAC1.0 Deluxe sounds richer, thicker, and more weighty in comparison to the Soul Note. It's one of those things where I don't necessarily notice how much I'm missing until I swap components, at which point the difference becomes striking. The Soul Note is beautiful in its own way, particularly with music like folk, acoustic singer/songwriter, solo piano works, or jazz trios. If that's your main musical diet, the Soul Note might be a perfect match. But I also enjoy all sorts of metal, electro, large scale orchestral works, reggae, etc, and for those genres the Soul Note feels less capable. Meanwhile I never feel like the Cen.Grand is too thick and rich to enjoy those lighter genres, but rather that it seems to balance everything very well and thus work with all music equally. And the Cen.Grand does not seem to miss any fine detail in the process, which to me is the most rare achievement of all. I could certainly work around the Soul Note sound by pairing the D-2 with suitably warm amplification and headphones, but that might miss the point - this DAC is extremely capable and deliberately voiced this way. Which for some listeners might be the ideal choice.

My other complaint about the Soul Note is that the USB implementation seems rather poor in quality. Which is disappointing since the device is otherwise so well designed and implemented. They use some unique data transfer method which requires special drivers and thus might not work with all Linux-based streamers. And even after fussing with drivers, the sound over USB feels somewhat flat and etched to my ears. I much prefer AES which is how I did the bulk of my auditioning. Given the ambitious design and the way Soul Note brags about their USB implementation, I can only conclude this was somehow an intentional choice, yet it baffles me to think that anyone would prefer this over the sound we get from AES or coaxial.

I didn't do my research beforehand, so I only discovered the D-2 having four software-selectable USB modes after the fact. Perhaps tweaking those settings would have made an improvement. Likewise I had no idea they allowed users to run the DAC in NOS mode, despite that term being right there on the front panel (in my defense, the text is quite small). There's a chance this may have altered the sonic experience in positive ways, but I missed my chance to find out one way or the other.

I've heard that "serious" audiophiles in Japan tend to run the Soul Note D-2 with fancy external clocking, and often in pairs for dual mono mode. Which may fix some of my complaints about politeness and tonal weight. But at that point we're talking at least $20k or beyond which is where I really lose interest. I remain intrigued by Soul Note in general, but the Cen.Grand feels like a superior DAC for my preferences, and for less money too.



Audio Research DAC 9 ($10,000)
I've never been terribly impressed with Audio Research DACs, and the DAC 9 did nothing to change that opinion. It's a fine sounding DAC, with a strong sense of resolution, good imaging, decent articulation, and a firm sense of low-end authority which makes for an engaging listen. It's built to the usual ARC standard, which is honestly where a lot of the price tag comes from. But as far as sound, to my ears it seems like a DAC which should cost around $2-3k rather than five digits.

I would put the Denafrips Pontus II (~$1,800) up against the DAC 9 any day and expect the Chinese contender to put up a good fight, even coming out ahead in certain areas. Ditto for the Schiit Ygdrassil ($2,699) or the Matrix X-Sabre III ($2,999). And I think the Yulong DA-1 with Power Station PSU (combined $4k) actually outshines the DAC 9 in most aspects. So it's not that the Audio Research DAC 9 isn't a very enjoyable DAC - it certainly is - but rather a question of value. If you honestly tell me that the DAC 9 is worthy of the $10k price, then logically the Cen.Grand DSDAC1.0 Deluxe should sell for at least double that. But thankfully it only goes for a little over $6k.

Audio Research makes great amplifiers and preamps. It's one of those brands like McIntosh that people tend to stick with, building out the system with visually matching gear and upgrading every once in a while to a newer or higher model. ARC users could do a lot worse than the DAC 9 - it really does have a pleasing sound to it - but they could also get a lot more for their money if they didn't mind deviating from the brand.



Rockna Wavedream Signature (non-balanced version, $13,000)
Rockna's Wavedream was already a popular high-end DAC, having been around in one form or another for some years by now. The Signature version contains various tweaks and upgrades which supposedly brings performance to an entirely new level. I haven't compared them directly so I won't comment on that, but I will say the Signature edition is a superb DAC and certainly in the top-tier of everything I have heard.

I really, really enjoyed this DAC. It has a very fast, extremely detailed signature, but also manages great realism and tonal richness. Soundstage is huge, imaging precise, and low-level detail immense. It also has superb dynamic punch, putting an energetic spin on everything it plays. All in all, this is among the very best DACs I've ever heard.

Compared to the Rockna, the Cen.Grand sounds very similar in many ways, but places more emphasis on musical liquidity rather than speed and punch. Everything flows more smoothly and has a slightly more relaxed feel, whilst the Wavedream is more lively. Both are extremely technically accomplished but the Rockna makes that more obvious or in your face, while the Cen.Grand hits you in a more organic way... yet the tiniest details are still there if you listen for them. I guess I'd call the Rockna presentation more bold and lively while the Cen.Grand is more graceful in focus, if that makes any sense at all. I could certainly be happy with either of them as my reference, from a purely sonic perspective.

When I first auditioned the Signature DAC, it was paired with the matching Rockna Wavedream NET streaming transport. That made for a dynamite combination, particularly via I2S which seemed to offer the best performance. I was later able to demo the DAC at my home with some other transports, and found the result somewhat less pleasing. It was still a great DAC but lost a little of the magic I heard the first time around. The best result I got was using my Euphony Summus music server - an extremely high quality transport in my experience - but it just didn't reach the same level as the matching NET transport. Thankfully Rockna has a mode to work with the PS Audio "standard" for I2S so it plays well with my Euphony as well as with my Cayin iDAP-6, but even so I feel the NET would be a required purchase to unlock the full magic of this device.

Factor in the significantly higher pricing, which climbs even more if you want a balanced version, and I'd say the value proposition is strongly in favor of the DSDAC1.0 Deluxe. But I would not argue with someone who chose the Rockna instead.



Merason DAC1 (original version, $6,000)
Switzerland has given us many high-end audio brands over the years. Weiss Engineering, Piega, Nagra, Thorens, CH Precision, Goldmund, Soulution, the list goes on and on. Merason is not a name I was familiar with prior to this project, but it is certainly on my radar now.

The DAC1 is an interesting device as it doesn't really do anything particularly unique as far as design goes. So no custom FPGA, discrete ladder network, or proprietary upsampling algorithm, and not even the latest/greatest DAC chip(s) on board - Merason uses a pair of TI PCM1794A chips circa 2004. The draw here is not buzzwords or emerging technology but rather good old fashioned engineering using established concepts executed extremely well. That means fully symmetrical dual-mono design, clean power via extremely well regulated PSU, discrete class A output stage, etc. And the result is something I feel challenges any new technology one might find in competing DACs.

The Merason, like the Cen.Grand, manages to combine speed and clarity with smoothness for a very high degree of listenability. I know that's probably not a real world but useful nonetheless. Both devices have weighty, authoritative tonality, both can do silky treble, and both project large performance spaces with excellent localization. Did I mention this was a very capable DAC?

In the end, I feel the Cen.Grand has the edge in a few areas, though the Merason is not without some victories of its own. The DSDAC1.0 Deluxe does even better at layering and soundstage depth, has a slightly sweeter upper midrange, and holds things together more convincingly when things get really busy. In exchange, the Merason does have a little more low-end slam, making the sense of dynamic impact greater. So it's a trade off and I could see someone choosing the Merason depending on their music preferences and system configuration. For me, more often than not I gravitated towards the Cen.Grand, but both presentations were more alike than different and both seemed to have merit.

Two examples really sealed the deal for me choosing the Cen.Grand over the similarly incredible Merason. First, the song "Le Veuve et Le Martyr" by technical death metal band First Fragment, off their superb 2021 release Gloire Eternelle. The track opens with guitars doing a sort of flamenco-inspired intro, and one would be excused for thinking they accidentally selected something by Rodrigo y Gabriela rather than a metal band. It then transitions to a bit of funky slap-bass action of the Infectious Grooves variety. Finally, about one minute in, things get going into the more expected tech-metal direction, with rapid-fire drums and soaring guitar notes. Both DACs handle this track beautifully but the Merason actually feels a little "one note" in terms of overemphasizing the drive and rhythm. The Cen.Grand has ample authority but feels more balanced across the board.

The second example is the upcoming Reference Recordings release of The Complete Beethoven Piano Concertos, with Sir Donald Runnicles conducting the Grand Teton Music Festival Orchestra. Playing the 24/192 FLAC version of this highly-recommended release (also available in 3-disc hybrid SACD format if you prefer), both DACs sound phenomenal, but I hear a greater sense of scale, space, and separation with the Cen.Grand in the system. It's just that much easier to pick out individual instruments, or identify their location on the stage. The Merason is certainly not bad in this regard, but I feel the Cen.Grand has the edge.

It fascinates me how closely these two DACs compete, and how much they have in common sonically. The Merason is a PCM only device which tops out at 24/192 and won't accept DSD at all. Meanwhile the Cen.Grand can only process DSD and converts everything to that format. While I ended up choosing the Cen.Grand, I would not hesitate to recommend Merason either. The brand just updated to the DAC1 to mk2 status and increased the price to around $8k. I haven't heard it so I won't comment about how much it improves and whether it is worth the cost.



Aavik D280 (~$11,000)
Getting to audition this device was a surprise as Aavik really doesn't have a big presence in the USA, despite being part of a major Danish hi-fi group that offers many high-end products under various brand names.

I thought it sounded pleasing in a sort of "audio show" way, where everything is great as long as you play Norah Jones, Stan Getz, or Pink Floyd. Very beautiful presentation, lit-up midrange, layered soundstage, and sweet treble. Also a really large, open presentation, with nicely defined imaging.

Digging into more real-world music - metal, funk, electro, punk, ambient - most of which would never fly at an audio show, and this DAC somewhat fades into mediocrity. It doesn't really do anything wrong per se, but neither does it do anything to elevate itself beyond something like a Matrix X-Sabre 3 which sells for less than 1/3 the price. In fact I would probably take the Matrix, or the Atoll, or the DiDiT, or the Musician Aquarius over the D280 based purely on sound alone, not even factoring in the cost.

Again, this is a beautiful looking and very well built DAC that screams high-end at first impression, but the sonics don't quite back it up. The performance is not bad at all, just not good enough for further consideration, and not even close to the Cen.Grand DSDAC1.0 Deluxe.



Wyred4Sound 10th Anniversary Edition ($4,500)
I've reviewed this device at Darko Audio roughly 6 years ago. Despite the years passing, it remains a very capable performer, and still ticks all my sonic boxes for a reference component. I do prefer the Cen.Grand by a decent enough margin, probably because it has a pretty similar sonic character and just takes things further in most aspects. I've always loved the Anniversary Edition, and despite the many differences in design theme, the Cen.Grand sounds like it could practically be an upgraded version of the Wyred DAC.

This DAC was a limited edition, initially only planned for a small run (hence the "Anniversary" designation). The company did extend it and release more batches over the years, but I don't know if they still ever have new units available at this point. I will see them on the second hand market from time to time, at pricing that seems very attractive for the performance this unit offers. So while I did not use this as my ultimate reference I can very strongly recommend it, no matter how you may be able to acquire it.



Bricasti Design M3 ($5,500)
This brand is one that I've flirted with owning for years. The M3 is relatively affordable compared to the bigger M1, and I actually prefer it in some ways. It trades some realism and detail retrieval for a more visceral signature that I find very satisfying. I also love the fact that Bricasti offers a Roon ready network card option, as well as an integrated headphone output on the M3h version. Seems like a great way to grab all-in-one functionality.

I don't find the M3 to be quite on par with the Cen.Grand though. It actually sounds a lot more like the Wyred4Sound Anniversary DAC that I love so much. That is to say, similar general sonic picture to the DSDAC1.0 but not as capable. If I didn't already have the Wyred DAC, and had not experienced the Cen.Grand, the Bricasti might end up being my choice for a new reference. It's that good.

I wanted to audition the newer M1S2 which sells for $12k. I've heard that model takes the technical excellence of the older M1 and bakes in some of the punchy, guilty-pleasure fun of the M3. But I wasn't able to track one down very easily.


Miscellaneous
Some others I tried got lost in the shuffle, so I don't feel confident giving specific comparisons about their sonic performance. But I do want to mention a few brief impressions of some noteworthy contenders:

*The Aqua La Voce S3 is probably my favorite offering from Aqua. I don't particularly care for their higher models and find them to have poor value for money. The La Voce S3 is not on par with Cen.Grand but remains a great DAC overall.

*I had never heard of Canor Audio, but their DAC 2.10 is a pretty decent option for the price. Lively and dynamic, it seems very flattering of poor recordings, and makes everything sound fun. Not the last word in realism, still enjoyable for what it is.

*MFE Tube DAC SE sounds very similar to the Audio Research DAC9, at less than half the price. Much more compelling in that price range, though still not enough to overcome the Cen.Grand sound. This is another brand I had never heard of but would like to try more.

*The Ideon Ayazi mk2 with their matching 3R Master Time external clocking box is a really musical, engaging device. Not the most detailed but tons of fun and really great for taming bright headphones. Similar in some ways to the Canor DAC 2.10 but with superior details and particularly a wider, deeper, more convincing soundstage. This makes me want to try the higher end Ideon Absolute DAC but the pricing on that puts it in Playback Designs territory so I probably won't bother. Plus I am not sure it can even be found very easily in my region. Still, the "affordable" Ayazi DAC is worth investigating if you like musical sounding DACs which retain good technicalities.




Conclusions
After auditioning several dozen worthy contenders, I can honestly say the Cen.Grand DSDAC1.0 Deluxe is the most impressive DAC I've heard. While not inexpensive, I actually find it to be a great value compared to what others have to offer, in the same price range as well as far beyond. The sound is simultaneously smooth yet transparent, wide open yet intimate, dynamic yet nuanced, to the point that I really don't hear any weaknesses.

Is it the "best" DAC out there? Of course not. There is no "best" DAC, only one that is "best" for a particular user or a specific system. For my musical tastes and associated equipment, the DSDAC1.0 Deluxe is the one, making it a perfect choice for my reference DAC.

(Note that I originally wrote this up as a combined review of the Aquarius DAC and Andromeda amplifier. HeadFi's category system doesn't lend itself well to that sort of thing, so I'm duplicating the content into entries for both products.)

I apologize in advance for all this, but it seems we can't talk about Musician Audio without also discussing its "sibling" brand Denafrips. So let's get it out of the way now: Musician Audio products are probably designed by the same team, and built in the same factory, as the well-known Denafrips equipment. Much (but not all) of the gear from each brand has a corresponding version from the other firm. Rather than rehash the topic, I'll point you to this fair and balanced summary by Srajan Ebaen, which lays out much of the behind-the-scenes info and gets details straight from the source - the head man Mr. Zhao over at Denafrips. He can't quite explain every single specific detail but I think it's pretty clear what the situation is.

Everyone is free to have their own opinions, but to my mind there are only a few things that matter. First, is any intellectual property being stolen? Despite the ambiguity of certain aspects with regards to sales/distribution, I'd say that IP theft is very clearly not happening. Second, will Musician Audio support their gear to the same standards that we've come to expect from Denafrips? That part I can't answer directly, but those of us in North America who purchase from authorized dealer Power Holdings can rely on Arthur's proven customer service should any issues arise. Lastly, is the gear actually any good? That's what I'm here to discuss.

To that end I've arranged to spend some time with the Musican Audio Aquarius DAC ($3199) as well as the Musician Andromeda headphone amplifier ($869) to see how they perform - as individuals, but also as a team. Perhaps it's an unlikely match considering the price spread between the components, but price doesn't always dictate performance, and Musician only has the one headphone amplifier at this point so I figured it was worth a shot.




Musician Audio initially launched their $1100 Pegasus DAC which lots of people seemed to enjoy. Now that they have more brand recognition, they've followed up with a fairly large spread of gear, including the more affordable Draco DAC and the flagship Aquarius DAC. There are also several digital to digital converters, a preamplifier, the Andromeda headphone amp, and even a set of nice looking monitor speakers.

Note that while some of these, based on price and size and capabilities, seem to have an obvious counterpart from the Denafrips stable, others are not so easily classified. The digital to digital converters are very similarly positioned across both brands. In the DAC lineup, I think it's safe to say the Musician Draco competes with the Denafrips Ares. And the Musician Monoceros preamp seems nearly identical to the Denafrips Hestia.

Then again I can't really say where the Musician Pegasus falls. Denafrips doesn't have anything quite like it. Same thing with the headphone amplifiers, where the only model Denafrips makes sells for nearly three times the price of the sole Musician Audio offering. As far as the Aquarius DAC goes, while it is priced roughly equal to the Denafrips Venus, it appears to be quite different from an internal perspective.

Build quality on the Musician gear is very impressive. Again, if you've encountered a Denafrips product, the experience is very similar with two notable exceptions. First, the Musician gear uses faceplates with some interesting angles to them, which is a subjective thing but I happen to really like. Second, the Musician stuff is a bit lighter in weight compared to the hefty Denafrips models. This reflects their internal design philosophies which go in somewhat different directions despite their many similarities. We do still get satisfying weight, surpassing many similarly-sized alternatives from other brands, but it just isn't quite as extreme.




Denafrips is all about massive power supplies - most of their DACs sport a pair of huge transformers and row after row after row of stiffening caps. I recently covered their Pontus II DAC which had over 150 of them in total. In contrast, the Musician Audio Aquarius (and all Musician gear as far as I can tell) tones things down to more sensible levels, more in line with what one might find in any number of other quality DACs on the market. There's still a large O-type transformer flanked by substantial capacitance but it isn't as overbuilt as the Denafrips gear. That likely accounts for the weight difference between them as well.

Aside from that, we see many similarities. That'd be the R-2R + DSD architecture, composed of hand-selected .005% resistors, marshalled by custom FPGA processing, with a bespoke USB solution rather than off-the-shelf chips. The spec sheets read mostly identical but the interior view for each brand shows the different approaches, and that does play out when we listen.







How about that Andromeda headphone amp though? So much attention is given to the digital products from Musician that it's easy to overlook this model. The amp appears sized to match the Draco or Pegasus DACs which probably make more sensible partners than the big Aquarius, but nobody ever accused me of being sensible.

Andromeda is a fully balanced discrete class A design with 3.8w on tap into 32 ohm loads. Connectivity is about as simple as it gets - just a single pair of XLR inputs, and a pair of XLR outputs which allow the device to double as a basic preamp. Up front it's a 4-pin balanced jack plus a 1/4" output which to my mind is not really the ideal choice - remember this is a fully balanced design so single-ended use will be a compromise. The simplistic theme continues when you realize there is no gain adjustment and no need for a button/switch for input selection. That leaves us with just a power button and a volume knob. It doesn't get much more singularly focused than this.

Internally we get a universal-voltage linear power supply which appears to be similar to that in the Pegasus DAC. That means a big O-type transformer with substantial (but not overly so!) filtering. I see some MOSFETs sprinkled throughout but those are likely part of the power supply - Musician uses a big heatsink to cover the output transistors so I don't have further details on that part of the design.




I actually received the Aquarius DAC first, so I got that set up with my Niimbus US4+ amplifier and later swapped in the Violectric V590 (obviously ignoring the internal DAC of that device, at which point I've basically got a V550). Music came via Roon though a Euphony Summus music server over USB or coax or I2S. All power and analog cables were from Audio Art, and the whole thing got balanced power from an Equi=Core 1800 conditioner. I primarily listened using XLR outputs, using the "slow" digital filter, and chose OS mode as the NOS option seemed a bit dull and disorganized to my ears.




Listening to the Aquarius/Niimbus combo with a Meze Elite produced bold tone colors, big dynamic slam, and a sweet, welcoming midrange that sounded excellent with everything I threw at it. From the intense chug of early Mastodon, to the piercing trumpet of Tiger Okoshi, to the expressively atmospheric vocals of Agnes Obel, this setup really delivered. It reminded me very much of the Wyred4Sound 10th Anniversary DAC, a significantly more expensive device that has always been among my favorites when it comes to enticing tonality with a hint of warmth that doesn't go overboard.




I swapped out the Elite for an Audeze LCD-5 and was able to hear even further into the mix, with superb delineation between instruments on the Jimi Hendrix classic All Along the Watchtower. The LCD-5 has exceptional capabilities when it comes to upper midrange and treble articulation, and that "snap" was on full display when listening to Impeach the President by The Honey Drippers. No wonder that track intro has been sampled a million times! Timbre and tone of string instruments was spot on, allowing me to enjoy everything from Crooked Still to Break of Reality to Yo-Yo Ma. This result would be impressive for any DAC but is particularly noteworthy in an R-2R design, where technicalities can sometimes take a back seat in favor of note weight and liquidity.

I then switched out the Niimbus amp in favor of the Pass Labs HPA-1, and heard a subtle shift in tone when listening via Audeze's LCD-5 as well as their LCD-24 Limited. The Pass amp has an excellent "Class A" presentation, with plenty of drive and gusto, but is also a bit less tonally rich through the midrange region - only obvious upon direct comparison to the much more expensive Niimbus amp. It's a subtle difference, and many otherwise excellent DACs I've had in my system were not able to really flesh it out. In other words, many DACs make the Niimbus and Pass Labs amps seem virtually identical, which they are not. So when the Aquarius makes that crystal clear, that earns it high marks for honesty.

One interesting aspect of the Aquarius is that the output levels are a bit lower than average. Musician lists the output voltage as 3.6Vrms via XLR and 1.8Vrms via RCA. The typical numbers are 4V/2V for many DACs, so we're talking about a roughly 11% difference - not drastic or perhaps even immediately obvious, but something to be aware of if the music initially feels a bit "soft". That applies to standard PCM material, but remember that when playing DSD, the Aquarius has its own separate signal path. That gives us 2.6V from XLR and 1.3V via RCA - quite a bit lower than you'll find from many other devices - the PS Audio DirectStream is the only recent DAC I can think of with similarly low output levels. During head-to-head comparisons, one has to take this into account, or else the louder DAC will have an unfair advantage.

Now, part of this has to do with DSD being inherently "quieter" than PCM, for various technical reasons. Many DACs automatically apply 6dB of gain to DSD in an attempt to level things out. Roon also gives the option of applying variable gain when converting DSD to PCM for exactly this reason. So this is not a novel thing. But since the Aquarius already has a somewhat lower than average output voltage, having DSD even lower seems much more noticeable than it does with most other DACs. Again, it's not a "problem" per se, just something to be aware of to make adjustments accordingly. In my case that meant changing the pre-gain settings on the Niimbus US4+. In other cases it might just mean cranking the volume knob beyond the range you typically use.





After more gear swapping, with various amplifiers (both tube and solid-state) as well as headphones from Focal, AKG, Sennheiser, 64 Audio, etc, I came to some conclusions about the Musician Aquarius. First, it's got a very neutral and natural tuning, which doesn't feel as obviously euphonically colored as many of the ladder-based DACs I've used over the years. It's not dark or smoothed in the treble, though it does have just a hint of extra warmth if one listens carefully for it. Compared to the Pontus II which I had in for review a few months back, the Aquarius has by far the more accurate, refined, uncolored presentation. It might not initially be as fun or engaging, but over the long haul it's the more rewarding sound as far as I'm concerned. It's also far more universal in terms of what other gear/music it plays nicely with.

At the same time, the Aquarius does have something special about it, in which it deviates from the standard (potentially boring?) reference sound. Is it the low end authority and subtle bass bloom? Is it the natural, organic midrange? Maybe the sweet treble that balances air and subtlety while remaining smooth and free of glare? I don't think these things are necessarily inherent R-2R traits any more than they are Delta-Sigma or FPGA, as you'll find good and bad examples of all types... but in this case the Aquarius does exhibit some of the stereotypical ladder-DAC behavior. Not necessarily in terms of frequency response but rather the general "feel" of the presentation.

Compared to the Denafrips Pontus II, the Aquarius is more open, with a wider soundstage and more precise imaging and superior layering. It's also more focused and intense, where the Pontus II portrays transients a bit softer. The Denafrips device is warmer and certainly more smooth on the top end, which can be welcome or not depending on the circumstances. Despite how much I enjoyed my time with the Pontus II, I definitely think the Aquarius is a higher tier DAC - as it should be selling for almost double. The real competition would be the Denafrips Venus II which unfortunately I have not experienced.

If forced to choose a popular DAC for reference, I'd say the Aquarius reminds me of the Sonnet Morpheus. Both are very snappy, fast paced but full-bodied, and give a vast holographic feel to the music. Contrast that with the Denafrips house sound which to me seems reminiscent of the Metrum Acoustics presentation - also from designer Cees Ruijtenberg. Unlike his newer Sonnet-branded devices, Metrum's focus is generally more on warmth, smooth/inoffensive treble, and an almost analog sense of liquidity to the affair. It's interesting how the same designer can make two different product lines and give each a unique perspective, as that also seems to be what happened with Denafrips/Musician Audio.

Like the Denafrips models, the Aquarius also sounds best via XLR outputs. It also sounds best, to my ears at least, with OS mode engaged. I don't know if Musician has the same faux-NOS mode as Denafrips - the user manual has a section tauting their "Self-developed and designed digital filters, working in 16 X Over Sampling mode, each level of filter has a 32-bit input resolution, which can greatly increase dynamic contrast, analysis, and more natural and delicate timbre, etc." Yet there is nothing explicitly stating whether NOS mode is truly non-oversampling, or if, like Denafrips, it employs linear interpolation for all PCM material under 768kHz. As I said earlier I find the NOS mode to sound comparatively muffled and dull, so I just stick with OS mode regardless.

I feel the Aquarius does respond quite well to upsampling, either to DSD or very high sample rate PCM. That's the way I prefer to run it (via I2S input, although USB is also excellent) if at all possible. That said, the distinction between that and just running a native signal does not seem as pronounced as it did with the Pontus II. I would happily run the Aquarius with native material straight from a CD transport, where I might not bother with the Pontus II - not that it would sound terrible, but I'd certainly know what I was missing. Aquarius seems more capable of dealing with relatively pedestrian transports and it doesn't suffer so much when using coaxial or optical inputs (though I do still prefer AES if we aren't doing upsampling over I2S/USB). Combined with the more universally appealing and even-handed presentation, this makes the Musician Aquarius an easy DAC to fit into any system and enjoy. Keep in mind this is a DAC and only a DAC - no volume control, no integrated streaming, no headphone amplifier, just a plain D/A converter doing a superlative job at its one and only task.



As for the Andromeda headphone amplifier, it is much easier to evaluate. Build quality is crazy good - it could easily be mistaken as costing several times the actual asking price. It's just a very focused/simple device with a wonderfully transparent sound. It's highly honest to the source, so no sugar coating and no euphonic coloration whatsoever. It's powerful enough to drive most headphones without issue, and has an impeccably clean background. To test this, simply plug in a headphone and crank the volume full blast with no music playing. You'll wonder if you forgot to even turn on the power - that's how silent it is. This translates to impressive dynamic swing and stunning contrast during musical passages.

The lack of gain adjustment was never an issue for me. I used all sorts of headphones from dynamic to planar models, low impedance to high, sensitive and more difficult loads, with no issues. I admit that I don't currently have a Susvara or HE6 here to try, but I suspect those would be some of the few out-of-bounds headphones one might find on the market. Aside from that pretty much everything else should be fair game as far as power goes.

That includes sensitive in-ear monitors, with the caveat that you'll definitely want to use the balanced output. Using the 1/4" jack sometimes gave me a mild "buzz" sound that was just loud enough to be troublesome. This could have been a byproduct of some sort of interference or perhaps a ground loop, but my system is normally dead-quiet so I'm not sure that's the case. Note that this only appeared with some of my more sensitive IEMs - others didn't pick up on this sound at all, and none of my big headphones did either. Still, using the balanced output eliminated this completely, and I still found I had a decent amount of play in the volume knob before things got too loud for comfort. Using the XLR out, the Andromeda paired harmoniously with the 64 Audio A18t, Empire Ears Zeus XRA, AME Customs Radioso, and even my old JH Audio JH13 that I haven't used in ages. The Andromeda's clarity and ability to resolve fine detail really is impressive.




I do wish I still had my old Auralic Taurus mkII since the Andromeda very much reminds me of that lovable device. The Taurus sold for $1900 almost ten years ago and was considered an excellent performer at that price, being chosen by both myself and my then-boss Tyll Hertsens as a reference headphone amplifier. So to hear the Andromeda perform so similarly, at a drastically lower price, is very welcome.

That said, Andromeda isn't always a great pairing in every system. I loved it with the Aquarius as that DAC has just a touch of warmth and a great sense of ease to the presentation. I also very much enjoyed it with the Wyred4Sound 10th Anniversary DAC and the Yulong DA1. And while I didn't get to try it, I suspect the Denafrips Pontus II would also be an appealing partner.




On the other hand, the Benchmark DAC3 B would likely not be an ideal match - a bit too caffeinated, too much energy in the upper mids and treble region. Same with the Matrix Element i and perhaps the various Mytek DACs too - depending on the headphone being driven. The theme here is that I prefer pairing Andromeda with DACs that are at least slightly warmish, since brighter/more articulate DACs (even very good ones) can be too much of a good thing.

Again, the Andromeda is a brutally honest amplifier that pulls no punches in revealing the true nature of a headphone. It's not one to "take the edge off" and certainly not one to add fullness, body, or tonal richness to an otherwise thin headphone. To put it another way - if you love a particular headphone, Andromeda will do a great job bringing out its character. But if you don't like a headphone, or are somewhat on the fence about it, Andromeda will not "fix" it for you. You've been warned.




In the end, I'm rather impressed with what Musician Audio has to offer. The Aquarius is a loveable yet very sophisticated and mature take on the classic R-2R presentation. It's extremely technically capable without veering too far off into analytical territory, and I think a lot of people would be really pleased to have it in their system. It's certainly one of the most capable DACs I've experienced in the ~$3k price range.

Meanwhile the Andromeda is an incredible headphone amplifier for the asking price of well under $1k. It may not be the most full-featured amp out there but it more than makes up for that with its surgical sonic precision. I love the amount of resolution it draws out of the music without being shouty or harsh, and although it isn't necessarily a warm, thick amplifier, it certainly can produce massive bass slam when the recording calls for it. Factor in the extreme build quality - it really does present like a more expensive piece - and the Andromeda starts to seem like a real bargain.

Maybe someday Musician will launch a bigger, more costly headphone amplifier, with more bells and whistles to better match their other top-range gear. But after spending time with the Andromeda, I'm not sure that's necessary at this point. Andromeda performs at such a high level that it really doesn't seem out of place with the $3199 Aquarius - and I can't think of another sub-$1k amp which can make that same claim.

I apologize in advance for all this, but it seems we can't talk about Musician Audio without also discussing its "sibling" brand Denafrips. So let's get it out of the way now: Musician Audio products are probably designed by the same team, and built in the same factory, as the well-known Denafrips equipment. Much (but not all) of the gear from each brand has a corresponding version from the other firm. Rather than rehash the topic, I'll point you to this fair and balanced summary by Srajan Ebaen, which lays out much of the behind-the-scenes info and gets details straight from the source - the head man Mr. Zhao over at Denafrips. He can't quite explain every single specific detail but I think it's pretty clear what the situation is.

Everyone is free to have their own opinions, but to my mind there are only a few things that matter. First, is any intellectual property being stolen? Despite the ambiguity of certain aspects with regards to sales/distribution, I'd say that IP theft is very clearly not happening. Second, will Musician Audio support their gear to the same standards that we've come to expect from Denafrips? That part I can't answer directly, but those of us in North America who purchase from authorized dealer Power Holdings can rely on Arthur's proven customer service should any issues arise. Lastly, is the gear actually any good? That's what I'm here to discuss.

To that end I've arranged to spend some time with the Musican Audio Aquarius DAC ($3199) as well as the Musician Andromeda headphone amplifier ($869) to see how they perform - as individuals, but also as a team. Perhaps it's an unlikely match considering the price spread between the components, but price doesn't always dictate performance, and Musician only has the one headphone amplifier at this point so I figured it was worth a shot. I'll create a separate review entry for the Andromeda, but I wrote this as sort of a combined evaluation so they will both have the same content.



Musician Audio initially launched their $1100 Pegasus DAC which lots of people seemed to enjoy. Now that they have more brand recognition, they've followed up with a fairly large spread of gear, including the more affordable Draco DAC and the flagship Aquarius DAC. There are also several digital to digital converters, a preamplifier, the Andromeda headphone amp, and even a set of nice looking monitor speakers.

Note that while some of these, based on price and size and capabilities, seem to have an obvious counterpart from the Denafrips stable, others are not so easily classified. The digital to digital converters are very similarly positioned across both brands. In the DAC lineup, I think it's safe to say the Musician Draco competes with the Denafrips Ares. And the Musician Monoceros preamp seems nearly identical to the Denafrips Hestia.

Then again I can't really say where the Musician Pegasus falls. Denafrips doesn't have anything quite like it. Same thing with the headphone amplifiers, where the only model Denafrips makes sells for nearly three times the price of the sole Musician Audio offering. As far as the Aquarius DAC goes, while it is priced roughly equal to the Denafrips Venus, it appears to be quite different from an internal perspective.

Build quality on the Musician gear is very impressive. Again, if you've encountered a Denafrips product, the experience is very similar with two notable exceptions. First, the Musician gear uses faceplates with some interesting angles to them, which is a subjective thing but I happen to really like. Second, the Musician stuff is a bit lighter in weight compared to the hefty Denafrips models. This reflects their internal design philosophies which go in somewhat different directions despite their many similarities. We do still get satisfying weight, surpassing many similarly-sized alternatives from other brands, but it just isn't quite as extreme.



Denafrips is all about massive power supplies - most of their DACs sport a pair of huge transformers and row after row after row of stiffening caps. I recently covered their Pontus II DAC which had over 150 of them in total. In contrast, the Musician Audio Aquarius (and all Musician gear as far as I can tell) tones things down to more sensible levels, more in line with what one might find in any number of other quality DACs on the market. There's still a large O-type transformer flanked by substantial capacitance but it isn't as overbuilt as the Denafrips gear. That likely accounts for the weight difference between them as well.

Aside from that, we see many similarities. That'd be the R-2R + DSD architecture, composed of hand-selected .005% resistors, marshalled by custom FPGA processing, with a bespoke USB solution rather than off-the-shelf chips. The spec sheets read mostly identical but the interior view for each brand shows the different approaches, and that does play out when we listen.





How about that Andromeda headphone amp though? So much attention is given to the digital products from Musician that it's easy to overlook this model. The amp appears sized to match the Draco or Pegasus DACs which probably make more sensible partners than the big Aquarius, but nobody ever accused me of being sensible.

Andromeda is a fully balanced discrete class A design with 3.8w on tap into 32 ohm loads. Connectivity is about as simple as it gets - just a single pair of XLR inputs, and a pair of XLR outputs which allow the device to double as a basic preamp. Up front it's a 4-pin balanced jack plus a 1/4" output which to my mind is not really the ideal choice - remember this is a fully balanced design so single-ended use will be a compromise. The simplistic theme continues when you realize there is no gain adjustment and no need for a button/switch for input selection. That leaves us with just a power button and a volume knob. It doesn't get much more singularly focused than this.

Internally we get a universal-voltage linear power supply which appears to be similar to that in the Pegasus DAC. That means a big O-type transformer with substantial (but not overly so!) filtering. I see some MOSFETs sprinkled throughout but those are likely part of the power supply - Musician uses a big heatsink to cover the output transistors so I don't have further details on that part of the design.



I actually received the Aquarius DAC first, so I got that set up with my Niimbus US4+ amplifier and later swapped in the Violectric V590 (obviously ignoring the internal DAC of that device, at which point I've basically got a V550). Music came via Roon though a Euphony Summus music server over USB or coax or I2S. All power and analog cables were from Audio Art, and the whole thing got balanced power from an Equi=Core 1800 conditioner. I primarily listened using XLR outputs, using the "slow" digital filter, and chose OS mode as the NOS option seemed a bit dull and disorganized to my ears.



Listening to the Aquarius/Niimbus combo with a Meze Elite produced bold tone colors, big dynamic slam, and a sweet, welcoming midrange that sounded excellent with everything I threw at it. From the intense chug of early Mastodon, to the piercing trumpet of Tiger Okoshi, to the expressively atmospheric vocals of Agnes Obel, this setup really delivered. It reminded me very much of the Wyred4Sound 10th Anniversary DAC, a significantly more expensive device that has always been among my favorites when it comes to enticing tonality with a hint of warmth that doesn't go overboard.



I swapped out the Elite for an Audeze LCD-5 and was able to hear even further into the mix, with superb delineation between instruments on the Jimi Hendrix classic All Along the Watchtower. The LCD-5 has exceptional capabilities when it comes to upper midrange and treble articulation, and that "snap" was on full display when listening to Impeach the President by The Honey Drippers. No wonder that track intro has been sampled a million times! Timbre and tone of string instruments was spot on, allowing me to enjoy everything from Crooked Still to Break of Reality to Yo-Yo Ma. This result would be impressive for any DAC but is particularly noteworthy in an R-2R design, where technicalities can sometimes take a back seat in favor of note weight and liquidity.

I then switched out the Niimbus amp in favor of the Pass Labs HPA-1, and heard a subtle shift in tone when listening via Audeze's LCD-5 as well as their LCD-24 Limited. The Pass amp has an excellent "Class A" presentation, with plenty of drive and gusto, but is also a bit less tonally rich through the midrange region - only obvious upon direct comparison to the much more expensive Niimbus amp. It's a subtle difference, and many otherwise excellent DACs I've had in my system were not able to really flesh it out. In other words, many DACs make the Niimbus and Pass Labs amps seem virtually identical, which they are not. So when the Aquarius makes that crystal clear, that earns it high marks for honesty.

One interesting aspect of the Aquarius is that the output levels are a bit lower than average. Musician lists the output voltage as 3.6Vrms via XLR and 1.8Vrms via RCA. The typical numbers are 4V/2V for many DACs, so we're talking about a roughly 11% difference - not drastic or perhaps even immediately obvious, but something to be aware of if the music initially feels a bit "soft". That applies to standard PCM material, but remember that when playing DSD, the Aquarius has its own separate signal path. That gives us 2.6V from XLR and 1.3V via RCA - quite a bit lower than you'll find from many other devices - the PS Audio DirectStream is the only recent DAC I can think of with similarly low output levels. During head-to-head comparisons, one has to take this into account, or else the louder DAC will have an unfair advantage.

Now, part of this has to do with DSD being inherently "quieter" than PCM, for various technical reasons. Many DACs automatically apply 6dB of gain to DSD in an attempt to level things out. Roon also gives the option of applying variable gain when converting DSD to PCM for exactly this reason. So this is not a novel thing. But since the Aquarius already has a somewhat lower than average output voltage, having DSD even lower seems much more noticeable than it does with most other DACs. Again, it's not a "problem" per se, just something to be aware of to make adjustments accordingly. In my case that meant changing the pre-gain settings on the Niimbus US4+. In other cases it might just mean cranking the volume knob beyond the range you typically use.




After more gear swapping, with various amplifiers (both tube and solid-state) as well as headphones from Focal, AKG, Sennheiser, 64 Audio, etc, I came to some conclusions about the Musician Aquarius. First, it's got a very neutral and natural tuning, which doesn't feel as obviously euphonically colored as many of the ladder-based DACs I've used over the years. It's not dark or smoothed in the treble, though it does have just a hint of extra warmth if one listens carefully for it. Compared to the Pontus II which I had in for review a few months back, the Aquarius has by far the more accurate, refined, uncolored presentation. It might not initially be as fun or engaging, but over the long haul it's the more rewarding sound as far as I'm concerned. It's also far more universal in terms of what other gear/music it plays nicely with.

At the same time, the Aquarius does have something special about it, in which it deviates from the standard (potentially boring?) reference sound. Is it the low end authority and subtle bass bloom? Is it the natural, organic midrange? Maybe the sweet treble that balances air and subtlety while remaining smooth and free of glare? I don't think these things are necessarily inherent R-2R traits any more than they are Delta-Sigma or FPGA, as you'll find good and bad examples of all types... but in this case the Aquarius does exhibit some of the stereotypical ladder-DAC behavior. Not necessarily in terms of frequency response but rather the general "feel" of the presentation.

Compared to the Denafrips Pontus II, the Aquarius is more open, with a wider soundstage and more precise imaging and superior layering. It's also more focused and intense, where the Pontus II portrays transients a bit softer. The Denafrips device is warmer and certainly more smooth on the top end, which can be welcome or not depending on the circumstances. Despite how much I enjoyed my time with the Pontus II, I definitely think the Aquarius is a higher tier DAC - as it should be selling for almost double. The real competition would be the Denafrips Venus II which unfortunately I have not experienced.

If forced to choose a popular DAC for reference, I'd say the Aquarius reminds me of the Sonnet Morpheus. Both are very snappy, fast paced but full-bodied, and give a vast holographic feel to the music. Contrast that with the Denafrips house sound which to me seems reminiscent of the Metrum Acoustics presentation - also from designer Cees Ruijtenberg. Unlike his newer Sonnet-branded devices, Metrum's focus is generally more on warmth, smooth/inoffensive treble, and an almost analog sense of liquidity to the affair. It's interesting how the same designer can make two different product lines and give each a unique perspective, as that also seems to be what happened with Denafrips/Musician Audio.

Like the Denafrips models, the Aquarius also sounds best via XLR outputs. It also sounds best, to my ears at least, with OS mode engaged. I don't know if Musician has the same faux-NOS mode as Denafrips - the user manual has a section tauting their "Self-developed and designed digital filters, working in 16 X Over Sampling mode, each level of filter has a 32-bit input resolution, which can greatly increase dynamic contrast, analysis, and more natural and delicate timbre, etc." Yet there is nothing explicitly stating whether NOS mode is truly non-oversampling, or if, like Denafrips, it employs linear interpolation for all PCM material under 768kHz. As I said earlier I find the NOS mode to sound comparatively muffled and dull, so I just stick with OS mode regardless.

I feel the Aquarius does respond quite well to upsampling, either to DSD or very high sample rate PCM. That's the way I prefer to run it (via I2S input, although USB is also excellent) if at all possible. That said, the distinction between that and just running a native signal does not seem as pronounced as it did with the Pontus II. I would happily run the Aquarius with native material straight from a CD transport, where I might not bother with the Pontus II - not that it would sound terrible, but I'd certainly know what I was missing. Aquarius seems more capable of dealing with relatively pedestrian transports and it doesn't suffer so much when using coaxial or optical inputs (though I do still prefer AES if we aren't doing upsampling over I2S/USB). Combined with the more universally appealing and even-handed presentation, this makes the Musician Aquarius an easy DAC to fit into any system and enjoy. Keep in mind this is a DAC and only a DAC - no volume control, no integrated streaming, no headphone amplifier, just a plain D/A converter doing a superlative job at its one and only task.


As for the Andromeda headphone amplifier, it is much easier to evaluate. Build quality is crazy good - it could easily be mistaken as costing several times the actual asking price. It's just a very focused/simple device with a wonderfully transparent sound. It's highly honest to the source, so no sugar coating and no euphonic coloration whatsoever. It's powerful enough to drive most headphones without issue, and has an impeccably clean background. To test this, simply plug in a headphone and crank the volume full blast with no music playing. You'll wonder if you forgot to even turn on the power - that's how silent it is. This translates to impressive dynamic swing and stunning contrast during musical passages.

The lack of gain adjustment was never an issue for me. I used all sorts of headphones from dynamic to planar models, low impedance to high, sensitive and more difficult loads, with no issues. I admit that I don't currently have a Susvara or HE6 here to try, but I suspect those would be some of the few out-of-bounds headphones one might find on the market. Aside from that pretty much everything else should be fair game as far as power goes.

That includes sensitive in-ear monitors, with the caveat that you'll definitely want to use the balanced output. Using the 1/4" jack sometimes gave me a mild "buzz" sound that was just loud enough to be troublesome. This could have been a byproduct of some sort of interference or perhaps a ground loop, but my system is normally dead-quiet so I'm not sure that's the case. Note that this only appeared with some of my more sensitive IEMs - others didn't pick up on this sound at all, and none of my big headphones did either. Still, using the balanced output eliminated this completely, and I still found I had a decent amount of play in the volume knob before things got too loud for comfort. Using the XLR out, the Andromeda paired harmoniously with the 64 Audio A18t, Empire Ears Zeus XRA, AME Customs Radioso, and even my old JH Audio JH13 that I haven't used in ages. The Andromeda's clarity and ability to resolve fine detail really is impressive.



I do wish I still had my old Auralic Taurus mkII since the Andromeda very much reminds me of that lovable device. The Taurus sold for $1900 almost ten years ago and was considered an excellent performer at that price, being chosen by both myself and my then-boss Tyll Hertsens as a reference headphone amplifier. So to hear the Andromeda perform so similarly, at a drastically lower price, is very welcome.

That said, Andromeda isn't always a great pairing in every system. I loved it with the Aquarius as that DAC has just a touch of warmth and a great sense of ease to the presentation. I also very much enjoyed it with the Wyred4Sound 10th Anniversary DAC and the Yulong DA1. And while I didn't get to try it, I suspect the Denafrips Pontus II would also be an appealing partner.



On the other hand, the Benchmark DAC3 B would likely not be an ideal match - a bit too caffeinated, too much energy in the upper mids and treble region. Same with the Matrix Element i and perhaps the various Mytek DACs too - depending on the headphone being driven. The theme here is that I prefer pairing Andromeda with DACs that are at least slightly warmish, since brighter/more articulate DACs (even very good ones) can be too much of a good thing.

Again, the Andromeda is a brutally honest amplifier that pulls no punches in revealing the true nature of a headphone. It's not one to "take the edge off" and certainly not one to add fullness, body, or tonal richness to an otherwise thin headphone. To put it another way - if you love a particular headphone, Andromeda will do a great job bringing out its character. But if you don't like a headphone, or are somewhat on the fence about it, Andromeda will not "fix" it for you. You've been warned.



In the end, I'm rather impressed with what Musician Audio has to offer. The Aquarius is a loveable yet very sophisticated and mature take on the classic R-2R presentation. It's extremely technically capable without veering too far off into analytical territory, and I think a lot of people would be really pleased to have it in their system. It's certainly one of the most capable DACs I've experienced in the ~$3k price range.

Meanwhile the Andromeda is an incredible headphone amplifier for the asking price of well under $1k. It may not be the most full-featured amp out there but it more than makes up for that with its surgical sonic precision. I love the amount of resolution it draws out of the music without being shouty or harsh, and although it isn't necessarily a warm, thick amplifier, it certainly can produce massive bass slam when the recording calls for it. Factor in the extreme build quality - it really does present like a more expensive piece - and the Andromeda starts to seem like a real bargain.

Maybe someday Musician will launch a bigger, more costly headphone amplifier, with more bells and whistles to better match their other top-range gear. But after spending time with the Andromeda, I'm not sure that's necessary at this point. Andromeda performs at such a high level that it really doesn't seem out of place with the $3199 Aquarius - and I can't think of another sub-$1k amp which can make that same claim.

Way back in 2016 - an eternity in audiophile years - Matrix Audio teased HeadFiers with news of an upcoming device called the X-Server. It was slated to be a dedicated music server with integrated DAC, Roon capability, robust connectivity, absurdly solid build quality, and stunning dCS-style looks to match. I spent some time with the prototype model and came away impressed by the overall experience.

The X-Server never ended up coming to market. I don't know the exact reasoning behind it, but the firm instead focused on their growing line of dedicated DACs, integrated DAC/amp/pre devices, and other related gear. Despite a decent contingent of HeadFi users showing interest in the X-Server, Matrix probably figured the DAC market was more lucrative, and less of a hassle in terms of software development (which is a fair point).

An interesting thing happened between then and now regarding the number of audiophiles doing their "serious listening" via streaming services. In 2016 Tidal was still somewhat new (as far as being available in the USA at least). Qobuz wasn't available in most regions yet, and Apple Music was still strictly lossy. Spotify and Apple Music were considered acceptable for "music discovery" or "background music" but not for critical listening on quality audio gear.

At that time, myself and most of my friends used file-based playback involving some sort of music server, grabbing lossless files either from a local network drive or directly from an attached hard drive, SD card, etc. Music was ripped from CD or purchased/downloaded from places like HDtracks. In that context, the X-Server made perfect sense, with its multiple USB inputs and internal SSD storage.

But now, many (most?) of us do a significant portion of our listening via streaming services. I personally use Roon which integrates Qobuz and Tidal alongside the roughly 5GB personal collection living on my Asustor NAS. The experience is seamless and ends up sounding great regardless of source, so I really don't have much use for internal SSD storage or USB attached drives.

The reason I bring all this up is to give some context regarding the Matrix X-Sabre 3. It's their latest top-of-the-line dedicated DAC which also does streaming, and as such acts as a worthy replacement for the old unreleased X-Server. Let's explore what this thing has to offer.



With the X-Sabre 3, we get a high-end fully balanced DAC with integrated volume control, XLR and RCA outputs, and a wide variety of inputs. The USB and I2S (over HDMI) options are capable of handling up to native DSD512, plus 768kHz PCM. Streaming options include Roon, Tidal Connect, Spotify Connect, Apple AirPlay 2, and DLNA/UPnP. The unit offers wired or wifi connectivity, but does not work with the existing Matrix app (that's for their Element series). Matrix instead recommends using the dedicated iPhone/Android apps for Spotify or Tidal etc, or else something like mConnect for DLNA playback. There's also a physical remote for folks who prefer that sort of thing. As an alternative, you can use the front-panel capacitive touch buttons, which cleverly disappear after a certain (adjustable) length of time for a cleaner look. The center front panel houses a round display which handles the menu system as well as showing album art, sample rate, and other useful info. It's quite small, but with the X-Sabre 3 being such a low-profile device, there really isn't room for anything larger. X-Sabre 3 sells for $2,999 and is available in silver or black. Did I mention the enclosure being milled out a single block of aluminum? It looks, weighs, and feels exceedingly "premium".

The device can trace its lineage back to the original X-Sabre, reviewed by yours truly way back in 2013. That device sported the original flagship ES9018S Sabre chip - hence the X-Sabre name - when it was still fresh and somewhat rare. More importantly, it also did DSD playback when that was not at all common. Nowadays Sabre chips and DSD are both standard checklist items but at the time it was fairly exciting. And while those features could arguably have been considered gimmicks, the sound of the device was not - it was almost universally enjoyed, and the build quality was shockingly good for the $1100 price tag.

Next came the X-Sabre Pro, followed by an updated Pro MQA version which looked identical. Those brought extensive upgrades including the latest ES9038Pro chip, a new and even more absurdly-well-built enclosure, and price increases in keeping with those improvements (plus inflation). The X-Sabre 3 at first glance has the same appearance but look closer and you'll see the enclosure is actually tweaked for a new look and somewhat lower profile. A new tempered glass top panel certainly looks the part though watch out for fingerprint smudges. Based on pictures (I refuse to ever open an X-Sabre model for internal pics again... see my link above, it was difficult enough on the original version and probably hasn't gotten any easier), the guts appear to be completely redesigned with improved layout and robust isolation between digital and analog sections. But the biggest functional change is the addition of integrated streaming, which I feel is very much worthwhile when implemented well - as Matrix clearly did here.




Listening
My initial listening was done using the stock settings, or at least the settings that were left in place by the previous reviewer when they sent the device my way. I started by listening via Ethernet using Roon, and pairing the DAC with a Niimbus US4+ amplifier driving the Audeze LCD-24 Limited Edition. Everything was fed balanced power from an Equi=Core 1800 power conditioner, and all cabling was from Audio Art.

X-Sabre 3 came across as a clean, neutral, articulate, and generally transparent DAC, doing a great many things right without showing any obvious weaknesses. Bass was tight, treble was detailed and precise, and soundstage was well defined, with tightly focused imaging and superb layering. The sound was fairly unforgiving with poor recordings - I might go so far as to call it brutally honest, for better or for worse. But it also excelled with well recorded material.

Performance was virtually identical when using XLR or RCA outputs. When I ventured away from the integrated streaming, I was pleased to note that all inputs sounded similarly excellent. Even in this day and age, many DACs I encounter have poor USB implementations, but I'm happy to report the X-Sabre 3 is not one of them. I got pretty much the same level of enjoyment from Toslink, AES, USB, WiFi, Ethernet, etc, which made system integration very simple. I was truly thrilled to confirm the quality of the integrated streaming aspect.

I2S was perhaps the lone exception, being ever-so-slightly more clear and dynamic than the rest when fed directly from my Euphony Summus music streamer. This improvement was significant enough that I would choose I2S whenever possible, yet not large enough to lose my recommendation for users who don't have I2S transports available. The good news is that Matrix gives us four different software-selectable configurations for I2S compatibility, which should make it easier to find a matching transport despite that format's frustrating lack of standards. The default configuration seems to be the same as that used by Euphony, PS Audio, Wyred4Sound, Cayin, and various others, with the remaining three options hopefully covering most outliers beyond those brands.

One more time for the glass-half-empty crowd - "unobjectionable" is a totally appropriate word to describe this sound signature, or perhaps "satisfactory" if we use the strict definition without any baggage. That doesn't seem like the highest praise I've ever heaped on a DAC, but I mean it in the best possible way (I promise). There are many, many devices out there which fall short of earning this status. If we stopped there, I would say the X-Sabre 3 is a fine performer which plenty of folks would be happy to own. Seriously - it's a really enjoyable DAC even in "default" form. I would call it sonically competitive in its price range and very likely the absolute best when it comes to build quality.




Optimization
Then it was time to tweak and see how much further I could ramp up the performance. I cycled through the seven selectable filters and settled on either "Slow Linear" or "Hybrid" as my favorite options. As is usually the case with digital filters, these differences were very subtle and difficult to discern in the short term, but made a somewhat appreciable impact when living with the device for a while.

Next I disabled the jitter reduction option, which gave slightly more tonal richness and improved solidity on the low end. This is similar to how my reference Wyred4Sound Anniversary DAC responds to disabling that feature, which is integrated into the ES9038Pro chip itself. As with the filter options, it isn't drastic by any means, yet the cumulative effect can be worthwhile depending on your equipment situation.

I then tried turning "Dither" off, but was unable to notice any difference one way or the other. After much listening, I developed an extremely slight preference for dither being set to on. Honestly this one was by far the smallest impact on sound of all the settings, and I am still not sure if I really heard any difference or just imagined it.

Now for the real fun - switching from "Async" mode to "Sync". This is a very unique option that is vanishingly rare in other devices. While Async mode uses the reclocking process of the Sabre Pro chip which is tied to the nearby pair of Crystek femto clocks, Sync mode disables that altogether. This makes the device far more dependent upon transport quality as the clocking gets synchronized from that end... which translates to potentially better/worse sound depending on the combo.

Using the Matrix X-SPDIF II ($439) USB to SPDIF converter in the chain gave a slightly richer, more fleshed out presentation which I found superior to the already-excellent standard mode. What really blew me away was that it didn't seem to lose any microdetail capability, or soundstage accuracy, or anything else - it simply filled in a bit of perceived tonal weight, with zero negative consequences, turning an already enjoyable DAC into something even better. I'm not exactly sure why this should be - both devices are made by the same company, and both use comparable-quality femto clocks, though not identical (Crystek for the X-Sabre 3, Accusilicon for the X-SPDIF II). I can't exactly explain it but neither can I argue with the sound I heard. Once again I liked I2S best, though all inputs were excellent and to my ears superior to using ASYNC mode.

Switching to different transports, I found SYNC mode to be much more variable than ASYNC in terms of performance. With some transports, such as digital out from an iBasso DX170, the result was just uninspiring. Darker, cloudy, less focused and transparent. It wasn't terrible by any means, but I'd rather just use ASYNC mode and call it a day. Which is weird because the DX170 usually does a respectable job as a transport. Less surprising was the fact that things sounded bland/dull via USB out from a Surface Laptop - again, ASYNC would be a much better choice there. But with the X-SPDIF II, or the ModWright Oppo 205 CD spinner, or direct from my Euphony Summus music server, SYNC mode sounded consistently richer, more robust and alive, without losing any realism. My takeaway is that using a high quality transport with SYNC mode is generally the way to go, with the caveat that some otherwise nice options may not pan out very well.

But wait a second - isn't the integrated streaming portion the key feature of this new X-Sabre model? Indeed it is, and used in that way I still found SYNC mode to be more engaging and full-bodied than using ASYNC. I don't quite know what happens in terms of clocking when using this method, but whatever the situation may be, I can't argue with the result. I hear beautiful textural complexity here just as I do with the other inputs, superior to anything ASYNC mode can muster.

Note that I run Roon from my Euphony Summus powered by a Keces S8 linear PSU, and I like to upsample to at least quad-rate DSD or 384kHz PCM. Perhaps that contributes to the streaming sound quality being so high. In retrospect I should have tried streaming Roon from a general purpose PC and/or without upsampling to see how that performed using SYNC, but the review loaner has already been returned. At the very least I can say that it is possible for streaming to deliver the same high-quality result as using a traditional transport with SYNC mode, even if I can't guarantee the same results with all equipment.




Pairings
As for synergy, I found the X-Sabre to pair extremely well with Violectric and Niimbus amplification. The Niimbus US4+ driven by X-Sabre 3 in SYNC mode over Ethernet gave a beautiful result - the word "pure" comes to mind for how clear and organic it sounded. I liked this combo best with the Meze Elite or the 64 Audio A18T. I later swapped out the Niimbus for a Violectric V550 which gave an even more weighty, thick presentation at the expense of a little speed and microdetail. I preferred this chain with the Kennerton Audio Thekk or the previously mentioned Audeze LCD-24 Limited, as both benefited positively from the very slight euphonic coloration imparted by the V550. Keen eyed readers might notice that I actually have the Violectric V590, which is a V550 plus an integrated DAC section in the same enclosure. When using an external source like the X-Sabre 3, the internal Violectric DAC becomes irrelevant, and the result is thus identical to a V550.







I also used the Cayin HA-6A tube amplifier which paired quite well with the Matrix. The neutral foundation of the DAC meant I was free to tweak the amplifier to whichever settings best fit the headphone/music I wanted to use. KT88 tubes in ultralinear mode? Single-ended triode with EL34? Higher output impedance to warm up my 300-600 ohm headphones? The Matrix really got out of the way and let the versatile HA-6A do its thing.




Comparisons
This was actually a tough one, since I don't currently own another DAC in the same price bracket as the X-Sabre 3. My reference Wyred4Sound Anniversary DAC is significantly more expensive. Both devices use the ES9038Pro but the Wyred has some tricks up its sleeve (custom made discrete voltage regulators, hand-matched Vishay Z-foil resistors, etc) which make it ultimately the better sounding machine. That said, it does give a smoother, more "analog" type presentation which is not always desirable in all circumstances. The Matrix is a bit more insightful at times and when you factor in the integrated streaming (which again sounds phenomenal), the value proposition is huge.

A much more fair comparison would be the Schiit Yggdrasil (latest OG version) I just had here as part of an upcoming DAC roundup for Darko Audio. That model sells for just a few hundred dollars less than the X-Sabre 3, which could easily be offset (and then some) by the cost of a quality transport. Unfortunately I had just sent the Schiit back home when the Matrix arrived here, so I was unable to go head to head. But from (very recent) memory I'd say the two devices were very close in overall performance. In Sync mode, the Matrix was faster and just a bit dynamically softer on a macro scale, though transients were more defined and snappy. It seemed to dig deeper as far as micro detail retrieval as well. Call it more of a reference-style sound if that makes sense, versus Yggy being a little more engaging. The Schiit device had a more open, wide, layered presentation, with Matrix just slightly behind but equal or perhaps even better in terms of specificity. Schiit was a little richer and had better tone density, though again, this was very close - far closer than one would expect when hearing "ESS Sabre versus Multi-Bit design". Seriously, these two sounded way more alike than different, and it came down to subtle cues and system matching that ultimately made me prefer one or the other - which I did in roughly equal measure. To put it another way: I could be very happy with either of them as my only DAC.

Beyond that, I would put the Matrix X-Sabre 3 above any of the recent DACs I've heard from respected firms such as Hegel, Burson, Audio GD, Chord, and Bryston, ranging from roughly $2k to $6k in price. I'm not going into specifics but that list is intended to give a general cross-section of where I rank the X-Sabre 3. I've also recently heard (well, within the past year or two), the Auralic Vega G1 as well as the Mytek Manhattan II, both of which are comparable streaming DACs selling for higher prices than the Matrix. In both cases I very much prefer the X-Sabre 3 - it's not even a close contest. Those are highly-regarded devices with excellent reviews, yet to my ears the X-Sabre 3 handily outclasses them both - and for significantly less money.



Conclusion
The X-Sabre 3 is a big accomplishment for Matrix Audio. While the brand started out making affordable yet surprisingly competent gear, they have steadily improved with each new generation to the point where they now compete with true "high-end" audio firms. The X-Sabre 3 looks, feels, and sounds right at home in a system priced well into the 5-digit territory, using some of the best headphones and amplifiers currently available.

Put another way: if this exact device had a name like "Weiss" or "Playback Designs" or "Esoteric" emblazoned on the front, it could easily sell for double or triple the price, and nobody would bat an eyelash. Owners would be very pleased with their excellent sounding/looking/feeling purchase, and at least some of them would fancy themselves superior to their friends who use Chinese gear. I don't mean this as a slight against people who own those brands (I love all three!) but rather am making a broader point about ego, preconceptions, perceived value, and ultimately the quality of the X-Sabre 3 experience. I would not hesitate to recommend the X-Sabre 3 to anyone looking for a clean sounding, reference-style DAC with integrated streaming.

That said, I can totally understand why some people might be a little disappointed with the device. Not its sound, definitely not its appearance, but rather in terms of features and thus pricing compared to its predecessors. The original X-Sabre model sold for $1100 in the USA. The X-Sabre Pro was around $1700, and the Pro MQA went up to $2000. At $3000, X-Sabre 3 is a pretty drastic increase, even accounting for 10 years worth of inflation.

I've heard the Pro model in my system and I do feel the 3 is audibly superior, but honestly the key improvement involves the streaming capability. So the person who already has their transport situation settled may see this as a large increase in price which mainly provides a feature they don't really need. I can sympathize with that perspective. Selling their existing transport will partially offset the cost but some people might not want to do that.

There's no easy answer to this situation. I've always thought Matrix should offer a dedicated stand-alone streamer, and of course should continue making dedicated DACs. But then again I love having both features combined into this one machine. It's a bit messy as Matrix also has the Element series of streaming DACs, and this X-Sabre 3 somewhat overlaps into that territory. Then again the Element models have integrated headphone amplifiers and the X-Sabre 3 doesn't, so there is a difference.

In the end it comes down to Matrix (and every audio company for that matter) having a somewhat limited capacity for how many products they have on their roster, and the end users deciding whether or not the gear on offer lines up with their needs. I can only speak for myself when I say that apart from my review gig, I would happily use an integrated streamer/DAC like this in my system, and in fact would happily own this very device. Your mileage may vary.

The older I get, the more I seem to enjoy complaining. To be frank, I've gotten pretty good at it, and there is certainly no shortage of topics to choose from. One of those topics has been around for a while by now but I am no less frustrated by it - that would be the omission of 3.5mm headphone jacks on nearly all high-end smartphone designs.

I could tell you all about how the first mainstream high-end device without a headphone jack (the Moto Z, announced in early 2016) actually made sense - it was a one-off, ultra-slim phone which just didn't have the physical space to shoehorn a 3.5mm headphone jack inside. I could discuss Apple's culpability in this whole thing by removing the 3.5mm jack in their iPhone 7 series - which had the same exact dimensions as their 6 series and thus did have enough real estate. I could explain how I was impressed with Samsung for bucking the trend, then disappointed when they reversed course on their S20 models. I could talk about how much I appreciated LG focusing on sound quality - via wired headphones - until the bitter end when they exited the smartphone market.

I could tell you all these things, but I won't. The point is, lovers of traditional wired headphones, wanting a flagship-level phone in 2022, are pretty much out of luck (outside of a few obscure offerings from brands that aren't available in most regions).

That leaves us with a few options.

First is the blossoming wireless headphone market, which thankfully has shown significant improvement since the iPhone 7 days. It's now possible to buy some great sounding Bluetooth headphones from "real" audio-oriented brands such as Sennheiser, Audio Technica, and Sony. I predict continued growth in this segment, with more and more traditional audiophile brands jumping on board in the coming years.

That's great, but what about those who already have sizable investments in headphones and/or in-ear monitors, and just want to make use of the gear they already own while on the go? The answer for those folks involves either using a so-called "dongle" DAC/amp (another blossoming segment) or else going semi-wireless with a Bluetooth DAC/amp. The former is the best choice for optimal sound quality, yet the latter can still perform admirably while offering additional convenience and versatility.

But what if we could have both?

iFi offers just that with their "Go blu" which sells for $199. It's an extremely compact integrated DAC and headphone amplifier which can be used wired via USB C connection or else paired wirelessly using one of several Bluetooth options, including high quality LDAC and AptX HD. iFi packs in much of the same technology found in their larger gear, which means we get familiar sound adjustment options like XBass and XSpace, both of which pretty much do what their names imply. We also get a 32-bit Cirrus Logic CS43131 DAC chip, Qualcomm QCC5100 Bluetooth chip, symmetrical dual-mono signal paths leading to a 4.4mm balanced output, and quality parts used throughout the design. The Go blu is built in typical iFi fashion which means it looks and feels like a very well made product.



Gear
I already own a bunch of decent sounding Bluetooth headphones/IEMs. And I definitely enjoy them under certain circumstances. But when given the choice, I will nearly always go with a higher-end custom IEM from the likes of Noble, JH Audio, Ultimate Ears, 64 Audio, etc. It's a matter of comfort for one thing, but also of sonic performance - no Bluetooth IEM design comes anywhere close to that level of quality.

Which is why I used all of those CIEMs and more while evaluating the GO blu. I also tried an assortment of big headphones ranging from the Campfire Audio Cascade to the Sendy Peacock to Meze's Empyrean and Elite, and found the GO blu surprisingly capable of driving them all comfortably. For sources, I fed the GO blu via USB C from my Surface Pro and then used LDAC to go wireless from my Pixel 6. Music was a mixture of lossless Tidal, Apple Music, Qobuz, and my own library of FLAC files.



Listening
Initial listening was done via Roon using the Surface Pro, with the GO blu driving a set of Campfire Audio Cascade headphones via the standard 3.5mm output. This allowed me to play up to 24-bit/96kHz files and worked flawlessly with no drivers involved. I immediately perceived an improvement from iFi's presentation, which was noticeably more refined and potent than the Microsoft tablet's integrated headphone output. Cascade is not difficult to drive but with the iFi in the chain I heard more clarity, more spaciousness, and - very important with the Cascade - more control. The Surface headphone jack was generally flabby and bumbling when it came to basslines and low frequencies in general, whilst the GO blu gave me the proper sound I'm accustomed to with Campfire Audio's unique headphone design.

I broke out Crooked Still's Shaken By a Low Sound (folk or "progressive bluegrass" as I've heard it called) to see how the GO blu/Cascade combo kept up with the cello and bass, which at times can become complex and fast-paced on this album. While the result was perhaps not quite up to the standard set by iFi's more expensive desktop models, it nonetheless performed admirably, sounding rich and clean without unwanted bloat. The sense of bow-across-cello-string was convincing, as were the plucks of the bass. Switching back to the Surface headphone jack turned those instruments to mush, and also revealed some unwelcome grit in Aoife O'Donovan's beautiful voice which had not been there via the GO blu. Overall, this was a highly enjoyable result.

Next came the Sendy Audio Peacock. This planar magnetic headphone has an interesting sound signature - it reminds me somewhat of the Sony MDR-Z1R in the sense that neither has a textbook frequency response, yet each can still sound very satisfying. To my ears they both give more of a speaker-like experience than most headphones, even if that speaker is something like an Audio Note design which is intriguing if not exactly neutral. I suppose the sound is more emotionally appealing than technically correct - I understand why the Peacock is polarizing but I've grown rather fond of it over the past couple months.

The Peacock comes terminated with a 4.4mm balanced plug, which makes it perfectly suited for the GO blu's balanced output. This duo was warm and inviting when playing Ahmad Jamal's The Awakening, or Sonic Boom by Lee Morgan. Power seemed plentiful, and there was no shortage of drive even when playing quiet classical or jazz pieces. Meanwhile bombastic material ranging from Phantogram's Voices to Meshuggah's 1995 masterpiece Destroy, Erase, Improve did not lose the plot as they did when run straight from the Surface alone. This level of performance is what I typically associate with midrange DAPs from the likes of Cayin, Shanling, or iBasso, so it's impressive to hear it coming from a tiny $199 device. Also note that the delta here was even larger than it was with the Cascade - I could still enjoy that headphone well enough straight from the Surface itself, but Sendy's Peacock sounds downright bad when I remove GO blu from the chain.

Lastly, my headphone experimentation brought me to the Meze Audio Elite, which is perhaps my favorite headphone at the moment. It's a superb do-it-all performer that isn't terribly difficult to drive and is thus - in theory at least - less fussy than some competitors like HiFiMAN's Susvara or the Audeze LCD-3/4/5 models. And the GO blu seemed to adequately power it, with plenty of volume and a significantly improved sense of drive over the Surface alone. Still, the Elite sounded a bit flat and uninspired compared to what I know it to be capable of. Activating XBass helped a little, though I still wasn't completely sold on the combo. It pushed things too far into the analytical realm, and lacked the authority and depth I'm used to hearing. The Elite thus became less balanced, less well rounded than it usually is. Switching to the original Meze Empyrean, with its darker, warmer signature made for a much better pairing - very impressive for the compact size of the GO blu, and no XBass needed. One of the largest headphones I own paired with one of the smallest DAC/amp combo units I've ever tried? An unlikely pairing, but it certainly works.

I suspect the Empyrean could sound better still if driven from a 4.4mm balanced connection, taking full advantage of the GO blu's capabilities just like the Peacock had done. But all the Meze-compatible balanced cables I own are terminated in 4-pin XLR, and I don't have an adapter for 4.4mm operation. Still, it sounds remarkable even via the less powerful 3.5mm output. And while balanced mode may help a bit in driving the Elite, I doubt it would change the general character enough to cause a synergistic pairing. You win some, you lose some.



IEMs
I then turned my attention to IEMs, which for me would be the more common usage with a device like the GO blu. I started with the Surface again, and later moved to my Pixel 6 using the LDAC Bluetooth protocol. In both cases, the GO blu reminded me of its older siblings from the iFi lineup. I've heard quite a few over the years, and generally speaking they all had a similar signature - nice detail with just a hint of warmth and smoothness, excellent staging, and a firm sense of control. The latter is something we would obviously expect from their higher-end models with gobs of power on tap, but I've experienced it on their $400-600 devices as well - and now add the $199 GO blu to the list.

Quality IEMs can often rival or surpass many of the best full-sized headphones when it comes to giving absurd levels of detail retrieval. I tried a wide range of custom IEMs ranging from certified classics like the Noble K10 and JH Audio JH13 to modern flagships like the 64 Audio A18t, along with some lesser-known models from Jomo Audio, Cosmic Ears, Lear, and more. The GO blu handled itself just as it had with the larger headphones mentioned above - excellent dynamics and speed, no shortage of drive, and a well-balanced tonality with just a touch of added warmth. I was unable to find even a single pairing that I did not enthusiastically enjoy.

Moreso with IEMs than headphones, I found the XBass and XSpace functions to be welcome depending on the situation. XBass helped highly-neutral models like my Jomo 6R and Cosmic Ears BA4 to sound just a bit less analytical. It's a pretty subtle change for the most part, and feels more like a compensation for slightly-bass-deficient models than an actual bass boost per se. I'd describe it as having the largest boost (roughly 5dB give or take) in the lowest sub-bass regions, with that boost steadily declining as the frequency increases. By 100Hz or so, XBass is done, so it doesn't touch anything above that range. The result is perhaps more subtle than some people might desire but for my preferences it worked quite well.

Speaking of XSpace - it's really subtle for the most part. A bit of crossfeed and perhaps a slight boost in the upper midrange/treble region which generally helps flesh out the feel of a realistic soundstage. I found it less obvious than XBass, but after careful listening I have to admit it does help out with certain music. Many IEMs have a fairly direct, focused presentation, and XSpace gives them a bit more of an open feel overall. On the flip side, something that already has a tipped up response just might be pushed over the edge. So while I would not use it universally, it's nice to have the option available. iFi does allow both XBass and XSpace to be active at the same time, which for some reason I found less appealing - but maybe I just didn't find the right situation where I needed both at once.

My biggest concern with IEM use was background hiss, which I've heard mentioned from several different folks when discussing the GO blu. I try to avoid reading about a product when it's under review, but I've been told hiss is a fairly common complaint. Keep in mind that hiss has been a problem with portable amps since the dawn of time, so it's not like iFi is unique in this. But I was surprised to find that very few of my CIEMs showed any sort of hiss at all. I've got a wide range of sensitivities represented in my collection, and the only ones I recall having any noticeable noise both happened to be from Empire Ears. That would be their previous flagship Zeus XR and their old Spartan from back when the firm was called Earwerkz. Both of these gave just a bit of background hash which for all intents and purposes disappeared when the music actually played. Using the 3.5mm output gave less than the 4.4mm balanced jack but as far as I'm concerned it was tolerable in both cases.

I'm not saying the hiss issue is overblown. My sampling of roughly a dozen CIEMs is nowhere near large enough to make that claim, and perhaps there are a bunch of universal IEMs out there with higher sensitivities. I also admit that I'm not up to date with the universal IEM market, as my ears tend to dislike them from a comfort perspective. So there could be a large number of popular IEMs out there which in fact do experience hiss when driven by the GO blu. My point here is simply to document my experience that there are also a significant number of well known, excellent (C)IEMs which in fact do not have any issues with hiss at all. As with any amplifier or DAP, your (hiss related) results may vary.




Wired or Wireless?
Notice that in the above section, I didn't really separate the performance in terms of wired versus wireless. Why is that? Because there really isn't a big distinction between them. This of course may vary depending on the music, headphones, and wireless protocol being used. My experience with LDAC, using decent recordings and quality headphones/IEMs, is that it is nearly indistinguishable from using a wired connection. I notice a mild reduction in top-end air and delicacy, giving less realistic staging and ultimately inferior reproduction of spatial cues. But it's not the sort of thing that immediately jumps out at you. I also get the vague impression of reduced dynamics, and perhaps slightly less clarity overall. Imagine looking out through a nice clean window, whereas the wired connection opens that window for an even more transparent view (there is no window screen in this analogy, if that helps the visualization). Again, all of these differences are such that they aren't obvious during casual listening, and certainly don't ruin the experience even when listening closely to excellent recordings with reference-caliber CIEMs.

Switching to aptX HD using the developer options on my Pixel 6, I notice a bit more of a "hazy" feeling, with softer transients and another general reduction in clarity. Overall still quite enjoyable and still not trivial to distinguish compared to a lossless/wired connection. Below that, standard aptX was where things became a bit more obviously "wireless" sounding if that makes any sense. Certainly not terrible but the difference was more pronounced when listening closely. I could (and did!) still enjoy the music even with this relatively-compromised choice. Lastly, dropping to the basic SBC codec is where I drew the line and said "no thanks", due to the very obvious compression artifacts and collapse of realism - as expected using SBC. I did not try out the remaining CODEC options (AAC, aptX LL, HWA) so I can't comment on how they perform. Bottom line: for me, using the higher-tier Bluetooth connections did not hinder my musical enjoyment in any way.



Odds and Ends
One thing which did hinder my experience was the lack of any sort of clip on the device. If I'm wearing a shirt with a front pocket, it's not an issue, but much of the time I don't have any reasonable place to store the thing during active use. If I'm just sitting at my desk listening, it's fine, but for on-the-go use I find it less than ideal.

I had a whole extra paragraph here with further grumbling about the missing clip (I warned you about my love for complaints!) but I now notice that iFi offers a great little leather case, complete with clip, for $29. So that solves my issue nicely. I would like to see it offered in additional colors in the future but I'm just happy to see anything at this point.

In terms of battery life, I'd call the GO blu "satisfactory" in the best sense of the word. I was typically able to get anywhere from 9-12 hours out of the thing, depending on CODEC, volume, and headphone choice. That's impressive for such a tiny device, and easily exceeds the rated time of 8 hours, so I'm quite pleased with it.

For those of you who think you might have zero need for a wireless solution - just be aware that hard wiring isn't always possible. For example, did you know the popular Google Pixel phones have historically had issues with USB audio out? It's not something I had heard anyone mention until I picked up a Pixel 4 XL back when it launched. I plugged in a USB DAC which I knew to work with various other Android phones, and went to install the always-useful USB Audio Player Pro, only to find... it crashed when I made a connection via USB. I tried various other apps such as Neutron and HiBy Music, and while they didn't crash, none could pass a digital signal out to my DAC. Did some searching online and found other people having the same issue. At the time I assumed it was a bug.

Fast forward to mid-2022, and Google is just now fixing the issue. USB Audio Player Pro did not even show up on the play store when searching on my Pixel 6, until last month or so. Sideloading the APK resulted in immediate crashing - the same behavior I experienced a few years back with my 4 XL. In short, until Google fixed this problem (which took them several years), wireless audio was the only option for Pixel users. If it could happen to Google on their flagship lineup, it could happen with future phone releases from other brands too. So it's nice to have a device which side steps that problem whilst losing very little in terms of sound quality.



Conclusion
iFi's little GO blu is the type of device that is easy to overlook. In a sea of more affordable "dongle" format devices from the likes of Questyle, iBasso, and others, it might not grab the same amount of attention. Likewise, iFi's own lineup contains half a dozen quality DAC/amp options in the sub-$1k price range, all of which have great reputations for performance and value. Still, none of those offer the one-two punch of USB plus Bluetooth connectivity, which the GO blu handles without compromise. It really does deliver a large helping of iFi's proven performance, despite its miniscule size and relatively low price.

I confess to not having been very familiar with Sendy Audio until recently. The brand is actually a subsidiary of Sivga Audio, a Chinese firm which focuses on relatively affordable headphones and in-ear monitors. Much like Toyota who launched Lexus in the 1990s to bring upscale designs to market, Sivga created the Sendy imprint for their more opulent offerings. Sendy's most ambitious model to date is the Peacock which sells for $1499 - not a budget product by any means, but also a far cry from Audeze, Focal, HiFiMAN, and others where flagship models go for roughly triple that price (or beyond).

I must also admit that I actually turned down Sendy's initial requests for review, due to having too many other projects in the works between Darko.Audio and HeadFi. But I eventually managed to catch up on those and decided I should give the Peacock a listen. As always, I made no promises about writing a review at all, much less a good one. But after spending time with the Peacock, I found it enjoyable in a unique way, and thus worthy of a write-up.


Design
The Peacock is a full-size headphone design using 88mm planar magnetic drivers. Sendy uses proprietary technology for their diaphragms, driven by their "quad-former" magnet array. The driver housing is CNC machined aluminum mounted in a wooden cup, with custom grills designed to direct airflow in a manner beneficial to the final tuning of the sound. The grills also look interesting and likely form the basis for the "Peacock" name, as they somewhat resemble an ornate presentation of feathers.



Build quality on the Peacock is excellent. I wasn't sure what to expect based on pictures, but in person everything is surprisingly well executed. The earpads are made from what feels like very soft genuine leather, and are stuffed with a thick, supportive memory foam. They have interesting angles which remind me of the pads on my AKG K812 in the way they get slightly thicker towards the lower rear portion that sits behind the ears. The headband assembly is made from a combination of metal and leather, using a double strap system where the lower part can slide to adjust for size while the top portion remains in place for rigidity. The whole thing looks and feels very solid, and does not seem out of place when sitting amongst my collection of significantly more expensive headphones. In fact, HiFiMAN could definitely take some pointers from Sendy in this regard.



Special points are awarded to the included cable, which ranks among the nicest stock cable solutions I've ever seen. The specs indicate 6N pure OCC copper, in an 8-strand configuration, with an interesting two-tone brown/clear color scheme. A wooden Y splitter and matching wooden slider seal the deal in terms of looks, and the whole thing is surprisingly flexible and easy to manage (as far as 8-strand cables go). It looks like something that would easily sell for hundreds of dollars from a quality aftermarket cable supplier - mixed in with my other cables from Effect Audio, Toxic Cables, Norne, Forza, etc, the stock Peacock cable easily fits right in. I typically replace all my headphone cables with something nicer but in this case I'm quite satisfied with the stock offering.



The cable joins the headphone using the same style connectors as Dan Clark Audio (formerly MrSpeakers) headphones. These 4-pin "Hirose" connectors are not all that common, but they work well in my estimation, with solid/secure connections and a nice aesthetic as they fit flush with the headphone cup. On the other end, Sendy's cable is terminated with a balanced 4.4mm plug. It also ships with a pair of adaptors for use with either a full-size 4-pin XLR balanced output or the classic 1/4" single-ended style. Sendy thus covers most of the bases save for single-ended 1/8" style, but that can easily be accomplished by the common 1/4" to 1/8" adapter. This system really is very well executed, and once again is the sort of thing Sendy's high-end competitors could stand to emulate.


Adjustable dual-band system

You can read about all of this and more on their product page so I won't repeat all that here. What I will do is point out how Sendy really seems to be taking the higher path in terms of design and research. Anyone can slap an OEM driver in a fancy earcup and coax a decent sound signature from the resulting headphone, but Sendy appears to be building up their own proprietary technology in all aspects of the design. This is certainly more expensive and difficult, but also makes for a better long-term strategy as they continue to refine and improve their platform.




Comfort
We all know comfort is subjective, but to my head the Sendy Audio Peacock feels slightly more comfortable than any of my Audeze headphones (I have yet to try the lighter LCD-5 or CRBN though). The key distinction is the headband system which distributes the weight to a wider area of my head. In that respect it reminds me of the modern HiFiMAN designs. The clamping force feels just about right to me as well, where Audeze headphones always feel a little loose and therefore less balanced when I wear them. Still, this is a large headphone and at 578 grams, a heavy one as well, so I can imagine it just not working for some users. I also personally really like the slight pad "bump" which helps seal behind my ears, though I have no idea if it works for everyone the way it works for me. The AKG K812 had a vaguely similar design (though a bit more exaggerated) and while it seemed to work for most people, I have heard talk of bad fit for some individuals, partially based on the pad shape. Of course, no design will be ideal for 100% of users, but in this case I do appreciate the shape of the pads.


Tough to capture the slight curves which make the pad so unique

System
I tried the Peacock with all sorts of gear, to get a better feel for how it reacts to different scenarios and voicings. I ended up settling on three different systems that I'd like to talk about - a reasonably affordable integrated DAC/amp solution, a midrange setup with separate DAC and amp, and a high-end rig capable of testing the limits of any headphone. That makes for three very different systems, covering a broad range of pricing and complexity.

The baseline of my system was an Equi=Core 1800 balanced power conditioner. All A/C, digital, and analog cabling come from Audio Art. To represent a fairly simple system, I paired Peacock with the brilliant DA-Art Aquila II all-in-one, fed over USB by a Surface Pro tablet running Roon. As a midrange system, I used the Matrix Element i network DAC which fed a Felix Audio Echo tube amplifier. Lastly, the high-end rig was a Euphony Summus music server, powered by a Keces P8 PSU and running Roon, feeding a Stack Audio Link II bridge paired with the matching Stack Audio Volt PSU. The DAC was a Wyred4Sound 10th Anniversary Edition connected to the mighty Niimbus US4+ amplifier.

(Apologies for any confusion, but I neglected to take good enough pictures the first time around. By the time I realized my mistake and found time to redo them, I had moved various components around in my system. So you'll see all sorts of other gear represented in the pictures here. While I did spend time listening to the Peacock with some of the pictured gear, the bulk of my evaluation was done with the three systems described above. So that's what I'll be focusing on.)


Euphony Summus server with Keces P8 PSU


Equi=Core 1800 balanced power conditioner and Stack Audio Link II network bridge with Volt PSU (also Euphony Server and Resonessence Labs Fluvius server/streamers, not used in this review)


Niimbus US4+ (with Violectric V590 and Wyred4Sound PS-1, not used in this review)


Matrix Element i DAC/streamer with Pass Labs HPA-1

Listening
I started with the relatively affordable DA-Art/Surface Pro system, to help get a baseline of what the Sendy flagship had to offer. Note that the Peacock is not a terribly difficult headphone to drive, with an impedance of 50 ohms and a sensitivity rated at 103dB. So while the DA-Art Aquila II is capable of feeding it roughly 3500mW in balanced mode, the headphone is also quite happy running single-ended which equates to around 1000mW. My initial impressions were... entertaining. Though perhaps not sounding completely "correct" in terms of neutrality and tonal balance, the Peacock nonetheless made for an engaging listen. Vocals were beautifully smooth and clear, with a nice sense of projection and sweetness. Bass was punchy and fast. While not offering the deepest extension I've heard, the midbass presence helped create the strong illusion of a full-bodied sound which never felt thin. Highs were clear, perhaps slightly boosted in the lower treble range (but not overly so) with excellent top-end air. Imaging was clearly a strong point - pinpoint accuracy, along with a very clearly defined soundstage that included height, width, and depth (assuming the recordings contained those spatial cues in the first place - many do not). In vague, extreme-oversimplification terms, I'd call it a W-shaped signature where at least some portion of the lows, mids, and highs are tastefully boosted.

The combined effect reminded me of a really nice set of big, classic monitor speakers in a small, untreated room. This would be the larger type with (required) matching stands which push the boundaries between "bookshelf" and floorstanding speakers. Think Harbeth or Spendor, though perhaps more in general spirit than specific signature. That means superb imaging and soundstage, punchy dynamics if not the last word in ultra-low extension, beguiling midrange that isn't completely sterile or linear but really satisfies with tonal richness, and treble that is not quite "textbook" yet often manages to make things sound a bit more exciting than it otherwise would. In short, there are some colorations here but they are pleasing in nature and tend to add to musical enjoyment rather than distract. Your mileage may vary with this analogy but this has been my experience over the years - feel free to substitute different speakers if your history with Harbeth/Spendor brought you to other conclusions.

Switching to the higher-cost combo of the Matrix Element i streaming DAC paired with Felix Audio Echo tube amplification, things got even more interesting. A bit of tube bloom was obvious, but again made for a pleasing coloration in most instances. Warmth was dialled up a notch, whilst midrange "purity", for lack of a better term, increased significantly. Vocals, ranging from Freddie Mercury to Etta James to Chris Cornell, all had a greater sense of palpable, in-the-room-with-me realism. That was augmented by a huge, open presentation which enveloped me in an ocean of sound. The spatial capabilities of the Peacock combined with the open feeling of the Echo amplifier made for a stunning combo, and certainly some of the most "out-of-head" listening I've heard this side of uber-expensive setups.

Treble clarity, while still being impressive, took something of back seat, with the sound prioritizing delicacy and finesse over pure resolution. I think this has to do with the interesting match of the highly-transparent Matrix feeding the Echo which has a stereotypical OTL tube amp presentation. Note that if I switched over to driving the Peacock directly from the headphone output of the Element i, the result was significantly more sterile and uninvolving. Not just compared to the Echo tube amp but also to the lower-priced DA-Art device. I typically enjoy the headphone output of the Element i but in this case it just isn't a great pairing.

Of course, any tube amp like the Feliks Echo can be tweaked depending on your choice of tubes. In my case I stick with the stock Russian tubes bundled with the Echo because they sound quite nice. Also, it doesn't make sense to spend huge money on tube upgrades for an amp that is positioned as "entry level" within the Feliks Audio lineup. That said, the tonality and character I get from this combo is a beautiful match for the Peacock, and I really can't think of anything I would want to change.

Lastly, I switched to the larger system, weighing in at over $20k worth of gear. This is a highly transparent setup with exceptional resolution that retains plenty of warmth and richness thanks to the impeccably tuned Wyred4Sound Anniversary DAC. Here the Peacock sounded fantastic in some ways yet bothersome in some areas which had not been intrusive on the lower-tier rigs. Resolution was slightly improved, allowing me to hear percussion minutia and other fine details more easily. Dynamic swells were also bigger, bolder, and more lifelike, no doubt owing to the massive power reserves of the Niimbus amplifier. And once again the soundstage was very impressive, offering even more pinpoint accuracy with regard to instrument placement. But I also heard more sizzle around the 7kHz range that brought out sibilance in certain vocals, and made snare drums feel too sharp at times. And I noticed the illusion of quality low-end response unraveling a bit. Here the lack of sub-bass capabilities was more glaring, and the punchy midbass - while still plenty fun - wasn't quite enough to convince.

In the end l actually preferred the Peacock using the mid-tier system. To my ears it accentuated the Peacock's strengths while minimizing its shortcomings. That's definitely related to the Peacock having a somewhat unique presentation which is very enjoyable in its way yet not universally compatible with all systems/music/ears. Factor in the inevitable fit variability of a large, heavy headphone with a pad profile that isn't the typical flat design, and I can absolutely see why opinions vary so much regarding the Peacock. It's not just that some might like them and some might not (which is in itself true of course) but also that different people are hearing different things based on how the headphone interacts with their anatomy and their equipment. This is true to some degree in all headphones but I suspect it to be stronger here than usual. And don't forget to account for the unit-to-unit variability that I assume is present here just as it has been with established competitors like Audeze and HiFiMAN (planar designs seem more susceptible to this phenomenon than traditional dynamic drivers), just to make it that much more difficult to find consensus.


Size comparison to an older Audeze LCD-2


Conclusion
Personally, I enjoy the Peacock quite a bit using two of the three systems I assembled for it. Paired with my relatively affordable all-in-one rig, the Peacock is different and interesting in ways other competitors are not. While some people would prefer the more straight-forward sound of a similarly priced Sennheiser HD800S or the Focal Clear Mg (and I wouldn't fault them for that choice), there is certainly something unique and appealing about the Peacock. This sort of musical, expressive signature is uncommon and I know plenty of people who would absolutely favor it over a more neutral, flat presentation. Meanwhile the mid-tier system I assembled brings out the absolute best in the Peacock. I could listen to that rig all day long. I'd call that a clear success, particularly for a newish company just setting out in the high-end market.

No, the Sendy Audio Peacock is not for everyone, nore will it thrive in every system. Think of it like a Sony MDR-Z1R or an Audio Technica woody. You don't reach for those expecting neutral, linear performance from top to bottom. They have character, and while some might find them off-putting, for others they will be just the thing to bring musical enjoyment. Of course I'd like to see Sendy work on lowering the price even more - lower is always welcome in that aspect, especially for a newer/less established brand - but even more important would be to somehow get a distribution network going where people around the globe could more easily demo these unique headphones. For some, the Peacock will be a very unique and enjoyable listening companion.


3 of these are premium aftermarket cables selling for $400 or more. Does the stock Sendy cable look obviously inferior?

Honda's 1990 NSX, sold under their Acura banner in North America, made a huge splash in the automotive market. It took Honda's well known traits - reliability, practicality, and value - and applied them to an exotic sports car design the likes of which has not quite been replicated even thirty years later. True, it had an asking price roughly four times that of the best-selling Honda Accord, but in return offered legitimate supercar performance typically associated with more expensive (and temperamental) firms. For those who have never had the joy of experiencing an NSX, go dig up some reviews from the 1990s. You'll find near universal praise, full of terms like "flawless" and "breakthrough" and "pure driving pleasure". Despite these accolades, many well-to-do buyers still chose more "prestigious" machines from Ferrari, Porsche, Aston Martin, and the like, paying significantly more money for inferior driving dynamics, poor ergonomics, and many more return trips to the dealership for maintenance and repair. The fact that NSX won nearly every shootout and direct comparison the car magazines could assemble, against these very same competitors, apparently wasn't enough for some buyers to overcome that prestige factor.

This is the situation that comes to mind when considering the new DA1, the most advanced DAC ever released from Yulong Audio. At $4100 for the complete package including a huge matching outboard power supply (which you definitely want), the DA1 is nearly four times the price of Yulong's typical top-line DACs (see DA10, DA9, etc). In return for that relatively large sum, it rewards listeners with superb performance, killer looks and build quality, a sublime interface that makes operation a breeze, and wide compatibility with a variety of formats and transports. It really is the DAC that many high-end customers have been searching for. And yet, as with the NSX, some won't be able to overcome their brand-name focus, and will instead seek out what they consider more "premium" gear. Still, the Yulong DA1 is here for those able to look past such things, and will reward them for their wise choice.

DA1 is what I consider a two part system. The actual DAC itself sells for $2499 and can be used as a stand-alone device, since it has an internal linear power supply. So let's start with that piece and build up to the whole system. I keep calling it a DAC but it's actually an integrated DAC, headphone amplifier, and preamp, as has been the brand tradition for a while now. While Yulong has released a few stand-alone headphone amplifiers, and the occasional DAC-only device, for the most part the firm is concerned with all-in-one functionality, and it seems like they intend to keep it that way in the future.



Design
The DAC section of the DA1 is based around the top-range AKM AK4499EQ, a quad-channel 32-bit design which promises the lowest harmonic distortion of any DAC chip on the market. This is paired with Yulong's FPGA-based 4th generation JIC circuit (Jitter and Interface Control) and a pair of the excellent Accusilicon AS338 precision femto clocks. A fully discrete high-current output stage drives speakers or headphone amps with authority (up to 12V XLR or 6V RCA). Speaking of headphone amps, the integrated amp section is a fully balanced Class A design capable of up to 5W output, with volume adjustments handled by Yulong's digitally-controlled analog solution featuring 99 steps of adjustment. Headphone outputs come in traditional 1/4", balanced 4-pin XLR, and the relatively new 4.4mm balanced formats - a little something for everyone.



Informed readers will be aware of the recent devastating fire which destroyed AKM's facility thus causing a chip shortage and forcing many audio designers to refresh their devices. Yulong happened to be sitting on a fairly large number of AK4499EQ chips prior to that incident, and is therefore able to continue offering the DA1 for the foreseeable future. An insider tells me AKM expects to have at least some of their chips back in production by mid to late 2022, so we'll see how that impacts things.

Back to the DA1 - Yulong has always been very up front and transparent about design and components. See their product page for diagrams and details. My biggest takeaway is that the DA1 takes their already excellent DA10 to its logical conclusion, with a similar basic structure but more refined and fleshed out due to a significant increase in chassis space and budget.

As such, I refer readers to my DA10 evaluation from 2019. It covers the basics which remain in use here - the efficient user interface and control scheme, the selectable filters, the high level of compatibility with smartphones/tablets for transport duty, the various modes of operation (full-scale DAC only, pre-amp mode), DSD512/PCM768KHz compatibility, etc. All of this remains in place, so I won't rehash everything here.

Instead, I'll focus on the areas of improvement, some of which are minor and some pretty drastic.

• The DA1 has a pair of XLR inputs which primarily allows for easy turntable or SACD player integration. This input could also be used for pairing a different DAC with the internal DA1 headamp. I find that to be an unlikely scenario outside of review/comparison purposes, but you never know - perhaps someone wants the option to get a contrasting sound once in a while, such as a syrupy NOS/tube DAC. In that case, the XLR inputs make it easy to integrate.
• Yulong's 4th generation FPGA-based JIC process is even more effective at extracting spectacular sound from lackluster sources. Users are now free to choose their transport based on aspects like UI, size, and price, rather than merely sound quality. USB, AES, Toslink - it really doesn't matter. I mainly used the excellent Stack Audio Link II via USB, just because that's been my transport of choice lately, but I got similarly enjoyable results from various laptops, phones, and tablets. This is a pretty amazing achievement.
• Build quality is even higher than the DA10, which was already very, very well done. The DA1 could easily pass for a very expensive piece of European kit if we didn't know better. I definitely see shades of Lindemann's 800-series models, each of which had 5-figure price tags if I recall correctly.
• While DA10 had several "sound mode" options which were essentially very modest premade EQ curves (hardly noticeable for the most part), DA1 instead offers a choice between clocking modes - Clock 1 essentially forces lower oversampling, whilst Clock 2 conforms to the default AKM recommendation for the AK4499. Switching between them makes a more substantial difference than the old "sound mode" options (though still not drastic), giving users their choice of unique sonic presentations. Both are good, and in some cases I prefer one flavor over the other pretty strongly (depending on system matching), so I really appreciate having both options.
• The addition of the 4.4mm balanced headphone output is also quite welcome. This option is present on the newer/more affordable Aquila II but absent on the DA10, so I'm glad to see it here. The DA1 sounds phenomenal with IEMs so as much as I am invested in multiple high-end IEM cables terminating in 4-pin XLR, the 4.4mm connection just makes more sense.

Complaints? My main gripe is that I would have liked to see a remote control option finally appear. I don't believe Yulong has ever done a remote before, which perhaps makes sense for the more affordable or compact models. But the DA1 has enough inputs, and enough physical space, to where it seems justified in my opinion. Since the volume is a digitally-controlled analog solution, it wouldn't even require a motorized pot, so I can't see a good reason not to add that feature. I also wouldn't mind seeing a pair of analog RCA inputs to augment the existing XLR option. That might make things a bit more complex, but would come in handy for a lot of turntable users. Yulong does suggest we use RCA to XLR adapters if needed, which I suppose is a reasonable solution. Lastly, an I2S (over HDMI) connection might be nice, although with the JIC feature making every input sound amazing, perhaps it isn't necessary after all.



Listening - Stand-Alone Mode
Comparing the DA1 (stand-alone, using the built-in power supply) directly to the DA10 seemed like a good way to establish a baseline, so that's exactly what I did. I have always loved the relatively compact form-factor of the DA10, so the DA1 would have to offer something worthwhile to justify the larger size and price tag. Thankfully the SQ does indeed scale accordingly.

Playing them both back to back, I hear increased resolution from the DA1. It's still got that beautifully smooth, "analog" tonality that we tend to hear from well-designed AKM-based devices, but DA1 feels more incisive. I hear deeper into the mix whether playing "serious" audiophile music (Coltrane, Holst, Alison Krauss) or just music I love regardless of recording quality (Bad Brains, Revocation, Aesop Rock). I hear more air, more openness, and a superior sense of layering over what was already a highly impressive presentation. This is evident whether using my speaker rig or a headphone setup - obviously better associated gear making it more easily discernible.

Using the 4-pin XLR output of both devices, the DA1 exhibits a greater sense of drive. A fairly simple load like my modified/balanced AKG K812 doesn't show a drastic improvement (though I still hear some benefits), but when using a more complex animal like the HD800 or LCD-4, the quality increase becomes more obvious. Richer tone colors, better midrange articulation, and silkier treble are the main benefits. The DA10 still sounds very grain-free until I switch to the larger device and notice the improvement. From there, switching back to the DA10 makes it all the more obvious - funny how that works, as I had zero complaints prior to hearing the superior DA1.

It's not just a matter of power or drive though, as I also hear significant gains when using quality in-ear monitors despite their high sensitivity. My favorite CIEMs - the 64 Audio A18t and AME Radioso - as well as various others from Noble, Jomo, and Ultimate Ears, all move up a notch when played through the DA1. Instrument separation is superior, and complex performances become easier to unravel. The level of resolution and clarity is really striking. I've said it before and I'll say it again: combining a killer all-in-one DAC/amp unit with a top-tier IEM feels almost like cheating. No bulky rack of components and corresponding cable mess, no headphone stands or massive imposing wood/metal cans, just a low key little system which performs at an extremely high level. It's a very satisfying way to listen, and the DA1 - while not quite as compact as some of its predecessors - is a perfect fit for this application.

I didn't spend any time on direct comparisons for the stand-alone DA1. I had already done those with Yulong's DA9 and DA10 models, and found them generally superior to most others in their price class (and often beyond). The same holds true with the DA1. Based on prior experience, I would confidently choose the basic DA1 over a Benchmark DAC 3 HGC, Bel Canto DAC 2.7, Chord Hugo II, and Luxman DA-250, all of which sit somewhere in the $2k-3k range. But that's not really my focus here so I'm going to skip to the most exciting aspect of this device.


Power Station
The Yulong DA1 Power Station ($1599) is a matching external linear power supply created specifically for the DA1 DAC. Its CNC milled chassis matches perfectly in size/design, allowing users to stack it or use it on a separate shelf in an audio rack, depending on their system layout. Yulong thoughtfully includes a very short umbilical cable for use when stacking (DAC on top please) along with a lengthier shielded umbilical for separate use. The Power Station sports a huge pair of toroidal transformers, each isolated in its own chamber for reduced noise, with the whole thing weighing in at around 19 pounds. Add the DAC portion and the combo exceeds 27 pounds, which is pretty beastly when you consider the 12.5 inch width rather than the typical "full-size" component width of around 17 inches. If we go by the old-school method of "heavier is better", the DA1 with Power Station is clearly a winner.





Adding the Power Station brings a whole new level of refinement to the system. I hear better drive and authority, an even larger more holographic soundstage, superior bass depth, improved midrange texture, and a certain effortless feel which is truly intoxicating. At this point I'm comfortable using the DA1 with any headphone, including a Susvara, HE6, etc, without feeling like I need more power from an external headphone amp. And yet the ridiculously competent performance with high-end IEMs remains in place - in fact it gets an even blacker background which means superior contrast and dynamics.



Yulong's marketing material calls the Power Station an "optional but significant upgrade" and I absolutely agree with that statement. I'd actually go a step beyond that and say most DA1 owners will want to add the Power Station when funds allow. It turns a class-leading $2500 device into a $4100 unit that can take on many challengers costing far more.



I know it's common for reviewers to throw out that sort of statement without any real point of reference. Is it hyperbole, or are they actually familiar with twice-as-expensive devices which can't compete? If so, why not name them? They usually don't, but I will - here are a few specific examples of the DA1 with Power Station outperforming some highly-regarded (and expensive) gear from other brands.


Comparisons

SimAudio 430HA ($4500 as tested)
Regular readers may recall that I did not get along very well with this component when I reviewed it several years ago. I quite like SimAudio as a brand, and was excited to see they had branched out into the headphone world, but the resulting component just didn't do much for me. Which is odd, as I know several users who are quite satisfied with it. In any case, one publication I wrote for at the time actually rejected my article, saying I either needed to spin it in a more positive way or they would just refrain from publishing. That marked the beginning of the end of my contributions to that particular site.

In any case, I recently had a chance to revisit the 430HA and compare it directly to Yulong's flagship device. The comparison is valid since this particular 430HA came with the optional DAC card, making the price just a few hundred dollars more than the DA1 set.

Long story short, my feelings for the SimAudio haven't changed, but I have a slightly better understanding of the target audience they may have been shooting for. To my ears, and particularly when contrasted against the DA1, the 430HA sounds soft, distant, and borderline reticent when it comes to transient attacks. It's certainly a smooth presentation that doesn't offend, but also fails to excite me in the least. The DA1 duo is more resolving, incisive, faster, far more open and accurate in terms of imaging, and has a beautiful midrange sweetness which engages in ways the SimAudio machine can't match. I even tried levelling the playing field by using a lopsided transport configuration - high-end Euphony Summus music server for the SimAudio versus an older LG G7 smartphone for the Yulong - and still could not find a single area where the 430HA took the win.

Once again I feel the DAC add-on in the 430HA is surprisingly good, and the pre-amp capabilities are top notch. But the headphone amp itself seems tailored for an audience of speaker loving, non-headphone-users, and is voiced to deliberately take away any of the strengths of headphone listening in favor of something that (somewhat) approximates the speaker experience. Meanwhile the DA1 is, to my ears, very clearly the superior device.


Mytek Manhattan II ($7650 as tested)
The Manhattan II used for direct comparison was augmented with the optional network card, causing the price to approach double that of the Yulong DA1 with Power Station. Obviously the network capability is an unfair comparison, as Yulong's device lacks such a feature. Then again, with their JIC system working so well, the Yulong can be paired with practically any transport (even very inexpensive devices) and still give outstanding performance.

After spending much time going back to back, I'd say Mytek's headamp stage is by far overshadowed by the integrated Yulong amplifier. Using the optional balanced headphone adapter brings out the best Mytek can muster, but it's still no match for the DA1. In isolation it sounds reasonably capable though somewhat flat and bright, pairing best with warm/dark headphones such as my vintage Audeze LCD-2 or the Ultimate Ears Reference Remastered. Switching to the Yulong opens up a whole new world of tonal richness, authority, and separation. Moving back to Mytek again feels threadbare, almost hollow in comparison, and to my ears does not satisfy considering the cost of the device.

While Mytek's headphone output is the weakest link, the DAC section shares some of the same issues, though to a lesser extent. There are times when the presentation works pretty well - particularly the spotlit top-end, airy and ethereal, throwing rapid-fire details without losing the plot. But again the switch to Yulong reveals more midrange body and sweetness, and significantly more low-end texture without becoming overbearing. Alto and tenor sax, violin and viola, the differences in timbre present themselves more readily.

If used as a stand-alone DAC only, I could probably assemble a system where the Manhattan II is the better fit - but only under very specific circumstances. I'd mostly be using the somewhat lean tonality to offset colorations from other gear in the chain. Meanwhile, for the vast majority of hypothetical systems out there, Yulong's DA1 with Power Station is the superior choice, and for significantly less money.

Now that I write this, I realize that switching the "Clock 2" mode on the DA1 may actually work better in those limited situations where Mytek had excelled. I forgot to do so during my direct comparisons, but did make that change at other points in my evaluation, and am pretty confident in describing the result. Clock 1 works with less oversampling, and Clock 2 uses the default AKM operating mode. The result is a tonal shift where the focus is more on detail and speed, for a lighter, more airy presentation. As much as I love the analog feel of Clock 1, Clock 2 does have its uses from time to time. Also note that Yulong gives us a choice of 3 digital filters: Sharp, Slow, and Short, all of which give extremely subtle contributions to the overall sonic flavor. These make about the same difference as they did in the DA9 and DA10 (which is to say pretty minimal) yet are still worth messing with in combination with the Clock selection.



Chord Hugo TT2 ($5495)
I've been struggling to make sense of the "Chord sound" for ages. I had their DAC64 back in the early 2000's and found it very interesting, somewhat unique, and utterly confusing. That device had beautiful detail, somewhat laid back in a nicely balanced way, accompanied by a certain overcooked warmth that could be fun or obnoxious depending on the situation. I really enjoyed it... except for when I didn't. This confusion was not alleviated with the launch of Chord's original Hugo, which I picked up at launch and found similarly tricky. On the one hand, the resolution felt at times off the charts, bringing to mind megabuck devices from Esoteric and dCS. The overdone low-end was no longer an issue, or at least not in the same way - now it felt a bit lightweight and fleeting, which is arguably a better problem to have. My key gripe, leaving aside the ergonomic and reliability issues on those early units, was the impression of hyper-focused "in-your-face" tonality, lacking weight and thickness, sounding at times like a cardboard cutout. I eventually concluded Hugo was an interesting yet flawed product that would probably lead to better things down the road.

One such better thing is the Chord Hugo TT2, the most recent evolution of the desktop-oriented Hugo TT. It's not perfect by any means, yet manages to fix or minimize many of my complaints about the Hugo and Hugo2 devices. I can certainly see (and hear) why some people swear by it. Still, at over $5k and with a headphone output that I find generally underwhelming, I don't often find myself recommending it over the many alternatives available in that price range.

Upon direct comparison with the Yulong DA1 combo, TT2 sounds more lit-up, speedy, and crisp. The low-end response is actually quite nice, hitting a sweet spot between the DAC64 and original Hugo's colorations. The TT2 sound is more energetic, and perhaps more neutral or "factual" than the Yulong, which sounds comparatively editorialized - but in a very pleasing way. DA1 has a bit of midrange sweetness that really brings out the character of vocals. It imparts a more defined sense of body to instruments, letting us hear not just the strings but also the wooden body of a cello or viola. Brass instruments sound larger, drums have more heft and a sense of reverberation going beyond just the heads alone. Overall there's a sense of dynamic punch present with the Yulong DAC that Chord's device can't match, making each unit appeal to the preferences of different users.

Note that the above mainly applies when using each device as a DAC with external amplification - I used my Niimbus Audio US4+ to flesh out these differences. When used as all-in-one DAC/headphone amp devices, the Yulong DA1 pulls ahead by a considerable margin. Here I find the TT2 severely lacking, much like the Mytek - both make clean, detailed sound suitable for "monitoring" in the most sterile way possible. But neither is the least bit engaging, nor does either remind me of what actual live music really sounds like. While I could see certain people choosing TT2 over DA1 for DAC duty, I can't fathom how anyone would prefer Chord's headphone output.

Here again, I neglected to try Yulong's Clock 2 mode until after the comparison unit had departed. And once again I feel it might win back some users who perhaps gravitate towards the Chord type of sound. When taking that into consideration, along with the vastly superior headphone section, significantly lower pricing, drastically lower transport requirements, and more straight forward ergonomics, I think the Yulong DA1 with Power Station is the clear winner.




Meitner MA3 ($9500)
I've always been a fan of EMM Labs gear, though I find the pricing somewhat absurd. But as far as "traditional Hi-Fi" audio firms go, EMM Labs is legit. Their spinoff Meitner Audio brings most of the sound quality in somewhat less expensive form - though still not affordable by any means.

The Meitner MA3 just launched earlier this year and is a fantastic sounding DAC with integrated Roon functionality. For just under $10k, we get a pure DAC with excellent volume control but no analog inputs. Meitner encourages direct-to-amplifier connection for speaker rigs, but those with analog sources are out of luck. Again, this is not what I'd consider a high value product, but the build quality, SQ, and pedigree do their best to help justify the cost.

I doubt many have heard the MA3 due to it being so new, so I'll just sum it up by calling it a generally neutral, well-balanced, unobjectionable DAC, with superb resolution that doesn't offend. That makes it seem pretty bland but trust me, the MA3 is anything but. Despite not having an easily describable sonic flavor, the MA3 performs up there with some of the best DACs I've heard. In fact it reminds me very much of my reference, the Resonessence Labs Mirus Pro Signature, stopping just short of matching that device in overall performance.

Yulong's DA1 with Power Station gives the MA3 a tough battle. I hear the Meitner as having a slight advantage in top-end extension, whilst the Yulong wins down low in the sub-bass region. Both have extremely well articulated soundstaging, with the DA1 being a touch larger and the MA3 feeling more precise in terms of imaging. Dynamics? Speed? Transient impact and decay? Both machines are neck in neck with no clear winner making itself known.

The biggest difference seems to be the Yulong again having just a bit more warmth and body in the midrange. There's a sweetness there which I typically find welcome, though it isn't all that obvious unless I listen back to back. When not streaming via the integrated network card, the MA3 does require a suitably high-end transport or else it falls behind, whilst the DA1 keeps on singing without concern for input format or quality. MA3 also has no headphone capabilities and no analog inputs, which to me tips the scales in Yulong's favor in terms of value and usability. Both devices are excellent and I could happily live with either, but the key takeaway here is that Yulong's DA1 remains very competitive even in highly-regarded company like this.



Violectric V590 ($4100)
The V590 is my reigning champion when it comes to one-box headphone amp/DAC/preamp devices. It sells for the same price as a DA1 with Power Station, so it seems like a logical comparison - and I happen to know that Yulong greatly respects Violectric and uses their gear in his own reference rig for inspiration.

Unfortunately, the full comparison doesn't really work. My V590 is the original model built around dual-mono AKM DAC chips, which are now unobtainable due to that previously mentioned factory fire. Violectric used up all existing stock, so like many other firms was forced to redesign their entire digital section. The new version sports an ESS chip and I have no idea how it performs - likely quite well, since Violectric knows what they are doing, but any DAC comparison done with my V590 is already irrelevant.

As far as the headphone amps go, the V590 is a touch bolder and more dynamic, with a slight focus on the lower regions though not necessarily what I'd call an actual boost per se. Yulong's device meanwhile gives preference to those midrange tones, drawing added attention to vocal shading while again not sounding unbalanced in the least. In both cases the shift in focus is very minor, an area of expertise that each respective amp is just more natural at accomplishing even if neither ultimately deviates far from neutrality in general terms. In many situations I would have trouble choosing one over the other.

The V590 does take a fairly clear lead when dealing with the worst headphone loads out there - Susvara and HE6 and the like. While DA1 remains very capable in these scenarios, it does take a back seat to the V590 which feels more effortless and full-bodied. Meanwhile the DA1 gets a victory with IEMs, giving an inkier background and thus more dynamic contrast. But again, neither device is lacking outside of a direct comparison.

I'm actually surprised at how close the DA1 comes to the V590 in terms of amplification. The Power Station really helps bridge the gap, with the base DA1 configuration falling significantly behind. In the end, each model performs in line with its design criteria - Yulong's device is primarily billed as a DAC which happens to include a killer headphone amp and pre-amp capabilities. Violectric starts with the headphone amp/pre-amp (based on an existing stand-alone product) and adds a DAC to round out the feature set. Again, not going into specifics on the DAC comparison other than to say I'm pretty confident DA1 takes the win in some key areas.



Conclusion
Like the original Acura NSX, Yulong's DA1 is superbly balanced, undeniably attractive, and - while certainly not cheap - ends up being a great value in relative terms. And yet, reading about either of them is not sufficient. They really need to be experienced to be understood. Unfortunately, as with the NSX, not enough people will get that chance.

If the opportunity ever presents itself, I highly recommend taking the Yulong DA1 for a spin, particularly with the optional Power Station. You'll be glad you did.

At this point, I don't feel the need to make long-winded introductions about the Violectric brand. I've written extensively about their gear over the past decade, along with their more studio oriented Lake People gear and now the ultra-high-end Niimbus imprint. Most readers have a good handle on the history and pedigree involved, so I'll cut right to the chase on this one.

The V590 is Violectric's latest flagship. It's a true all-in-one DAC, headphone amp, and preamplifier, with equal attention given to all three aspects. This deviates from prior Violectric models where the headphone amp was the main attraction and the optional (and relatively low cost) DAC add-on was nowhere near the same caliber. Conversely, other mainstream devices (Benchmark, Mytek, Antelope Audio, etc), to my ears at least, give the DAC top billing whilst the headphone out is more of a bonus feature for the sake of convenience. But the new Violectric V590 is a true all-around performer.

To set the scene: Violectric designed the amplifier portion of V590 as a successor to the legendary V281. This means at some point down the line a stand-alone version will no doubt become available, but I don't have any further details regarding their time frame. In any case, this is a balanced, discrete design with all the typical Violectric talking points: massive power reserves, high internal operating voltage, low output impedance, and low noise. This amp is potent enough to drive any headphones with ease (5 full Watts at 50 ohms) while being quiet enough to synergize with sensitive IEMs (V590 has 8dB less noise than V281).



The DAC section was built from the ground up specifically for the V590. I don't know what the company has planned for a future stand-alone version, but I wouldn't be surprised to see that happen - the performance is certainly high enough to justify it. The design is based around dual-mono AKM DAC chips along with AKM's exceptional 32-bit resampler which also acts as the digital audio receiver for incoming signals. The Amanero Combo384 handles USB signals which then route to the resampler as well. Front panel controls give us several options for resampling including the ability to disable it completely. A Femtoclock ties the whole operation together with utmost precision.

Note that while designer Fried Reim has been notably critical of DSD audio in the past, and had thus far stubbornly refused to implement DSD capabilities in his prior devices, he nonetheless relented and made V590 capable of playing up to quad rate DSD (the specs indicate it goes higher to DSD512 but I have not tried). This is a long overdue update in my humble opinion. While it's true that the vast majority of music is only available in standard CD quality PCM, we still want the ability to play hi-res PCM as needed - I can't imagine Fried releasing a DAC limited to 16-bit/44.1KHz. Similarly, even though many of the "DSD" releases are just glorified/upconverted PCM, there does remain a small amount of superb "real" DSD recordings to be enjoyed. Fried may still be critical of the merits of the DSD format but it's always a good idea to let the users decide for themselves. And at this point, most DAC chips and USB implementations have that functionality on board anyway, so why not?

The standard V590 configuration ($3499) involves a motorized Alps pot, allowing the very nice bundled remote to adjust volume, select inputs and outputs, and most everything else which has a front panel control. You do have to set the pre-gain and line-out gain manually around back using the traditional Violectric DIP-switches, but that's more of a set-and-forget situation. An additional $599 gets us the "Pro" configuration which replaces the Alps pot with a 256-step reed-relay solution similar to the one found in the mighty Niimbus US4+. My review unit came with the Pro configuration so I can't really speak to how it contributes to the sonics compared to a stock build.

As for pre-amp capabilities, V590 has a pair each of RCA and XLR connections on both the input and output sections. Add in the digital inputs (USB, AES, coax, and optical) and most folks likely have enough connectivity for their entire rig. At one point I paired V590 with a music server via USB, a CD transport via AES, a warm gooey tube DAC (to give an alternate take on the presentation) via analog XLR in, and still had a set of RCA inputs remaining for connecting a turntable if I wanted to go that route. I then used the RCA outputs to feed a warm gooey tube headphone amp (again, for variety) whilst tapping the XLR outputs for my speaker amp. It's certainly possible that someone would need even more connectivity, but for the majority of users V590 is likely to be enough.

In terms of sound, my initial concern was how much of an upgrade the V590 amp section might be over the already excellent Violectric V281 amplifier. Way back when the V590 first arrived, I posted my initial impressions in the main V590 thread, and honestly my opinion has only solidified since then. After spending many hours listening and comparing back to back, I would say V590 amp section offers something of a mixture between the V281 sound and the Niimbus US4+ signature.0



While evaluating the amp section, I used my reference DAC - the Resonessence Mirus Pro Signature - feeding V590 via Audio Art XLR cables. This was fed by a Euphony Audio Summus music server by way of a BMC PureUSB1 active cable. Audio Art supplied all interconnects and AC cables, and headphones included a wide range of models such as LCD-4, Susvara, Utopia, HD650, Elex, K812, Kennerton Thekk, Empyrean, and various others which I'm probably forgetting.

The general Violectric "house sound" (for all their gear but particularly the higher-end headphone amplifiers) involves rich tonality, huge dynamics, superb low frequency slam which is also very controlled, a suitably wide and deep soundstage which is set back somewhat from the front row, and a wonderfully balanced treble response that walks the line between smoothness and detail retrieval. I'd say the V590 continues that same tradition but moves slightly closer to the Niimbus sound: more focus on top end refinement and clarity, slightly less bass slam yet more texture and nuance down there, superior imaging, and faster transient response. It could still be considered just a touch on the warm side, particularly so if compared to something like the headphone out of a Chord or Benchmark product, but I'd call it generally more neutral then the V281 overall.



The V590 amp stage has enough grunt to make even the mighty Susvara sound exceptional. Is it on par with the Niimbus US4+ or a good speaker amplifier? Not quite. But it is, without a doubt, the best I've heard Susvara sound from an all-in-one device. Ditto various other difficult loads like HE-6. Dynamics are astounding and bass punch is palpable, while treble remains totally controlled - three areas which usually expose poor matches on these challenging headphones. Meanwhile sensitive headphones have no trouble in terms of noise - V590 is practically silent, and the pre-gain settings work wonders for optimal pairing with a variety of sources. As I type, I'm listening to the AME Radioso custom IEM via my classic Toxic Cables balanced silver/gold cable, and there's not a hint of background noise or hiss to be found. I do use a balanced power conditioner and I know that contributes to a lower-noise experience, but I've got many other devices here which aren't as quiet as the V590. So obviously Violectric is doing something right. Perhaps some uber-sensitive IEMs may bring out more hiss but in my (fairly extensive) collection I'm just not having any issues.

I also need to mention how much I love the reed relay volume control, a $599 upgrade over the standard version. It is very similar to the design used in the Niimbus US4+, which I've always maintained is the best volume control I've yet to encounter on an audio product. It just feels "right" to me. Having 256 volume increments is vastly superior to most stepped attenuators out there using 24 steps (Burson, Cayin), 36 steps (Bottlehead), or 48 steps (Wells Audio) - with so many steps in such small increments, I'm able to dial in just the right levels regardless of my headphone choice. Operation is practically silent, unlike the V281 stepped attenuator option which produced audible clicks that bothered some people (I never minded it). I can't say for sure how much the sound changes when going from the relay option to the standard Alps volume solution, as I have not heard that version. With V281 it was a small but noticeable difference, and I imagine the superior reed relay design plus the general increase in resolving capability on this device would lead to a more substantial improvement, but I am obviously speculating here.



Swapping in various sources to evaluate how the V590 amp responds to changes proved illuminating. This amp section is more responsive than its predecessor, with the ability to showcase the best and worst aspects of upstream gear. The punchy, dynamic Wyred4Sound 10th Anniversary DAC. The lush analog sound of the ModWright UDP-205. The intense yet slightly overcooked tonality of my aging SimAudio Moon Orbiter universal player. The effortless realism of the Sonnet Morpheus. V590's incisive amp section exposes the true character of each DAC or player with minimal editorializing of its own, save perhaps for a hint of (welcome) warmth. Repeating this experiment with V281 just doesn't capture the same amount of variation from one source to the next.

Overall the amplifier section feels like a very sensible evolution of the existing Violectric presentation, without going so far as to abandon the core appeal. Anyone reading the above and thinking the V590 might be too thin or analytical has nothing to worry about. This is still a meaty, rich, textured amplifier that will bring out all the authority a headphone can muster. It won't turn an HD800 into an Empyrean, but rather will give the Sennheiser a fighting chance for those on the fence about its spartan presentation. I am particularly enamored with how fast, open, and holographic it sounds - attributes I encounter more often with brighter, more caffeinated sound signatures. Experiencing them combined with such a rich tonality is very satisfying.



Moving on to the DAC portion. I fed it through the same Euphony Summus music server, and isolated it by using XLR outputs to the Niimbus US4+ headphone amplifier. I later swapped in my Nativ Vita music server in order to investigate differences between digital inputs, since the Vita offers a wide array of connectivity options (spoiler: V590 doesn't have much variation at all, meaning USB is very well sorted). Again I listened with a variety of headphones and music to give the DAC a real workout.

The DAC section lays a great foundation for the integrated amplifier but also pairs quite well with any external amp one might use. It's clear, highly resolving, and has a bit of sweetness to it which keeps it from becoming too analytical. This midrange sweetness - actually the entire presentation if I really think about it - reminds me of the Yulong DA10, which is one of my absolute favorite DACs in the sub-$2k category. Both are based around AKM chips but of course there are many other factors involved, and I have plenty of other devices with AKM chips that sound completely different.

Bass is weighty but not boosted. Texture and clarity seem to be the main focus, and are sufficient to give the impression of excellent authority. Using the Niimbus US4+ is a great way to highlight any shortcomings one might hear in the low-end of a DAC, and it has sometimes exposed weaknesses that I had not heard when using lesser amplification. The V590 has no such problems.

Treble is superbly balanced - accurate, very detailed, but also very slightly laid back in comparison to some of my other DACs. It's not what I'd call "soft", nor does the DAC veer into "dark" territory in the least. The focus is just more on textural flow than transient attack and extended decay trails. This DAC seems meant for listening enjoyment rather than gawking at obscure details. So while it does dig plenty deep into the mix, that doesn't quite seem like the main focus. Still, if we look back to the older Violectric V800 and V850 products - both known for their neutrality and resolution - this new DAC digs deeper while at the same time flowing more smoothly. So ultimately this is still a generally neutral component.



Playing with the resampling options gives us subtle variations to the resulting sound. The differences are small, and at times almost impossible to notice, but in the right situation I do hear changes. This is noteworthy since V590 does not allow users to cycle through DAC filters as some products do - but these resampling options seem to make a larger difference anyway. My general impression is that higher sample rates (particularly x4) bring a sensation of more treble energy and top-end air. This is occasionally welcome, such as when dealing with dull recordings or darker headphones, so it's nice to have the option. I generally leave resampling set to "best" and call it a day though. Choosing "off" disables resampling altogether, which to my ears is usually not a desirable choice as it makes the system far more dependent on transport quality. With a top-caliber transport, the difference between "off" and "x1" is essentially zero, but with most associated gear you'll want to keep some form of resampling engaged.

As much as I love the improved amplifier section, the ground-up V590 DAC sometimes feels like the star of the show. Unlike most integrated DACs, it doesn't feel like a weak link at all. Synergy with the integrated headphone stage is absurdly good, with the signatures playing extremely well together. In short, only the finest external DACs will offer any significant improvement over what V590 has built in.





Lastly, the preamp section. I spent less time there versus the headphone amp and DAC section, as I just don't spend much time with my speaker rig these days. But from my somewhat limited experience I can say that it feels generally on par with the rest of the V590. The extensive inputs and useful remote (included with base model and Pro configurations) make this a more realistic option than, say, the Pass Labs HPA-1, which is by all accounts sonically amazing but also extremely minimalistic in terms of features. My system includes Merrill Audio Thor monoblocks and Usher Mini One Diamond speakers, and using V590 as preamp easily fits in the same class as the old Jeff Rowland and Krell units I have laying around. The V590 as preamp is neutral, dynamic, refined, and has that sense of "drive" which seems to characterize better preamplifiers. While going direct from DAC to amp can often lead to a sort of stark, bleached signature (particularly with already thin-sounding products such as Benchmark), V590 performs on the level of a true high-end preamplifier.


In the end, it's obvious that I'm enamored with the Violectric V590 and its superb performance. Whether used as a DAC, headphone amplifier, preamp, or any combination of the three, V590 is nothing less than brilliant. Faced with the difficult task of updating their classic V281, Violectric really nailed it, going above and beyond with the new DAC section as well. Those who only need a DAC or stand-alone headphone amplifier should obviously wait for the company to launch them as individual separates... but for anyone who might find an integrated solution useful, V590 earns my highest recommendation.

I've been using the Euphony PTS music server for a few years now. It's a simple, compact, silent source which does everything I could ask from a device like this. To get the most out of it, I pair it with a Keces Audio P8 linear power supply, and the resulting performance holds its own with far more costly devices I've owned or auditioned.

The secret sauce to the PTS is the Euphony OS, which is a dedicated audiophile platform based on Linux. I won't go into the technical bits here, but I'm convinced of its efficacy based on what I hear. They also sell the OS separately which can be installed to the hardware of your choice, and my experiments have shown it even brings a benefit to the most basic general-purpose PC. Particularly when combined with a quality PSU such as the Keces P8. All this time I've been building my own music-server/streamer devices with special USB outputs, SATA filters, and other audiophile-oriented parts, yet I've been missing out on significant performance gains by running Windows or my favorite Linux builds rather than a purpose built OS like Euphony.



As good as it is, the PTS was technically not an "official" Euphony product - the North American distributor had it specially made, and it worked quite well, but the Euphony team was not really involved in its creation. But now the group has decided to launch their own purpose-built hardware device, customized to work in conjunction with their operating system. They call it Euphony Summus and it starts at $2599 for the standard configuration with a 1TB SSD (more storage available for an upcharge). They do still sell the OS for DIY folks, but the version included with Summus has some further optimization which you won't get using your own hardware.

Summus includes a powerful i7 CPU with a generous amount of RAM, plus anywhere from 1TB to 8TB of onboard Solid State storage. The system lives in a fanless passively cooled case which is completely silent. Stock power is the standard switch-mode brick but aftermarket supplies from 12v-19v are supported. Note the stock supply is 19V/5A so any replacement will need to be fairly stout - my Keces P8 is 12V/8A so more than capable of doing the job.

The Euphony team has incorporated various tweaks specific to this hardware platform. Highlights involve severely throttling down the powerful CPU during playback to control noise, and evenly distributing playback tasks across all cores for even lower utilization. Then there's the "Ramroot" feature which has the system kernel boot entirely into RAM to avoid all disk interactions. These, plus various other behind the scenes optimizations, add up to an ideal platform for audio playback.




From a usage perspective, Summus is totally straight forward. Simply plug in power, Ethernet, and connect USB output to your DAC of choice, boot the system up, and (after giving it a moment to complete startup) point any browser to this location: https://euphony-audio.com/launch-euphony/. Clicking the button on that page will show any Euphony-based devices on your network, and choosing one will launch their Stylus interface. From there, we can play files directly or switch to one of several different modes - Roon Core, Roon Bridge, HQPlayer Embedded, Squeezelite, and various combinations of those options.



Switching to a Roon mode lets the system show up in Roon as any other Roon-ready device would. That can be combined with HQPlayer if that's something you might use. There's also an AirPlay mode which I've never tried. The standard Stylus mode is a full-fledged playback interface which rivals Roon in effectiveness if not quite overall polish - as a long-time Roon fan, Stylus is one of the few systems I don't mind using, but Roon just hits the spot for me. Still, Stylus is definitely attractive, intuitive, and full featured. It can manage the local library or connect to network storage (an Asustor NAS in my case) along with Tidal, Qobuz, and an option for internet radio (which I haven't messed with but intend to at some point). It all works quite seamlessly and I'd rank it just a few steps behind Roon overall, but definitely above JRiver, Foobar, and the other common choices as far as I'm concerned. This is of course a highly personal thing, so I'd never argue with someone for having a different preference than mine.



While I tend to use Stylus mode on a regular basis, I also appreciate having Stylus EndPoint available. That's a mode which sort of "tricks" Roon into "seeing" the Summus as an HQPlayer network endpoint, allowing it to work with the device in Ramroot mode. Just running Roon core or bridge mode alone does not enable Ramroot and thus, while sounding excellent in its own right (particularly with DSD upsampling in play) is not getting every last bit of SQ out of the system. I like to keep Roon running on a big server anyway, using my old (and water-cooled/massively overclocked) i7-5960X 8-core/16-thread CPU. So the StylusEP mode taps into that and generally sounds best when I want to use Roon. There is also an option for running Roon Core plus StylusEP from the Summus itself, removing the need for an external server.





In terms of SQ, I'm still exploring and contrasting Summus to the PTS. But so far I really like what I hear. There's an improved sense of immediacy, a more fleshed out soundstage with superior layering, and an inky black background which is even better than the already excellent PTS. In my experience, this level of performance easily takes us beyond 95% of the competition out there. I'm talking about the $5k-15k music servers of the world, but also crazy-expensive disc transports which cost as much as a Honda Civic. I've had more than my fair share of that sort of thing in my rig and I'd take the Summus over any of them.

One interesting thing I've noticed so far - if I remove the Keces P8 linear power supply and just stick with the stock PSU, Summus performs roughly on par with the PTS/P8 combo. Which, if I recall correctly, was priced within a few hundred dollars of the Summus. The newer device sounds very slightly better in a few areas but it's not far off. The difference then involves significantly higher system horsepower for those running Roon Core mode with upsampling, convolution, multiple zones, etc. Adding the upgraded PSU takes us up into the stratosphere.

Interestingly, I do have a single DAC (BMC UltraDAC) on hand which pairs better with the original PTS device. It just gives more solidity and despite a less expansive soundstage, I still find the PTS/BMC combo more convincing. But system synergy is a real thing and I never expect any single DAC/amplifier/headphone/transport/etc to be the best in every single case. Aside from that one instance though, Summus is the clear winner with every other combo I've tried - and I've tried a bunch.

I also enjoy using it with a quality DDC such as the Matrix X-SPDIF II. Power that by the same Keces P8 and the combo is superb for feeding my older/non-USB capable DACs - in my case from Sonic Frontiers and Theta, but I can think of many others from Levinson, North Star Designs, Audio Research, etc which pre-date USB. Not to mention the various modern DACs I've experienced which just don't sound as good via USB, regardless of using the latest technology. I then add my Titans Audio Lab Helen to the mix to de-jitter/clean up the signal even more, and the result is a world-class (if somewhat overly complex) chain that will challenge all comers regardless of price. The Helen doesn't do USB so I only use it with AES paired with the Matrix DDC - it does add a small amount of SQ which in some cases is worthwhile, but in other cases ends up not being very noticeable.



I'm not just using hyperbole here - over the past few months I've directly compared the Euphony Summus with the Keces P8 to some heavy hitters and it's always come out on top. These contenders mostly sell for significantly higher prices... at least double what Summus costs, but in some cases quite a bit more than double. I'm talking Totaldac d1-player, Aurender N10, Wadia M330, Lumin X1, and some of the higher recent models from Linn and Naim. All of them had enjoyable aspects (except maybe the Wadia... I didn't care for that one) and I'd particularly like to revisit the Lumin as an integrated streamer/DAC when I have the time. But in terms of transport usage, none of them impressed me as much as the Euphony Summus with P8 power supply, with or without the Matrix/Titans gear in the chain. The closest some of them came (such as the Totaldac) was matching the performance of my complex little chain, which is fine until we consider pricing where Summus comes out far ahead.


The Euphony Summus - while certainly not cheap - reminds me once again that it is rarely the most pricy gear that performs best. Sometimes a clever designer will extract more sonic goodness, for a lower price, and do so without much fanfare. The Summus is just this sort of device. If you are seeking top-caliber playback in a music server/streamer/whatever you might call it, Summus is definitely worth a look.

Kennerton Audio is the upscale division of Fischer Audio, which - although perhaps still not so well known in some circles - has been around for quite a while by now. I remember first using the Fischer FA-003 about a decade ago, and that same pair is still going strong after being gifted it to a friend who uses them daily. While I always felt Fischer made high-value gear, Kennerton is significantly more upscale - with pricing to match.

Kennerton is not just throwing together fancy looking headphones and charging a bundle for them. The company is known for their beautiful wood cups which come in a wide variety of finishes, but they also have their own custom-designed 80mm carbon fiber planar magnetic drivers. As far as I can tell, these same drivers are utilized in several headphones in their lineup, with each model using unique tuning, cup designs, and headband assemblies to achieve very different results. Kennerton also has some dynamic models using interesting designs (graphene drivers, horn loading) which fall beyond the scope of this writeup.

The focus of this writeup is their latest model, the Kennerton Thekk. I haven't spent a ton of time with their other products such as the Odin, Magni, or Vali, and I've never heard the Thror or Gjallarhorn, so this review will mainly focus on the Thekk and its place among other headphone models from outside the Kennerton stable.

The Thekk sells for €2,680 which is roughly $3,000 at time of writing (but will fluctuate depending on the exchange rate), and can be purchased directly through their website. I'm told there is not currently a specific North American distributor at this moment, but readers in other countries may have more luck with local dealers.

$3000 is not an insignificant sum for a set of headphones. In terms of pricing alone, that puts Thekk in the same general ballpark as Audeze, Sennheiser, HiFiMAN, Focal, and others, all of whom would like to claim supremacy in the high-end headphone category. So what does Kennerton give us for that money?



Build and Presentation
Thekk's packaging and cable both feel appropriate for the price. I've seen more luxurious packages but also far more sparse presentations for the money, and Kennerton's leather storage bag is actually more useful than most headphone boxes. The stock cable is good enough to where I don't immediately want to replace it, and feels reminiscent of the Fostex TH-900 cable which I've always appreciated. But thanks to their choice of mini-XLR connections, Thekk can accept the same cables as Audeze, ZMF, Meze Empyrean, and probably others that I'm forgetting at the moment. I love that, and frankly wish it would just become the standard at this point.

My Thekk has Bubinga cups which are simply gorgeous. There are plenty of other options, all of which look nice from the pictures I've seen, so you really can't go wrong. Being an open design, I doubt the wood choice influences tonality much, but you never know.



The best part about Thekk is the comfort. The light weight (listed as 390g but I swear it feels like significantly less) plus the well-done self-adjusting headband system makes for an effortless feeling when worn. Clamping force is light enough for extreme comfort but not so light that it feels unsecure on the head. The (real leather) pads are excellent as well, making the whole experience nearly perfect.

My main issue is the fact that the cups do not swivel from front to back. As you can see from the pictures, the headband assembly holds them in place with no swivel mechanism in that direction. The design relies on angled pads to help conform to the human head, which is of course wider behind the ear. This mostly works for me but I could use another few degrees of angle for optimal fit. I've messed with trying to bend them slightly - they are somewhat flexible - but the design does not lend itself well to permanent bending in that particular direction. And frankly I don't feel I should have to resort to that in a $3000 headphone. In the end I do achieve a good fit but I could see how it might be an issue for some people.



My other (very minor) quibble is the fact that this same headband assembly is used in several of Kennerton's significantly less expensive models. That feels a bit weird when dealing with a headphone in this price class. I suppose Focal and Audeze do the same thing, but at least they throw in some carbon fiber bits to help differentiate the higher models. Again, I really like the self-adjusting design overall (apart from the non-swivel cup mounts) as it reminds me of one of my favorite old-school designs - the K1000. No, not AKG's classic ear speaker, but the rare Kenwood KH-K1000 which had a conceptually similar headband design to Thekk (but it actually swiveled). If Kennerton could somehow make Thekk swivel it would be just about perfect for me.




Sound Signature
The Kennerton Thekk is characterised by its lightning-fast, highly-detailed presentation, which manages to avoid feeling cold or clinical. Low-level detail is fantastic - in the right system, with a good enough recording, it will make the hair on the back of your neck stand up. While that aspect is what initially jumped out at me, I need to clarify that this is not an in-your-face bright headphone. The presentation is generally neutral overall which makes for a versatile experience without any significant flaws.

The treble approaches the most accurate and resolving I've heard regardless of driver technology (dynamic, planar, even electrostatic), which was actually something of a surprise for me. The entire Kennerton line, while very nice to look at, has a sort of steampunk vibe to it, in contrast to the more high-tech feel of a Focal Utopia or Sennheiser HD800. So I figured it might have a lovely tuning with only moderately strong technicalities (which is how I would describe some of the older Kennerton models such as the Vali). And yet, when it comes to digging out extreme detail from a great recording, Thekk is right up there with Stax, HE-6, Utopia, etc, all of which I feel outperform the HD800 in treble performance if we consider the total package and not just the initial "wow" factor. HD800 initially might feel superior in resolution but flaws are readily apparent after further listening. Meanwhile Thekk is not overly sharp or peaky, not sibilant (unless it's in the recording), and does not feel bright when paired with neutral equipment. Seriously - if you love treble clarity, this is a headphone you really need to try out.


The great part about Thekk is that it isn't a one-trick pony. That exceptional treble performance is accompanied by engaging midrange and very fast, well textured lows. Tonality is thick enough to keep from being shrill or losing timbral accuracy. Bass speed is phenomenal, and impactful enough to feel believable. There might not be enough slam to satisfy bassheads, but those who like a generally neutral presentation should be pleased. I personally wouldn't mind a roughly 2dB bump below 80Hz but it's not a big issue assuming I use the right amplification. Did I mention the bass is fast, tight, and really accurate?



Listening to double bass virtuoso Larry Grenadier's recent release The Gleaners is thoroughly impressive, as is pretty much all jazz and folk. Metal is also fantastic - I very much enjoy using Thekk for highly technical metal from Obscura, Meshuggah, Revocation, Fleshgod Apocalypse, etc, as the speed and accuracy make for a nearly perfect match. Grimey underground hip-hop or rumbling drum n bass? Still quite good though perhaps not ideal for this sort of presentation - I tend to go with a warmer/more forgiving headphone in those cases. Still, I'd say Thekk is highly versatile overall, and never sounds completely out of touch with any genre.

Thekk is also very open and layered, which makes it beautiful for all sorts of classical music. While staging is slightly less wide than HD800, depth is clearly superior, which to me is the more impressive achievement. Utopia feels a lot more intimate, and HE-6 can be huge but imprecise, but Thekk feels "just right" to me.



A Few Comparisons
The last Kennerton I heard was the Vali and that was a fun, somewhat laid back headphone with very punchy bass. Thekk is something different entirely, playing in a significantly higher tier. Unfortunately I haven't heard their Thror model so I don't know how Thekk compares - Thror is priced slightly higher but Thekk is newer, and from speaking with the folks at Kennerton I get the impression it may be the more complete sonic package. I have heard Thror described as having lovely detail but lacking a bit in fullness - if that's true, Thekk solves that issue, but I'm speculating here.

Since using Thekk for the past few months, I haven't had any desire to reach for an HD800, HD800S, or Utopia. Thekk just scratches my itch for detail whilst retaining "musicality" - that dreaded generic word - more than any of those. HE-6 or Susvara, when driven by herculean amplification, can match or even exceed the Kennerton, but that's easier said than done... only a select few amps can take us there.

I do hear some similarities to the HiFiMAN HE1000 family. Not necessarily in overall frequency response but rather with the lightness and speed of the sonic portrayal. Thekk fits somewhere in between my particular HE1000 v1, which sounds slightly richer, and HE1000SE which is lighter and more threadbare in tonality. But Thekk avoids the sense of "softness" in the transients that I hear from HE1K (both models, to some degree) so I think it may be the superior headphone in many cases.

Lastly, the Meze Empyrean is an interesting comparison. Both headphones sell for just about the same price, and both are attractive and well built - each in its own way. Empyrean has a beautifully organic, warm, smooth presentation which still does detail retrieval quite well. Conversely, Thekk has more of a focus on treble brilliance, but does not go so far as to become cold or clinical, and it offers very clean, tight bass response despite that aspect not being the focus. It really comes down to whether one prefers a top down (Thekk) or bottom up (Empyrean) approach. I tend to reach for one or the other based on what I'm listening to, as well as the system I have assembled at that moment.



System Matching
The level of performance I've described does require a very nice system to achieve. When used with more modest gear, Thekk takes a few steps back in terms of clarity and fullness. And with a bright or thin setup it does become somewhat unbalanced to my ears. So despite being fairly easy to drive in terms of volume, it ends up being fairly demanding when it comes to system matching.

My reference amplifier for Thekk is the Niimbus US4+. That's pretty much the ideal amp for getting neutral, fast, highly resolving results from this headphone. I then choose my DAC for a slight flavoring - Resonessence Mirus Pro Signature for supreme resolution, Exogal Comet Plus for a more reserved top end with added midrange oomph, or ModWright Oppo 205 for a focus on textural thickness. Thekk plus Niimbus US4+ very clearly highlights the signature of each source, and is great as a review tool.

Interestingly, the SparkoS Labs Aries comes very, very close to the Niimbus level of performance for significantly less money, so that's definitely a solid recommendation as well. It's single-ended only but so is the stock Thekk cable so that's not a major drawback. Not a ton of info about the Aries around these parts but I highly recommend it - pretty ruthless in terms of demanding a great source, but it really rewards the listener.

The Cayin HA-6A SET amp brings out a bit of midrange bloom and opens up the soundstage even more than the Niimbus. Upper mids and treble still sparkle, but the focus shifts more towards the presence region. Again, an excellent combination, which could be tweaked even further with tube rolling.

Schiit's Asgard 3 makes for a surprisingly good match considering how affordable it is. It takes a bit of emphasis off the treble but still allows for clarity with reasonable extension. Mids are rich, bass is tight, and the whole thing is really engaging. Apart from some refinement and resolution, the largest sacrifice I hear is soundstage, which feels restricted compared to the above amps.



All-in-one DAC/amp units offer a great value - assuming both aspects perform to a high enough level. Thekk doesn't always pair well with these types as it tends to sound best with fairly stout amplification. A lot of the integrated units I tried (Mytek, Benchmark, Chord) left Thekk sounding threadbare, overly soft, or just plain boring.

Wyred4Sound's intimo is a great match though, and actually proved to be one of the best ways to enjoy this headphone. It gives a large portion of the performance I hear from Wyred's upscale Anniversary DAC paired with my Pass Labs HPA-1, but for a mere fraction of the cost. $1500 isn't cheap, but this is among the best "value" propositions I've heard from Thekk.

Also sounding excellent is the Violectric V590, which takes the amp stage of their famous V281, upgrades it, and then packs an excellent AKM-based DAC in the same chassis. V590 is in the same sonic ballpark as intimo in terms of signature, but expands on it in both performance and price. I recently had a friend visit who is big into speaker-based audio but totally out of the loop with modern headphones. This is the combo I chose out of all my gear to show him how far the hobby has progressed since the old HeadRoom Desktop/AKG K701 days. He was absolutely floored - and that's coming from someone who owns some crazy nice Rockport speakers.


Conclusion
Kennerton Audio has always felt a bit mysterious to me. Their wood-clad headphones are consistently beautiful yet have a wide range of sound signatures - I don't really think they have any persistent "house sound". That's actually refreshing as plenty of headphone makers tend to churn out one incrementally different headphone after another.

With Thekk, the company has achieved something very close to my view of neutrality. The focus is squarely on detail and resolution, without compromising listening engagement across the rest of the spectrum, and the presentation is extremely open and layered. Feed it mediocre music on a bright system and you'll likely find yourself unimpressed, yet Thekk handsomely rewards a quality chain - in that context it performs in an elite field with only a select few competitors.

I still wish the cups swivelled. And - like all high-end headphones these days - I wish it was less expensive. Apart from that I love this headphone. Despite my natural preference for the warmer Empyrean-style sound, I find myself choosing Thekk regularly for all genres of music. Anyone looking for an alternative to Utopia, HD800S, etc may find Thekk to be their ideal match.

Custom IEMs tend to be expensive, and there's a good reason for that. Even when dipping into the murky waters of no-name IEM brands from overseas (I'm based in the USA) - where admittedly there are occasional gems to be found - I generally prefer sticking with universal designs. Customs require a level of care and customer interaction that doesn't really lend itself to language barriers or unproven operations. This translates to entry level customs from 64 Audio and Ultimate Ears starting at $499, whilst Noble Audio, Empire Ears, and JH Audio charge $599 for their most affordable offerings.

That said, many people (including myself) find custom IEMs generally superior to their universal counterparts. I've had the opportunity to try many designs in both universal and custom form, and nearly every time the custom sounded easily superior. Not saying universals can't sound amazing as well, but perhaps it is easier to extract high performance from a custom build where the designer has more room for ideal component placement. Whatever the case, comfort is obviously a strong point as well - once a good fit is obtained, CIEMs just disappear into your ears, and I can use them for hours without any issues.

That creates a bit of a problem for those desiring CIEM advantages but lacking the budget for even an entry-level model. If only there was an established firm, known for high quality (in both product and customer service), who sold a CIEM for significantly less than the previously mentioned $499-599 starting point.



Enter the Lear Audio LCM-Skyline.

Based in Hong Kong, Lear has been around for well over a decade, and has built a reputation for making great CIEMs at what I consider reasonable prices. I still use their BD 4.2 dual-dynamic/quad-armature hybrid on a regular basis, despite it being almost five years old by now (practically an eternity in the rapidly advancing high-end CIEM world). Lear does have some new cutting-edge designs in the works but is also very focused on the other end of the price spectrum - a segment with far more potential customers.

The Lear Skyline carries a base price of roughly $175 (HK$1388 to be specific). That price gets you the Skyline in your choice of translucent, opaque, or glitter black, with a standard cable featuring an MMCX connection. Before we go on, I feel the need to point out - $175 is a crazy low price for a custom molded IEM! Obviously there's an additional cost for ear impressions, plus shipping (in this case to Lear's Hong Kong office), but those are costs which apply to any CIEM purchase.



Additional options will bring the price a bit higher, but it remains considerably lower than other entry-level CIEMs. I opted for Lear's proprietary "Combo Con" connection system which made the whole thing HK$1688, or approximately $215 based on the exchange rate at time of writing. Opting for Lear's Bluetooth cable, along with the Combo Con, brings it up to HK$1888 or ~$240. Beyond that it's just the usual upcharges for aesthetic touches if you want a fancier shell or faceplate design.

The LCM-Skyline is a single-driver design, which is not unexpected at this low price. What is slightly unusual is the fact that it uses a dynamic driver rather than the more traditional balanced armature style - where I've seen many budget examples floating around (which usually don't sound all that great). Lear uses a custom-made 8mm driver with a PEEK diaphragm, N50 neodymium magnet, and CCAW coil, wired with high-quality OFC wiring. Using a single full range driver makes crossovers unnecessary - a good thing in my view, as coherency is often something of a weak spot in budget multi-BA-based CIEMs.



Lear's "Balanced in-ear pressure" system incorporates a vented shell to reduce pressure and thus listening fatigue. Think of it as a simple version of the Adel/Apex systems used by Empire Ears and 64 Audio respectively. This design has no removable module, and the vent is thoughtfully placed at the "top" of the CIEM, which inserts into the cymba concha (or under the ear "flap" as some call it). Based on the anatomy of the human ear, this portion has an angle to it which means the port won't be blocked by direct skin contact. But it's also tucked away such that it doesn't really leak sound. That said, care should be taken to watch for earwax that could theoretically clog up the ports - I've seen other brands use a layer of mesh as protection, but these are just open. It's small enough to where I'm not worried about it based on my own personal earwax situation.

Let's talk about the Combo Con system. Independent stress testing shows standard MMCX loses 30% of its mating strength after as little as 15 inserts/removals (sorry not allowed to link that report), and is typically rated for a minimum of 500 pairing cycles. That's enough to last a while but could be problematic for long-term users or just people who remove their cables often. Meanwhile the Combo Con system is rated for 3000 cycles and can then be replaced if it does wear out. Whether designing a budget CIEM or an expensive flagship, this seems like a great foundation to build upon.



As mentioned previously, I got my Skyline in black with the Combo-Con upgrade, representing a roughly $215 expenditure. That's less than half price compared to entry-level models from mainstream CIEM makers. Despite that, build quality is solid on the 3D printed shell (typical for every Lear product I've seen), and there's nothing about the Skyline that screams "budget". My fit is excellent, but I've been doing my own impressions for years and as a result rarely suffer with fit issues.

The package is pretty basic: IEM, a fairly standard (for modern CIEMs) quad-braid cable with a nice looking 45 degree plug, and a carrying case, plus a wipe cloth and the usual cleaning tool we've seen so often over the years. This all comes in a nice looking Lear-branded box, but it's nothing fancy. Which is fine considering the cost... overall I have no complaints whatsoever.



Skyline is rated at 16 ohms with a 106dB/mW sensitivity. This is a fairly easy IEM to drive. I used an LG G8 ThinQ, Shanling M0, and even an old iPhone 6S (last iPhone with a real headphone jack) and all were plenty capable driving the Skyline to absurd levels. I did not experience hiss with any DAP or phone I tried, though big desktop amps will often reveal a bit of noise as they usually do with IEMs. This also relates to the DAC being used, the transport, and your power situation in general, as ground loops and other gremlins will be most obvious with a sensitive IEM. I find that using a quality power conditioner helps quite a bit in my particular system, and I achieve great results from an Equi=Core 1800 for balanced power, but there are many other worthy directions that could be taken.


Listening
The signature of the Skyline immediately stands out as warm and somewhat smooth up top, with a meaty tonality and prodigious bass extension. It's definitely punchy down low, but also has some midbass warmth - though not so much to be troublesome in my opinion. My initial listening involved various underground hip-hop releases: Little Brother, The Weathermen, Homeboy Sandman, The Dynospectrum, Propaganda, and Atmosphere. The Skyline signature is just about ideal for that sort of thing - warm where you want it, clear enough in the midrange, and forgiving in just the right ways.

Despite the midbass - which definitely seems like a deliberate tuning rather than unwanted bleedover - the low bass remains deep, textured, and surprisingly well controlled. Particularly when we consider the price. From the elegant double bass of Ron Carter to grimy Bass Mekanik test tracks and everything in between, Skyline gets the job done right, and surpasses the performance of many/most entry-level CIEMs using balanced armatures. Bassheads of all types should be thoroughly satisfied here.

Although there is clearly a low-end focus, Skyline is not over-the-top mushy or dark. Detail retrieval is commendable, resolving enough of the intricacies on Tigran Hamasyan's Shadow Theater to keep me engaged. Vocals, whether from Damien Rice or Etta James, have a nice sense of projection, and Hiromi Uehara's Yamaha CFIIIS concert grand piano has convincing enough extension to not sound muffled - as is often the case with overly-warm IEMs. While the Skyline is not a detail monster, it does well enough to where I don't feel short changed as I sometimes do with bass-oriented IEMs.



Where it really shines is that murky world of "musicality" - a term many reviewers detest (perhaps rightly so). Still, I think most people understand what I mean. Older audiophiles may use the term "PRaT", which is of course another imperfect label attempting to capture both a sound and the feeling it invokes. Whatever we decide to call it, Skyline has it. I find myself bobbing my head, tapping my foot, and - in various other ways - being more engaged with the music on a physical level. This is a very uncommon reaction when listening to most single or even dual armature-based designs on the more affordable side of the spectrum.

Soundstage is surprisingly large, though with more width than depth, and imaging is accurate enough to not feel vague or blurry - another shortcoming I've experienced with entry level custom IEMs. Again, the lack of a crossover likely makes this possible, and I suspect the vented shell design may help as well. The overall feel seems appropriate considering the signature and target market of this design.

Is Skyline perfect? Obviously not. No IEM, including those selling for several thousand dollars, can make that claim. Lovers of classical or jazz who find themselves gravitating towards the more detailed signature of an HD800, classic Etymotic, etc, will likely not find Skyline an ideal match. And folks who prefer a more neutral, low-impact presentation may find the bass little overwhelming. But I think the general music enthusiast who enjoys a wide variety of genres, and levels of recording quality, will find Skyline appealing. It has that generous low-end kick without going completely overboard, whilst maintaining enough presence and impact in the mids and treble to still feel reasonably well balanced.



If you are looking for something higher up the food chain, Lear certainly provides many more advanced models, but Skyline is still very compelling as-is. When we consider the extremely low price, it's an easy recommendation for anyone seeking their first CIEM, or perhaps building on their CIEM collection with an affordable and fun option.

Long story short: Oppo Digital used to be a player in the audio market. They had some of the best universal Disc spinners around, which were very highly regarded both for their audio and video performance. They also had an excellent line of planar magnetic headphones - the most popular of which seemed to be the PM-3 ($399).

It's been nearly 2 years since Oppo left the market, and many of us still miss their contributions. I regularly use the Oppo UDP-205 player (with ModWright upgrades, though the original was great too) and wish I had purchased a few more when I had the chance (second hand pricing has skyrocketed). But the Oppo product I arguably see most demand for is the PM-3. It occupied a relatively niche spot - a great sounding, portable, planar magnetic design which is still fairly appropriate for home use. Not a ton of competitors in that category.

Drop (formerly MassDrop if that name is more familiar) made a deal to purchase some of Oppo's intellectual property, and has put it to good use with the new Panda headphones. It takes the beloved PM-3 and augments the design with wireless capabilities, whilst maintaining the same $399 asking price. On paper, that sounds like a winning combination right?


As with any headphone design, implementation is key. Transforming a traditional wired headphone into a wired design offers many opportunities to screw things up, but fortunately Drop did a great job here. They not only kept the wonderful PM-3 sound signature but actually improved it to a small degree, which is not something I expected.

In order to make this happen, Drop uses built-in discrete amplification with THX AAA technology. This drives the headphone directly - no intermediate sound shaping through digital signal processing or noise cancellation, which are usually found in this sort of headphone. That means excellent battery life (30 hours or longer) and no major change in sound when switching to passive mode using the included headphone cable. There are two microphones on board but so far they are only for use with phone calls. Drop has toyed with the idea of later adding some sort of ANC and/or DSP for user adjustable EQ, but that's all speculative at this point. Just know that for now, what Panda offers is a direct-to-driver wireless experience that mirrors what we get when driving it via quality (external/wired) amplification.


Bluetooth capabilities are fairly comprehensive: I could be forgetting a few, but so far I recall basic SBC plus AAC, aptX, aptX HD, aptX Adaptive, and LDAC (although LDAC was not yet implemented in the firmware on my review prototype). I don't always care about having so many options but with a headphone of this caliber, the improvement from one codec to the next is more noticeable. I mostly listened with aptX HD and was very satisfied with what I heard. LDAC - at the highest 990kbps setting - has potential to sound even better, so I'm glad it is present here even if I could not test it out for myself.

Controls are handled via 4-way rocker which also does push-to-click. It's really low profile and I initially had trouble finding it while wearing the headphones. Eventually I got used to it, and actually quite like it now. I find it simple and intuitive but your mileage may vary, and I know some people may prefer side mounted touch-controls or other solutions. Charging is accomplished via USB C and going from empty to full supposedly takes around 2 hours. Note that I've heard talk about connecting Panda directly to a phone or other device via USB, which would be a handy option to have, but I never actually tried it myself. Not sure if that will end up being a feature in the production model or not.


Panda relies on strong passive isolation to account for its lack of noise cancellation. That means a fairly strong clamping force - no way around that if isolation is the goal. Combined with the pads which don't quite fully envelope my ears, I find the Panda less comfortable than larger at-home models like the HD6XX or K7XX. That's sort of the nature of these things though, and I felt pretty much the same way about the original PM-3. The prototype model caused a bit of a "hot spot" on the crown of my head due to a lack of padding, and Drop says they are working on improving that aspect. They are also adding a bit more extension (1 or 2 clicks) as some of us with big heads are pushing the limit for the prototype.

Now, before you go assuming this is merely a regurgitated PM-3 with some electronics stuffed in the cups, take a look at pics and see how different Panda looks compared to the original Oppo. The general appearance is very similar but when you get down to details you'll see substantial redesigns at every turn. Which is logical - it's not like Drop could just take over production with a few minor tweaks, as PM-3 hasn't been produced in several years. They basically had to built it from the ground up, using PM-3 as a target and working in all their extras without compromising anything. And I'd say the mission was a success.



As for sound, well, I found the Panda to be quite similar to the PM-3 I had a while back. That means somewhat U-shaped signature, though generally neutral enough to not sound overly colored. Bass impact is generous but not overwhelming by any means, and seems improved over the PM-3 which at times felt just a little muddy. The somewhat forward lower treble keeps things interesting and is followed by a subtle roll-off in the higher frequencies which makes the sound smooth and non-fatiguing. Imaging is pricise, with a somewhat focused soundstage that is accurate if not all that massive... again, very similar to the original Oppo design.

The end result is a slight improvement over the PM-3, both at the top and the bottom of the spectrum, without losing any of the fundamental charm so many people loved. Disregarding any of the wireless aspects and just using this as a traditional wired headphone, it seems worthwhile for the asking price. I had a blast using it in a ridiculously mismatched system featuring Pass Labs amplification and a McIntosh CD player/DAC, though it also sounded great from my "quad-DAC" equipped LG phone. True, the market has moved on since the PM-3 originally sold for $399, but I'm not aware of any other headphone that came along and filled that particular gap. Used PM-3 models still command a hefty price on the second-hand market so based on that alone, the Panda looks pretty appealing. Granted the bundled cable is rather short for home use but that can be remedied fairly easily.

The part that makes Panda outstanding is when I unplug and walk away from the big system whilst maintaining the same general sonic performance via aptX HD. Between this and the Sennheiser Momentum True Wireless, I'm enjoying cable-free listening on a level I never imagined just a year ago. I'm not even really the target market for this thing as I mainly listen at home, but even I am excited to see the Panda launch - I suspect it will be among Drop's biggest projects to date, right up there with HD6XX and K7XX.

Since this whole thing is based on a few weeks of use with a prototype Panda, I will have to stop short of going further in-depth in terms of comparisons etc. I guess the bottom line for me is this - I'm not much of a portable user, and even I want to buy the Panda. That's about the best real-world praise a product can earn. Panda currently has just over 2 weeks left on the Indiegogo campaign (which has proven to be wildly successful) so take a look and give it serious consideration if you may have use for such a device.

It's been well over a decade since I started reviewing audio gear. As you can imagine, I've ended up on the mailing lists of quite a few audio companies and PR firms since then. I get solicited for reviews on a regular basis but unfortunately only have time to accept a very small fraction of the offers. This means the item in question must be fairly compelling - maybe a new product from a well-established firm, or an item that has lots of buzz on the forums, or perhaps a new technology that I find interesting... something to stand out from the many other options available.

The item under review today falls into the latter category. Some months back I was contacted by a company called AME Custom, based out of South Korea. AME is a relatively new firm who at the moment offers just a handful of models, in your choice of universal or custom designs. My regular readers know that I vastly prefer customs due to standard IEMs generally fitting my ears poorly. As such I've got tons of experience with CIEMs but am not usually all that up to date on the latest/greatest universal options. So AME agreed to send me a custom molded set of their Radioso ($1350) for review.


Design
Radioso is what AME terms a "Tribrid" design. That means a triple-hybrid using three different types of transducer technology, arranged to leverage their strengths within the sonic spectrum. We get a dynamic driver specifically handling low frequencies, a balanced armature for midrange, and a quartet of electrostatic drivers dedicated to the highs, with signals routed by a 3-way crossover. Again, these are the stereotypical areas where each type of driver generally performs best, so the design seems to make a lot of sense. Note that the dynamic driver gets a small vent, placed just below the cable socket and protected by a filter to avoid any debris entering the shell. Unlike some other vented CIEMs I own, this does not significantly impact isolation.



There are several other tribrid options out there using some variation of this same theme, including the Noble Khan, Jomo Trinity, Vision Ears Elysium, and probably some others I'm forgetting. They all use different arrangements in terms of driver count and configuration, not to mention crossover tuning etc, which means they should all sound unique despite the presence of all three types of drivers. Worth noting is the price difference involved - those models all live in the $2,500 to $3,000 neighborhood, whilst AME sells their tribrid for less than half that price. Note that a universal version is available for $1200, while the custom option is $1350.

AME offers another model called Argent which is a totally different 6-driver design using dual electrostatic highs and quad BA drivers with a 4-way crossover. It is priced identically to Radioso, so I guess we could say co-flagship status. It may have launched just a bit prior to Radioso, and there are a few good reviews to be found here. As far as I can tell, AME seems to have positioned Argent as semi-neutral. Their words are "immersive, well-rounded sound featuring warm lows and sparkling vocals". Meanwhile Radioso is described as having "rich low-end response while fully highlighting the unique characteristics of the electrostatic tweeters". I have not heard Argent so I won't comment any further as far as comparisons go.

There are some (non-sonic) compromises apparent when looking at the Radioso compared to the more established tribrid competitors. Where those models offer numerous customization options and extravagant aesthetic packages (some included in the base price, others for an extra fee), AME limits Radioso buyers to fully clear or clear with wooden face plates. They mention "custom artwork" available for a $300 upcharge, though I'm not certain what that entails. Regardless I don't think AME is trying to compete when it comes to fancy builds. Don't take me the wrong way though - my (completely clear) review pair seems built to a very high standard, and pics of the wooden face plate option look very attractive too. The Argent model has abalone pearl instead of the wood option and that also looks to be quite well done. So it's a matter of simplicity and limited options rather than poor quality. Still, for less than half the price, I don't feel this is unreasonable, and my experience is that newer CIEM companies tend to branch out with more options as they become more established.



As far as packaging - the box is a simple and functional affair which showcases the IEMs as soon as you open it. Included is a metal storage case similar to that used by a few other brands, along with the typical cleaning tool and 1/4" adapter. Universal versions ship with a few tip choices, and both models get a leather cable organizer with "AME" logo which I don't find particularly useful. It's not the most comprehensive bundle I've ever seen but it gets the job done well enough.











The AME site mentions their "6N OCC Pure Copper Silver Plated Cable" with the classic 2-pin style connection. The one I received appears different from what I see listed on the product page or reviews of the Argent model. The pics show a quality thin silver cable with nice chrome-looking Y-splitter and angled plug, but mine appears more like the classic cables used by Westone and Ultimate Ears back in the early CIEM days, complete with plastic molded bits. It does the job well enough but feels somewhat delicate - I'm glad most IEM brands use a more beefy cable these days. Granted, back in the early Westone/Ultimate Ears days these cables proved very reliable, so who knows. I did most of my listening with a bespoke Effect Audio pure silver cable anyway, so I can't speak of any compromise involved in the stock cable nor the specific cable I happened to receive.








Setup
After a quick test to confirm proper fit, I left the Radioso to burn in for well over 100 hours. That's usually the goal I shoot for, as a sort of "just in case", but this time I got tied up and couldn't get back to it as planned... so it probably got more like 200 hours total. I did not monitor to see what (if any) changes took place during that time.

My initial listening, as far as portable devices, was with a Samsung Galaxy S10. I then upgraded to an LG G8 thinQ, followed by a side-grade to the diminutive Shanling M0, and finally the Cayin N5iiS. I don't currently have any higher-end DAPs on hand, as I'm generally not much of an on-the-go listener at this point in life.

So I brought out the big home system - Equi=Core 1800 balanced power conditioner, Nativ Vita music server streaming via Roon from an Asustor NAS in the other room, McIntosh MCD600 spinning CD/SACD titles as well as acting as DAC for the Nativ via SPDIF connection, and Pass Labs HPA-1 amplifier, with Audio Art cabling all around except for the previously mentioned Effect Audio IEM cable. This rig is resolving and powerful enough to bring out the best in any headphone, yet utterly silent even with the most sensitive IEMs.








Sound
My first impression came straight from a Samsung Galaxy S10, playing Qobuz lossless files. Right off the bat I could tell two things about the Radioso: first, that it had a beautiful, dynamic, and very slightly U-shaped signature with superb extension on both ends. Second, it really needed a better source to shine. I found myself cranking the volume to nearly the maximum levels on the S10, and I'm not usually a high-volume listener by any means. The bass had great impact but clearly ran out of steam down low, and staging was essentially flat. Inexperienced listeners might stop right there and dub Radioso merely a decent IEM, but I took these issues as signs pointing to the need for a better player.

Switching to the LG G8 thinQ (still playing Qobuz) which has the same DAC implementation as their audiophile-approved V-series, I proved my theory correct - Radioso performance was excellent. It still had that mild U-shaped tuning overall, but really opened up the presentation in all dimensions, and the frequency extremes both started filling out. I played through most of Avishai Cohen's Seven Seas, then a few tracks from Philharmonics by Agnes Obel, and finished up with a mixture of Carpenter Brut, Moose Blood, and classic hip-hop supergroup Def Squad - just to see how Radioso deals with a wide range of genres.



Through it all I heard the same superb clarity in the upper mids and highs, controlled yet weighty midrange, and rather generous low-end impact. At this stage I would call it a somewhat bassy signature though not overdone in most situations. Midrange was just a touch set back in the mix with respect to lows and highs, though not enough to feel recessed or lacking - in fact mids were very articulate and thick enough to be satisfying with all sorts of material. And those electrostatic highs? Despite there being a quartet of drivers handling the upper frequencies (compared to just one driver each for lows and mids) the presentation did not feel overly bright, sharp, or aggressive. While detail was plentiful, and clearly more extended than most (or all?) of my other CIEMs, it felt very smooth and natural. I could listen for hours or use higher-than-normal volume levels and not experience any treble fatigue, which is not a common occurrence for an IEM that isn't dark or rolled-off up top.

Complaints? I felt the bass was just a touch overbearing at times with the G8, particularly with more subdued music like Agnes Obel or the slower tracks from Avishai Cohen. While prodigious extension is always welcome, I could have used a bit less midbass from time to time. And again I still had a suspicion that I wasn't properly driving the Radioso to the fullest potential - I was still cranking the volume significantly higher than usual to get the desired outcome.

Switching again to the Shanling M0 brought out a different set of strengths and weaknesses. This time the lows were more controlled and midbass did not intrude where it wasn't welcome. Midrange was rich and bold, and the upper midrange had excellent energy. Unfortunately the highs now seemed a bit more vivid than I would find ideal. Not a problem at all with many tracks, but at times I found it a tad overwhelming. What may start as being dynamic and fun ends up being fatiguing as time goes by - trumpets have too much bite, certain vocals become crispy, and cymbals in particular have too much energy. Some people are more sensitive than others to the issue, but for me it makes for one of the few times I would actually prefer the LG G8 to the Shanling M0.



Moving upscale to the Cayin N5Sii solved this problem. Now everything worked in harmony to provide a big, bold take on the mild U-shaped signature, without any of the previously mentioned shortcomings. Bass was tactile, weighty, yet very well controlled. It had a sense of impact and definition that only a superbly design IEM can provide - it's as if I could feel the air pressure moving as I would when listening via speakers. Midrange texture was beautiful, with gobs of detail and rich tonality. Highs were ethereal and airy, with seemingly limitless extension. The presentation was open and spacious on a level only a handful of top-tier IEMs could hope to match. In short, this was among the very best I've ever heard from an IEM at any price. Which is impressive considering it "only" sells for $1350. That statement may sound absurd but remember that top CIEMs these days go for at least double that.

The Cayin N5Sii, while excellent for the price, is certainly not in the conversation for best sound available from a portable device. So I switched to my desktop rig to finally get a sense of how high the Radioso can scale. What I heard ended up raising the bar for what I thought possible in an IEM.



The McIntosh/Pass Labs rig takes everything I loved about the Radioso and turns it up a few notches. The level of realism is just astounding. Any questions I had about coherency with the tribrid setup went out the window once I heard Radioso in what I consider its final form. Interestingly, the low end actually feels a touch more subdued here, as do those stunning highs. Or maybe I need to think of it from another angle and say the midrange becomes more forward and full. Whatever the case, the U-shaped signature is not as immediately noticeable in this scenario (though it never quite becomes a ruler flat signature, which I'm actually thankful for).

But talk about dynamic - this thing shakes the foundations, giving some of the hardest hitting, most convincing bass I've heard this side of a huge subwoofer. I mentioned it before but I'll reiterate - at times it really does feel like some air is being pressurized, in a way that even my favorite armature-based IEMs just can't match. And that treble? Zero grain, totally liquid and flowing, with decay and sustain and beautiful tone that just sounds like real life. The top end air reminds me of my big Stax rig - a very rare achievement for IEMs. Can you tell I love the Radioso sound on this system?



Two other things worth mentioning. First, the sense of space surrounding the music is palpable. Not only is the soundstage wide and deep, but also very precise. Usually when an IEM has a presentation this "out of head", it comes somewhat at the expense of accuracy. Sort of a 3-dimensional blob. Which is still impressive, until you try and focus on any single instrument only to find you can't really place it all that well. Either that, or the signature is just thin and lacking tonal density, which always seems to improve focus in the imaging/soundstage department (think mini-monitors in the speaker world). Radioso manages to fill out that note thickness while achieving both spacious and accurate staging.

The other aspect to point out is the fact that while these can play very loud and clean (on the right system), I actually find myself listening at lower and lower levels when I'm not critically evaluating. I take that as a good sign - I'm not trying to compensate for any perceived lack of detail or impact.

After spending time with Radioso, switching back to CIEMs that I otherwise enjoy is just a disappointment. The UE Reference Remastered sounds comparatively dark, lifeless, and slow. Empire Ears Zeus XR sounds bloated. Jomo Audio 6R sounds thin and nasally. Even the mighty 64 Audio A18t - probably my favorite CIEM overall - can seem a bit dull and uninvolving compared to the Radioso, though it does make a good comparison of sorts. If we start with the general A18t signature, crank up the bass texture and impact, then mix in the insane treble clarity of the Unique Melody Maven, we might end up with something similar to the Radioso signature. But don't forget to slash the price by more than 50 percent. Am I saying I actually prefer the Radioso sound to the A18t, regardless of price? Perhaps. I need more time to directly compare them, but at this stage I'm leaning in that direction.

I really wish I had a half dozen more DAPs here in the $500-3000 range, just to see if any of them could approach this level of performance. Obviously my big setup (approaching $20k or so) will likely still have an advantage over even the best portable players, but I suspect some of the best DAPs out there may come close enough. Whether using a portable or home system, you'll really want the best possible chain to get the most out of the Radioso.



A comparison can be made to the venerable Sennheiser HD650. Not really in terms of sound signature, but rather the fact that both models can scale ridiculously well. They are both make sound yet seem fairly underwhelming straight from a phone. Yet both respond incredibly well as your system improves. And at some point they both start outperforming more expensive flagships when paired with a really nice setup. While the Sennheisers respond more to amplification, I'd say Radioso is equally picky when it comes to DAC and even transport quality, so make sure all of those are on point to really hear them at their best.

Conclusion
I've had about a dozen requests to review CIEMs over the past year or so, and have declined them all. I didn't find a single CIEM I felt worthy of writing about in 2019. Not that I think the industry was stagnant during that time, but I felt the generally incremental sonic improvements just didn't warrant my attention. To put it another way - nothing I heard in 2019 felt like a significant jump over models I reviewed the year prior. Now it's 2020 and the AME Radioso feels important enough to grab my attention.

It may seem weird for me to rave about this $1350 CIEM as if it were some sort of uber flagship release. But honestly the pricing is just another thing AME got right with this design - it's actually kind of a sad sign of the times when most people dismiss sub-$2k releases as something less than state-of-the-art. I've heard most of the competing tribrid designs and while they all sounded enjoyable in their own ways, none really stood out to me like the Radioso does.

I admit that the mild U-shape is one of my favorite signatures, which no doubt impacts the level of enthusiasm I have for this CIEM. But I'd say it has a fairly universal appeal rather than being more of a specialty sound like the UM Maven, and I think most listeners would really enjoy it. The beautiful tuning combined with superb technical abilities add up to a very compelling product which I can easily recommend - just make sure your playback chain is up to the task.

My first experience with Yulong Audio was covering their D100 DAC way back in 2010. At the time, nobody had really heard of the brand apart from perhaps the local enthusiasts in Shenzhen. The D100 offered features and sound quality typically found on devices costing several times the price - very impressive for a small upstart audio firm.

9 years and many products later, Yulong no longer needs any introduction. While still not what I'd call "mainstream", the brand is quite well known across a wide range of forums including headphone and speaker enthusiasts. Yulong (the designer himself) continues pushing the limits without going overboard on price. His flagship releases tend to hover just over the $1k USD mark, yet like the original D100, they remain competitive with others costing significantly more.

The latest iteration of Yulong's top-dog is the DA10. As with most of its predecessors, the DA10 is an integrated DAC and headphone amp, and offers volume control for DAC-direct connection to your amplifier, all in the same compact form factor. Internal design could be described as an evolution of the DA9, which if you recall was substantially different from the prior DA8 and DA8mkII models - sort of a "reset" if you will. That 8-to-9 transition marked the switch from ESS Sabre chips to the AKM AK4497 DAC, along with a slew of other fundamental design changes. The resulting sound was sweet, clear, and nuanced, with a performance that continuously drew me in as I logged additional hours with it.


DA10 takes that same winning formula - AKM AK4497 chip, robust linear power supply with ultra-low noise regulators, fully discrete balanced Class A headphone output - and further refines it. The main change is the way signals are processed prior to the actual D/A conversion stage. This time around everything is handled via FPGA - there's no off the shelf digital receiver or reclocking chips here, but instead it is all custom programmed. The FPGA handles all SPDIF inputs, acts as a jitter-reducing FIFO buffer, and does PLL clock duty in partnership with a Crystek "Femto" clock. This pristine signal then passes to the AK4497 for conversion to analog before being sent to the analog stage which uses the same OPA1611/OPA1612/OPA1622 analog stage as the DA9. Or, route the signal to that discrete, fully balanced headphone amp which again is identical to its predecessor. Note that USB signals first come in through an updated XMOS U208 which is capable of DSD512 and PCM768kHz - both of which are doubled from the prior model. FPGA processing still applies as the XMOS chip hands off the signal.

And that's the bulk of the internal changes. Sounds simple enough, but in practice that's actually a pretty significant change. The new design puts Yulong squarely in control of the signal for a huge portion of its journey from input to output, and the proprietary FPGA process can be perfectly tailored to suite the desired result.


We also get some external changes which I find to be rather welcome. First, the enclosure is more low-key and "traditional" looking, swapping out the gold faceplate of the DA9 for a solid black or silver design (there is a red version available for those seeking a bold look). The general layout is very similar but upon close inspection, I notice the enclosure has more details and tighter tolerances compared to the (already excellent) prior model. The new design reminds me a bit of when Bel Canto upgraded to their .7 series of components. Check out the beveled edges on the faceplate of a Bel Canto DAC 2.7 compared to the older (and more expensive) DAC 3 and you'll see what I mean. It may not be obvious in the pics, but in person the DA10 looks like an upgrade.


I also like the display which uses a generously-sized wide angle IPS panel. It seems more logically laid out and easier to read at-a-glance compared to the more simplistic text-based design of the DA9, while also avoiding the (infrequent) issues some experienced with the OLED panel of the older DA8 series.

Again, none of this is drastically different, but keep in mind the DA9 was an excellent device that launched just 2 years ago. It remains competitive to this day, but it's nice to see Yulong isn't standing still. Custom FPGA solutions are seen by many as the wave of the future, and remain beyond the reach of many competing audio firms.


Gear
I inserted the DA10 into my usual test-rig. That means Equi=Core 1800 balanced power conditioner and Audio Art Power1 ePlus AC cables as a basis. Just for fun, I also swapped in Yulong's own CP1 heavy duty cable, which is surprisingly nice. Transports varied but for the most part I used either a Nativ Vita, Resonessence Labs Fluvius, ModWright Oppo 205, or a Surface Pro. For testing the built-in amp, I used Meze Empyrean, Sony MDR-Z1R, Sennheiser HD650 and HD800, Fostex TH-X00, and HiFiMAN HE1000v1. When testing the DA10 as DAC only, I paired it with external amps including the Pass Labs HPA-1, Niimbus US4+, and Cayin HA-1A mk2. I also threw in my Stax rig, which is a KGSSHV with SR-007mkII or SR-4070. Interconnects were from the Audio Art's IC-3 e series in both RCA and XLR.




Listening
I'm not going to reinvent the wheel here - everything I said about the sound of the DA9 pretty much still applies. That means clean, clear sound with a beautifully sweet tonality, particularly in the midrange, and superb treble articulation. But the DA10 builds upon that performance, with several key improvements that I believe make it a worthy successor.

Perhaps it will be easiest to explain in context of my few minor gripes about the DA9. While I found it beautiful to listen to, I also mentioned it wasn't quite the last word in tonal saturation. Meaning at times it lacked a bit of weight to the presentation. DA10 adds just a hint of richness, particularly in the midrange, which results in more solidity and "oomph". This is most noticeable with vocals, where the DA10 captures a bit more of the chest voice versus merely the head voice. Again, subtle difference, but a positive one. Note that the sweet, almost delicate nature of the midrange presentation remains in full effect despite the boost in thickness, so everything I loved about the DA9 is still applicable... this update does not transform it into a drastically different presentation.

Another little complaint I had was that DA9 didn't have the best low end impact. It was nicely articulate but lacked just a touch in authority and slam. That changes here, with DA10 sounding competitive with the best I've heard in this price class and even beyond. Thankfully it remains agile and totally free from bloat, so it doesn't stray from what made the DA9 enjoyable in the first place. Ultimately DA10 remains a generally neutral sounding DAC rather than a warm design with a prominent low end boost - I don't want there to be any confusion about the Yulong becoming something it is not.

Lastly, soundstage and imaging have improved by a significant margin, likely due to the superior processing horsepower on board. The DA10 now competes with class leaders such as Benchmark's DAC3 in providing a wide open, lifelike soundscape, whilst avoiding the somewhat steely, artificial treble tone the Benchmark sometimes displays. In this case the improvement is more significant - where the other two aspects were relatively subtle, this one is more obvious when listening with good recordings and using quality associated gear. My advance copy of the new Reference Recordings Holst: The Planets featuring Michael Stern at the helm of the Kansas City Symphony (which I very highly recommend!) is totally immersive, with a clearly defined sense of space that the DA9 - while still doing a fine job overall - doesn't quite reveal. Recordings with less spacial information obviously don't emote as clearly but the DA10 does what it can with the material I give it.

All in all, a superb performance when compared directly to the DA9 or, for that matter, most other DACs in this price class. I wouldn't call it a night and day difference over its predecessor but with careful back to back listening, it's pretty clear which one is superior. We still get the same sweet, beguiling tone along with improvements in several areas, ultimately making for a more satisfying experience which is extremely competitive for the price.


How competitive, you may ask? Well, I had the chance to give it a go head to head with the excellent PS Audio Stellar Gain Cell DAC, which goes for $1699 - that's $500 more than the DA10. Both devices make great DACs and have very nice integrated amplification. Both also drive external speaker amps directly, sans preamp, without any appreciable loss of quality that often accompanies a DAC-direct hookup.

The differences? I thought the Stellar DAC had a bit more bass bloom, combined with a tuneful midrange which could be appealing on some recordings (particularly classic rock which can be a bit thin). But the DA10 was clearly more nuanced and textured in the lower end, making it the more "accurate" sounding of the two in that respect. The accuracy continued into the more neutral, slightly sweet midrange, making it superior for vocal projection and realism. And then there's the treble - whilst the PS Audio has very nicely controlled highs, inoffensive and somewhat musical, the DA10 is far more insightful and resolving - yet still not harsh. It's in another league entirely. The Gain Cell DAC is very enjoyable and I never fell I'm missing any details until I switch to the DA10 and discover that yes, there is more to be heard.

Stellar offers plenty of analog inputs plus a remote, making it the more useful device in some circumstances. But for pure SQ alone the Yulong is superior without question, at least in the majority of system configurations I assembled. Since the Gain Cell DAC is one of my favorites in the sub-$2k price range, that's a big with for the DA10.

Without going into too much detail, I think I also like this DAC more than the RME ADI-2 DAC, the Bryston BDA-2, Chord's Qutest, and the original Auralic Vega. All are strong performers in their own ways, yet none matches the DA10 in terms of overall sonic refinement, sweetness of note, and tonal balance. Considering 3 out of those 4 sell for quite a bit more, that again tells a story about the Yulong and how competitive it is.

Transports
But general sonic enjoyment is not the only aspect worth mentioning. Perhaps the biggest change over the DA9 is just how easy it is to unleash this level of performance. The DA9 already had a custom DSP algorithm for processing all incoming data, which made it fairly undemanding of transport quality. DA10 handles everything via custom programmed FPGA, and the result is an even more low-key device that really doesn't care about transport quality in the least. It's hard to believe, but the DA10 sounds just as good with a standard Surface Pro or off-the-shelf Dell workstation as it does with excellent music servers and disc transports - that would be the SimAudio Orbiter, Resonessence Labs Fluvius, or Innuos Zen Mini mk3 with the matching linear power supply upgrade. In all cases, the result was essentially identical - or at least close enough where the strain to identify subtle differences just wasn't worth the trouble.

Favorite tweaks such as the BMC PureUSB1 active USB cable, the Wyred Recovery USB Reclocker with Wyred PS-1 linear PSU, or even the superbly capable Titans Audio Lab Helen jitter reducer/signal enhancer don't make an appreciable difference in the overall performance. Again, you can switch back and forth and try to convince yourself of some minor change, but in the end I was satisfied enough to not bother with any tweaks.

Frankly, I don't recall ever hearing a DAC quite so ambivalent to transport quality. I have heard some with euphonic coloration which tended to impart its own flavor across the board, making the transport contribution smaller than usual. This is not that. The DA10 is quite resolving and transparent overall, it just doesn't care about transport quality - an intriguing combination.

This ultimately means the end user is free to choose a player based on other aspects such as user interface or price, without regard for output quality. That's new to me, and pretty exciting.

Note - this may explain the omission of the increasingly popular I2S via HDMI input. That's typically the highest quality way to feed a DAC these days, but when transport quality is made irrelevant by proprietary means, then the superior I2S connection becomes pointless. Makes sense to me.


Amp Stage
Everything I've said above applies broadly to both the line out as well as the headphone stage. The integrated amplifier is identical to what we got in the DA9, but again, much has changed in the path leading up to that amp section. Leashing headphones directly to the DA10 allows us to pick up on those changes fairly confidently, as long as we use the right headphones.

Like most prior Yulong designs, the DA10 pairs best with lower impedance cans. 32 ohm loads see 3 full watts from the balanced output or 2 watts from the 1/4" jack. That's potent enough to drive most cans with authority, though obviously you won't maximize an Abyss or HiFiMAN HE-6 without adding a powerhouse dedicated amp. Sticking with the 4-pin XLR jack is generally the way to go, but single-ended connections don't suffer too terribly, and this is particularly true of less demanding cans. IEMs are fair game unless they are particularly sensitive (certain Empire Ears and 64 Audio models come to mind) and the digitally-controlled analog volume control is uniformly excellent.

Overall I'd happily drive my Sony MDR-Z1R, Meze Empyrean, or Fostex TX-00 directly from the DA10. Adding a modest headphone amp such as the Cavalli CTH changes the flavor a tad but in my mind is mostly a downgrade. Moving up to an Arcam rHead or Rupert Neve RNHP is a sidegrade at best - perhaps certain headphones benefit based on synergy, but as a whole the system is not elevated to any significant degree. It was only when I went up to the powerful Cayin iHA-6 that I noted a clear improvement, and then even more so when going to something like a Pass Labs HPA-1. So overall I'd say the integrated headphone stage is excellent as long as one keeps expectations in check and uses a lower impedance headphone.


Conclusion
Yulong has done it once again. Not only does the DA10 sound sweet, organic, neutral and clear, it does so with superb aesthetics and build, at a price that seems more than fair considering what some others are asking for lesser performance. Bonus points for being so accommodating with transport quality - use whatever you want, the DA10 will make it sound like a mega-transport.

I don't particularly care about the increases to DSD512 and PCM768, but I recognize that the market has something of an arms race going on in that area... and Yulong is not immune to it. On the plus side, it makes the DA10 is rather future proof, and allows it to handle all sorts Roon or HQPlayer upsampling shenanigans. So that's a good thing.

In any case, my only real complaint is the lack of remote, which is a common one in this segment for whatever reason. Beyond that, the DA10 does pretty much everything I could ask for. While not matching the tonal heft of a Metrum Pavane, or the wide-open clarity of my Resonessence Mirus Pro Signature Edition, the DA10 nonetheless does everything I could possibly ask for in a $1200 unit. Anyone looking for an all-around performer in that price range (and beyond) should give it serious consideration.

Resonessence Labs has always specialized in DACs. From the original Invicta which put them on the map back in 2011, to the diminutive Herus DAC/headphone amp, the Concero series, their superb Veritas, and the latest flagship Signature edition of the Mirus Pro - when people think Resonessence Labs they think D/A conversion.

That changes with the recent launch of the Fluvius, a streaming audio device designed with the same attention to detail as the rest of the Resonessence lineup.

Fluvius sells for $1750 and is a perfect aesthetic match for the Veritas DAC, though it also looks great stacked with pretty much any DAC (Resonessence or otherwise). Fluvius is very simple: USB output for audio, two more USBs ports for storage/accessories, Ethernet connection for networking, HDMI output for audio and/or interface, and a power port which requires 5V/4A (but more on that later). And that's about it. No analog outs and no legacy SPDIF connections. There is a microSD slot but a stern warning in the user guide tells us this is for maintenance/factory operations only - not to be used by the end user under normal circumstances.



The Fluvius enclosure is classic Resonessence Labs. That means fully CNC milled, incredibly solid, with an illuminated push-button logo that doubles as standby switch. No front panel display or other controls, as everything is handled remotely. Fluvius doesn't have integrated WiFi (but a USB WiFi adapter will work if you really need that), nor does it have internal storage apart from what is used by the OS. It can support connected USB drives and play music from there, which worked fine with my 256GB Sandisk USB stick, but the main focus seems more on streaming. My library lives on a NAS and Fluvius easily accesses all 5+TB worth of music without issue. It also supports Tidal and Qobuz for lossless streaming, which my informal surveys of the community tell me are becoming pretty widely used.







Internally, Fluvius is built around an 8-core Samsung processor along with 2GB RAM. This ends up being more than enough juice for playing any files I've thrown at it, including hi-res PCM and DSD, and it can do hardware upsampling if you want it to... with a clear warning that - more often than not - the designers find native sample rates to sound best.

On the software side, Fluvius runs on Android 7.1 with USB Audio Player Pro acting as the interface for music playback. This allows the device to sidestep any OS-based resampling and spit out a pristine native signal, including hi-res and DSD. It also allows use of streaming services - Tidal, Qobuz, Google Play Music, Shoutcast radio, and of course UPnP/DLNA from a NAS or other local source. Anyone who has used UAPP knows it is a simple yet effective solution, nice and stable thanks to years of development and updates.



UAPP uses a custom driver for hi-res audio, allowing phones (and other devices) to bypass the standard 48kHz Android limitation. This can make a significant impact on audio quality, which is why UAPP has become commonly used by many enthusiasts. Resonessence Labs takes it a step further by actually writing their own driver to improve things even more. Note how the UAPP site mentions all Samsung Exynos-based devices doing a forced resample (technically upsampling) to 192kHz regardless of input - the custom Resonessence drivers allow Fluvius to bypass that problem and maintain the integrity of the native sample rate.


In the user manual, Resonessence spends a good amount of time explaining their design philosophy. Since I don't see this manual uploaded to their website yet, I'm going to quote the relevant section here rather than try to paraphrase it myself. It's a bit lengthy and covers a range of topics, but worth the read if you like this sort of thing:



"In the modern era of digital music Resonessence products excel: the DACs such as MIRUS and VERITAS
are designed with the best of our understanding of how to maintain quality and deliver perceptively
better audio performance.

We note that many of our customers gather and maintain music collections either on dedicated servers,
cloud-based subscription services such as Tidal, or on their own cloud accounts. This means that playing
high quality music now involves streaming – moving music as a digital stream from one place to another.

Audiophiles will have noted that the means to achieve this streaming affects the music quality: a non-
optimum streaming source can never be improved upon, no matter how high performance the DAC and
amplifier that are connected at the end of the stream.

In our INVICTA products we gave our customers the opportunity to use an SD Card as the music source –
“streaming” that music from the SD Card to our internal DAC parts was handled as carefully as we know
how to do – the clock is managed optimally and even the means to access the card (the firmware that
accesses the raw bits on the card) was designed by Resonessence and so under our control to be
optimized to the listening experience.

But if you plug any of our DAC products into a third-party streamer, we are left to do the best we can
with the stream of data we are sent. Listening tests confirm that some third-party products are very
good, some could be improved upon.

Streamers are digital products – they move digital data from one place to another but where does any
part of this process degrade, in any way, the listening experience?

We find it is to do with the software in the streamer– it is not related to the precise type of USB port or
Ethernet connection used (but faster is better even when not running at the fastest speed).

Our experience is that problems, or potential degradations, relate to the marshalling of the data.
“Marshaling” as the name suggests, is to do with how the data is packetized, collected and assembled
from the input source and prepared for output at the output source.
Using any streamer product, the most convenient connectivity involves WiFI. FLUVIUS may be
connected to your WiFi router by inserting a USB to WiFI dongle. Bear in mind that any WiFi connection
demands the most intense marshalling efforts: WiFi is a remarkable set of protocol layers sitting above
an outstandingly clever digital radio substrate: that WiFI works as well as it does is a testament to the
skill of the radio and software designers.

But WiFi demands a complex effort in marshalling of data: it neither arrives continuously nor does it
arrive at a predictable time. A layer of the protocol assembles the data into order and prepares the data
to be streamed out, but this layer is in constant communication with the physical layer that is master of
the data availability. The stream may exhibit asynchronous behavior as the inherently asynchronous
radio layer is marshalled into a synchronous stream for output.
We mention all this to explain to you why, when you have listened to music streamed over a WiFi
connection, it will have been noticeably inferior to that which you can get from, for example, the SD
Card in the INVICTA.

However, we do recognize the convenience of a WiFi connected system and we document in this note
how to configure FLUVIUS in that mode. It works by plugging a third party WiFi dongle into the FLUVIUS
product and then streaming via WiFi – but for the reasons explained above, we think you will enjoy the
listening experience a little more if you use a high speed wired connection or USB source.

What works optimally in our experience is as follows: first, use the 1Ghz wired Ethernet if possible.
Ethernet, unlike USB, is galvanically isolated and you will find it to be the most reliable and artifact free
source of remote data. Secondly, use the USB3.0 ports – one to the DAC, one to the data source. The
data source can be a local hard drive (that is, you may plug a USB drive into the FLUVIUS directly, we use
the Samsung T5 series in our testing, but any USB 3.0 drive should work).

___________________

We have developed the FLUVIUS streamer product to embody all we know about audio into our first
audio product that has no analog signals in it: FLUVIUS is essentially a software product – it is an Android
based streamer with custom Resonessence software to avoid all the known pitfalls of non-audiophile
software.

Listen to our FLUVIUS product and compare it with other streamers. If you can, use one of our DACS so
that the aspects of the streamer are a little easier to perceive. We think you will be able to hear a
difference.

The FLUVIUS is just moving digital data from one port to another and data is data so where does any
perceptible audio degradation occur?

The first and easily understood aspect is ensuring that the average data rate input is the same as the
average data rate output. For reasons that all readers of this note will understand, the master must be
the DAC – the final sink of the data – not the source. This is what Asynchronous USB is all about: the
precision clock of the DAC must determine the data flow.
In a typical Android based streamer the request for an asynchronous data source made by, for example,
our VERITAS DAC, may be silently ignored. Android systems accept the notification that asynchronous
mode is requested, but in general they may not act upon it.

The reason for this is that the Android system is commonly synthesizing the audio stream data: it is
down-converting from a common format that is chosen to optimize the playback of video, not audio.
(Most cell phones have this “feature”. Only with special software can a typical Android cell phone
support asynchronous audio data mode.)

Beyond the synchronicity of the data is the variability of the timing of the data during the marshalling
process. A CPU on a chip, such as the excellent and popular Raspberry Pi, Beagle, Arduino etc. is
programmed to perform this data marshalling and the time that it takes depends upon its clock speed
and how many individual cores it has available. More cores and more speed means that well-written
software can complete the data marshalling process faster - leaving less risk of asynchronicity and never
having to drop or insert a “fake” data sample.

FLUVIUS uses the Exynos core as the computing engine and runs the Android operating system on
it. The Exynos chip is a cell phone chip: it has considerably more power than the typical simple Android
system (it is able to run the entire cell phone environment) but we use all its power in just the audio and
GUI processing. The Exynos has four cores running at high speeds and four more cores running at a
lower clock rate dedicated to managing the IO – the USB and Ethernet etc.

We do not rely upon the built-in audio drivers: we write our own drivers that do respect the DAC’s
request for Asynchronous mode, and because of the raw speed of the system, marshalling activities
complete quickly and do not create any perceptible listening degradation."


So there you have it. Various competitors place their main focus on things like power supply quality, processing power, or noise filtering (among other things), while Resonessence is all about "data marshalling" via their custom built software. It's a fairly unique approach that I don't think I have seen elsewhere, or at least haven't seen it explained in such a way. Note that the use of a WiFi adapter is possible but ranked last in terms of SQ, behind Ethernet and direct USB playack, so I didn't bother trying it.

Resonessence doesn't use a specialized USB solution from the likes of SOtM or Matrix. But perhaps they don't need to - this isn't a big PC with 400+ watt switching power supply, half a dozen PCI lanes, and hundreds of background processes running. The Fluvius is purpose built with an eye (and ear) towards maximizing playback quality, so everything should be well sorted right from the start.




SYSTEM
I tried the Fluvius in many different system configurations over the coarse of several months. Of those, I spent the most time with the following components:

Power: Equi=Core 1800 balanced power conditioner with Audio Art power1 ePlus AC cables for all components
Library Storage: Asustor AS6404T NAS with 24TB storage plus SSD cache, storing my entire 5+TB library, wired with gigabit Ethernet from the other room
DACs: Resonessence Labs Mirus Pro Signature, Yulong DA10, ModWright Oppo 205, Airist R-2R, Sonic Frontiers SFD-2 mkII
Amplification: Niimbus US4+, Cayin HA-300, Pass Labs HPA-1, custom hotrod KGSSHV
Headphones: Meze Empyrean, Sony Z1R, HiFiMAN HE1000v1, Lear BD4.2, 64 Audio A18t, Stax SR-4070 and SR-007mk2
Analog Cables: Audio Art IC-3 e for XLR and RCA connections, Audio Art HPX-1 SE for Empyrean, Moon Audio Silver Dragon for Z1R, Effect Audio custom 8-wire design for IEMs
Digital Cables: BMC PureUSB1, iFi Gemini3, Audio Art D1-SE








In Use
First off, you really do want a monitor in the system - at least during initial setup. I was eventually able to get everything configured to the point where I could remove the monitor and control things via Android tablet using the HiFi Cast app (my personal favorite among many UPnP control options) but the initial setup needs a monitor. I happen to have an ancient and fairly rare Acer device which is sort of a combination Android "tablet" (version 4.2!) plus 21" monitor using an HDMI input. The Android aspect is pretty worthless these days but the thing can actually run on battery power, and I can get a couple hours of use in monitor mode - plenty for getting everything set up.



Fluvius comes with a bundled remote that is a sort of generic design, with many variations floating around on Amazon and other places. It's got a keyboard and touchpad for mouse use, and surprisingly it works quite well. The USB dongle will take up one USB port but there are still two more available so it's no problem - just make sure to use the slower 2.0 port which is intended for just this type of thing.



Browsing the menu system with the remote is fairly simple and intuitive. If you've used USB Audio Player Pro on your phone, you'll know what to expect - just imagine mousing around instead of using a touch screen. I was satisfied with the various options for sorting music by artist, album, genre, folder, etc, and the system handled my enormous library without complaining. It's not the most flashy looking user interface in the world (definitely no Roon or even JRiver) but it gets the job done and covers all the basics for album art, sample rate display, sorting by genre/artist/album, etc.










One interesting thing is that the custom drivers allow for volume control on certain DACs. Pairing Fluvius with the Resonessence Labs DACs, and probably some others (I don't have a list though), we can adjust things in software and it will change volume on the DAC itself. That could definitely be useful depending on your control preferences.


Listening
Regardless of which DAC I used, Fluvius sounds tight, precise, accurate, and clear. Those are some loaded words, and I can see some people being turned off by what they might assume is a cold, sterile sound. But I would disagree with that assessment. In my mind, a transport should be as clean as possible, to allow maximum performance out of the connected DAC/amp/transducers.

And that's exactly what happened with the various DACs I used. Yulong's DA10 had a touch of sweetness to it, while the ModWright 205 was bold and rich, and the Mirus Pro Signature startlingly transparent - all native attributes of the DACs themselves without any sonic influence from the transport. Which is precisely what I am looking for. A transport that imparts its own signature into the mix can be fun, but ultimately this approach is the better method.

My Sonic Frontiers DAC predates USB by quite a few years, so the USB-only Fluvius won't pair with it as-is. Adding the Matrix X-SPDIF 2 gave me a usable AES signal with which to feed that vintage DAC, resulting in glorious sound that I could happily live with indefinitely - despite age and lack of hi-res playback capabilities. The SFD-2mkII is very sensitive to transport quality so to get a superb result like this speaks very highly of the Fluvius (and the Matrix for that matter).

I like to call the Fluvius a USB-only transport, but that's not strictly accurate. There's also the HDMI out which, while not commonly used by headphone aficionados, can actually be a viable connection. My mind immediately goes to NAD's unique M51 DAC, but also the numerous good-to-great quality processors and even AVR units out there which all pack HDMI inputs. There is tons of nice gear from Classe, McIntosh, Theta, Datasat, Levinson, and others which make viable DACs/preamps, and I wouldn't hesitate to use these in an audio-only rig - but of course a "home theater" type setup would be just as relevant. I did throw Fluvius in my living room system with an Anthem processor, Merrill Audio monoblocks, and Usher Mini One DMD speakers, and it performed exceptionally well via the HDMI connection. I have used quite a few music streamers/servers in that rig and I would rank Fluvius as a top-tier performer when it comes to audio quality.




Comparisons
Though music servers and streaming players are everywhere these days, the Fluvius goes in a somewhat unique direction in both hardware and software. I don't really see it competing with the Aurenders of the world - that's a different approach for a different sort of customer. That said, I did have recent experiences with three other devices that I thought might make for interesting comparisons.



Playback Designs Syrah ($6500)
The Syrah shares a similar design concept with Fluvius in that both models tap general purpose computing platforms stuffed into rather substantial enclosures. Playback Designs opts for a "carefully selected" fanless Intel NUC of unknown specifics paired with a switch-mode PSU of their own design. They don't claim any fancy customization of the NUC board but rather focus on their Windows-based software optimization. Like Resonessence Labs, Playback Designs is very focused on the timing of the data stream as it passes from one component to the next, and they similarly embrace the concept of finding adequate processing power rather than using a more potent (and therefore noisy) CPU.

While Fluvius has basic integrated playback software which can be supplemented by your choice of UPnP apps on a tablet, the Syrah leverages JRiver's Media Center software, simplified and paired with a custom app for a streamlined user experience. Syrah also handles CD ripping directly - again via tweaked 3rd party app in the form of dbPoweramp, plus a (bundled) external USB optical drive. Resonessence assumes users already have another way of ripping, or perhaps intend to bypass a local library altogether and just stream lossless music from Tidal/Qobuz.

With the massive differences in price and size between these two units, one might think the Playback Designs solution is automatically in another league. And it actually is - when it comes to being approachable. The Syrah is easier to get up and running, has a more robust user interface, and of course allows for building a library via ripping CDs directly. Fluvius assumes a certain level of technical prowess, along with a pre-existing library. If someone was approaching file-based playback for the first time, and trying to decide between these two units, I'd point them to the Syrah without question.

When it comes to pure sound quality though, I find the diminutive Fluvius to be generally superior. Obviously this will vary from system to system, or more specifically from DAC to DAC, but in most cases the Resonessence Labs solution gave a cleaner, more believable presentation, with a particularly superior sense of space and size. Imaging is terrific - listening on an HD650 won't necessarily show this all that well, but switch to HD800 or Empyrean (or speakers...) and it becomes pretty obvious that Fluvius does a better job. Treble is also very slightly more clear and refined. Drastic changes? No, both devices make excellent transports. But for extracting the absolute best from a high-end DAC, I'd choose the Fluvius - and save big money in the process.

The less refined user experience with Fluvius does not necessarily bother me. I've already got a massive collection of ripped CDs, plus a bunch of purchased music from a variety of sources. And I get a ton of use out of Qobuz and Tidal for lossless and even hi-res streaming. I'm accustomed to various apps which I can use to control Fluvius, and I know enough to get it all up and running despite some potential difficulties early on. Is it the ideal music server? Nope. But for those tech savvy enough to get it going, Fluvius offers unbelievably high sound quality for the relatively modest asking price.



Sound Science Music Vault Diamond Ultra ($5500 as tested)
This device basically takes the opposite approach from the Fluvius and Syrah. Those two both spec for adequate power while keeping a low "footprint" for minimal EMI/RFI, heat, etc, Music Vault goes for massive overkill - a very fast Intel i7 CPU and far more RAM than could ever be needed in this application. The theory is that stressing a CPU is not ideal, but the potent i7 can cruise along using a very low percentage of its power and thus have more headroom.

I'm not here to judge the merits of the different design philosophies, but I will say the utility of a powerful system can't be denied. The Music Vault system can run Roon, including extreme DSD upsampling, without issue. It has a Blu-Ray drive if you need that for some reason (which of course also rips CDs). It is loaded with internal storage - fast SSD for system use supplemented by a large spinning drive for the library, so there's no need for a NAS or other external storage.

With those advantages comes some drawbacks. For one, this rather expensive server looks like a piece of industrial equipment - and not a particularly appealing one at that. It's just an ugly enclosure in my personal opinion, unable to even remotely hide its status as a giant PC. While the Fluvius and Syrah look right at home in a rack full of tastefully designed high-end audio gear, the Music Vault product is one you'll want to hide away on the bottom shelf (or ideally behind a cabinet door) if at all possible. Also, the system is not quite silent. It does use a passive heatsink for the CPU, but there's still a case fan spinning at low RPM which I can't help but hear. An issue while listening to music, particularly via headphones? Probably not, but I'm accustomed to silent devices by now so this threw me off a bit.

In terms of the actual experience, Music Vault is great as long as Roon is used. It does support various other methods but Roon is by far the most well executed option as far as I'm concerned. Using this makes me wish Fluvius worked as a Roon Endpoint, as it really does elevate the user experience.

Sound quality for the Music Vault is a mixed bag for me. With certain DACs, and with DSD upsampling in particular, I find it arguably the best of the bunch. It is particularly superb feeding DSD512 to the Yulong Audio DA10 - a $1200 DAC which, when used in that fashion, sounds like a far more expensive device. The Exogal Comet Plus also pairs quite well just using native sample rates, as does the BMC UltraDAC.

In other cases the Music Vault comes in dead last behind the Syrah and Fluvius, sounding loose, flat, and fairly unfocused. I particularly dislike it with older DACs from Sonic Frontiers and Audio Research, but the relatively recent Keces S3 doesn't seem to like the Music Vault either. Overall it's hard to predict how it will pair, which makes things frustrating.... good matches are spectacular while poor matches would be better served sticking a $500 sMS-200 in the chain and relegating the Music Vault to server duty only. I feel like a ~$5k product should guarantee decent results across the board, and this one just doesn't in my experience.

Nativ Vita ($2399 as tested)
I recently reviewed the Nativ Vita at Darko Audio, and it makes for an interesting comparison to Fluvius. Why? Because despite the $2399 price listed above (which includes 2TB worth of solid-state drives), Vita can be had for $1799 if we skip internal storage and just use it for streaming. That's within spitting distance of the $1750 Fluvius. The "interesting" part is how different both units are from a design and usability standpoint, and how despite those differences I really enjoy each one quite a bit.

Vita is dominated by a large touch-capable HD display. It has multiple outputs in various legacy formats. It has built-in WiFi, Bluetooth, support for Apple Airplay, Google Cast, and several other streaming formats. It can run over a dozen streaming services, and is fully Roon capable. All of this stands in contrast to the Resonessence device which is USB-centric, Ethernet only, no display, no Roon, and focuses mainly on Tidal and Qobuz. Obviously two very different approaches, and at face value we would choose the more feature-rich device every time.... right?

I do love the Nativ Vita for a variety of reasons laid out in the above-linked review. But I also quite enjoy the Fluvius for its simple, direct approach, and - most importantly - for its superb sound quality. More often than not, when using the same DAC (via USB, obviously) with each device, I get slightly better performance out of the Fluvius. While the Vita sounds outstanding in every way, Fluvius gives just a touch more realism. Wider, more expansive soundstage. More vibrant tone colors. Superior imaging. The delta is admittedly small and is not always apparent depending on the other associated components... but in many cases I did indeed prefer the little Resonessence device. Of the field mentioned above, and strictly focusing on SQ alone, I would rank it Fluvius, then Vita, then Syrah, with the inconsistent Music Vault alternating between last place and competing for first. That makes the Fluvius the top dog in terms of sonic performance per dollar.


Tweaks
I mentioned earlier that the stock power supply is a 5v switch-mode design. With PSU upgrades being an obvious path for system improvement, I figured it was worth exploring... unfortunately the specs call for 4 amps which is not very common and thus disqualified any of the nice power supplies I have on hand. I ended up buying a suitable PSU on eBay, unbranded but clearly aimed towards audiophiles looking to upgrade, for roughly $300.

Long story short - despite the new PSU inspiring confidence with its beefy construction, large R-core transformer and claimed low-noise output, I could not hear one bit of difference compared to the stock unit. This could mean the Fluvius is well-sorted internally to where it doesn't need a linear supply, or it could just mean the eBay unit I purchased isn't very good. I suspect the 4A requirement is there to cover power-hungry USB attachments such as portable hard drives, which would mean the basic requirement for the solo Fluvius is a lot lower... but I didn't want to test that theory as my only other PSU capable of 5V tops out at under 2A.

I rotated through a slew of additional tweaks, hoping to find any way of bringing sonic improvements. I tried the BMC PureUSB1 active USB system - no change. I used various combinations of the iFi iPurifier3, DC iPurifier2, and iSilencer 3.0 - no improvement. I enlisted the iFi Gemini3.0 split USB cable, with the power leg being fed by a Keces Audio P8 linear power supply - still nothing.





Bottom line is this: the Resonessence Labs Fluvius is extremely well optimized and does not require any additional tweaking.


Conclusion
The Fluvius is among the very best sounding transports I have ever experienced. It competes with megabuck disc spinners, and outshines several far more costly streamer/server devices upon direct comparison. From memory, I'd put it above many others I've owned in the past as well, all of which sold for double, triple, or up to five times as much. Needless to say, Fluvius is a superb sonic performer.

As for the total user experience, that really depends on what one might be after from their transport. Those who demand Roon are obviously out of luck, and power users in general may find the UAPP interface a bit simplistic. The lack of SPDIF outputs could be problematic for some users, as could the lack of any integrated disc-ripping solution. But
Fluvius does handle all the basics of playback in a competent, easy to use fashion, so I find it difficult to fault in that respect.

Ultimately different folks are after different things in a device like this. Only you can decide what your needs are in terms of connectivity and other features. For straight-forward, no frills USB or HDMI playback with unsurpassed sound quality, the Resonessence Labs Fluvius makes for a compelling choice.

Cayin has been around for quite some time, but within the last few years have really made a push into the personal audio space. Their iDAC-6 D/A converter, iDAP-6 file-player, and iHA-6 headphone amplifier make a brilliant stack. Their HA-1A mk2 delivers sumptuous tube goodness. And their numerous portable audio players have something to offer at almost every budget.

Having offered some extremely high-value designs for the headphone crowd, Cayin recently moved upmarket to try their hand at an extremely high end desktop amp. This resulted in the massive HA-300 headphone amplifier - a two-box, direct-heated triode design selling for $3,999 and weighing in at a combined 64 pounds. I spent considerable time with the HA-300 in my system but have been struggling to find time for a proper write-up... so here we go.



First off, I recommend perusing the HA-300 product page over at the site of North American distributor Musicteck. There you will see many details about the design and execution of the HA-300. Cayin has always been refreshingly transparent about the inner workings of their products, and this one is no different. Highlights: direct-heated triode design, point to point wiring with quality silver-plated wire, high-precision 24-step custom potentiometer encased with aluminum shielding to avoid interference, and - possibly most important - the custom designed, in-house wound transformers (toroidal for the power supply, EI core for the output stage). There are very few firms out there doing their own transformers these days, and it's such a crucial aspect of the resulting sound that it's nice to know Cayin has full control of the entire process.



Although I consider HA-300 to be primarily a headphone amplifier, keep in mind it also has speaker outputs at 8 watts of Class A power per channel - perhaps not enough for inefficient full-size, multi-way speakers in larger listening environments, but probably adequate for desktop or bedroom use with most stand-mount/desktop models. Or, with the right speakers (think Zu, Tekton, certain Klipsch, etc), even larger listening rooms could work quite well depending on your listening preferences. Still, I'd say the main draw here is firmly on the headphone side.



The HA-300 is, by its nature, a single-ended amplifier. However it does provide XLR inputs and 4-pin XLR outputs for maximum compatibility. All else being equal (which isn't always the case) I found that I didn't notice a significant difference in sound regardless of using RCA or XLR inputs. Same goes for the 1/4" versus 4-pin XLR output options. That said, some of my DACs are clearly superior via one output type or the other, so it's nice to have both inputs at hand.




The HA-300 can deliver up to 5 full watts per channel to headphones. It has adjustable output impedance which I tend to prefer on the lowest setting, but again more options are always welcome. The outboard power supply unit connects to the main amplifier via massively thick umbilical cable. We're talking nearly "garden hose" status, though it remains fairly flexible and easy to work with. The tube array is a combination of 6SN7 (and variants) for input stage, 300B for driver tubes, and 22DE4 rectifier tubes in the power supply. With just a pair each of RCA and XLR inputs, and no line-level outputs to speak of, the HA-300 is a purist design which does not double as a pre-amp. Along those same lines, there is no remote control, so speaker listeners will need to walk across the room to adjust.






For stock tubes, Cayin starts us off right - NOS RCA 22DE4, Shuguang WE6SN7 gold base, and TJ Full Music 300B/n dome top. Or at least this was the combo I was given on my review unit. It actually initially shipped with a different variant of the Full Music 300B but those arrived damaged, and I was told Cayin was switching from that point forward, so hopefully my review experience represents the current state of affairs. Users will certainly try their hand at tube rolling but I stuck with the stock configuration as it is A) excellent overall, and B) more relevant to the experience of a new HA-300 user.





I did some brief searching and found that the included Shuguang tubes appear to sell for around $160/pair, whilst the TJ Fullmusic set goes for over $300 (there are some significantly more expensive variations, and some cheaper ones, but I believe this is correct for the bundled tubes). Compare that with many amps (even very expensive ones) where the stock tubes go for $10-15. Not saying price always determines quality, but it's an interesting fact worth mentioning. Lastly, the rectifier tubes sell for around $24 for a quad, and I was unable to find any expensive examples of the 22DE4 which might be a significant upgrade. So despite this being the most affordable tube in the stock configuration, it may already be hitting maximum potential. Cayin had their choice of pretty much any 22DE4 tubes, and NOS RCA is what they went with. So I think that says something.


System
Since the HA-300 lived in my setup for quite some time, I got to try it with a large array of gear. The base setup, which remained constant throughout, consisted of the following:

*Equi=Core 1800 balanced power conditioner with Audio Art power1 ePlus AC cables for all components
*Native Vita music server with 2TB solid state storage, playing local files as well as Qobuz/Tidal via Roon
*Asustor AS6404T NAS with 24TB storage plus 512GB SSD running Roon Server (which is then controlled by the Vita)
*Titans Audio Lab Helen reclocker/jitter reducer
*Bettercables Silver Serpent Anniversary Edition RCA cables
*Bettercables Blue Truth II XLR cables
*Audio Art D-1 SE BNC cables
*B.M.C. PureUSB1 active USB system



I rotated several excellent DACs through the system while evaluating the HA-300. These included the B.M.C. UltraDAC, Exogal Comet Plus with WyWires umbillical upgrade, ModWright 205, iFi Pro iDSD, Wyred4Sound 10th Anniversary Limited Edition DAC, Audio Research DAC 2, and my reference, the Resonessence Labs Mirus Signature Pro.





For headphones, the selection included AKG K812 (recabled and modded to fully balanced) and K7XX, Sennheiser 650, HD660S, and HD800 (with Effect Audio cable upgrades), HiFiMAN HE1000v1 and Susvara (Effect cables), Focal Utopia, Elear, and Elex (Moon Audio Silver Dragon cable), Audeze LCD2/3/4 (Audio Art HPX-1SE cable), and Fostex TH-X00.




Listening
As a general overview, I find the HA-300 extremely quiet if not silent like my Pass Labs HPA-1. When I crank the Cayin's volume knob to full blast I can hear the slightest indication of vague background noise. This will vary depending on tube choice and I'm sure my balanced power conditioner helps lower noise floor, but overall I'd call this top notch performance from a tube amp. In my experience, SS amps seem to have more potential for complete silence, but Cayin comes pretty close to that ideal with the HA-300.

On the flip side, I find the gain to be generally higher than I'd prefer. I admit to being a rather quiet listener compared to many folks - I've got very sensitive ears and thus play at lower volumes than most. That said, the volume on the HA-300 comes on pretty strong even from the first "click" of the 24-step attenuator, and goes up in rather large increments from there. This is not a problem at all with most planars or higher impedance cans, but it can be tough to dial in sensitive headphones for those of us seeking modest listening levels. Again, I recognize that I'm probably the outlier here, and most people will be cranking the volume much higher than me anyway... but sensitive listeners with efficient headphones need to be aware of the situation. Note that this can be mitigated using a DAC with quality volume control, which I found did not impact performance since I was only trimming a small amount. Some sources don't have any volume control so that's not a universal fix.

Obviously, with this many headphones in the mix, it's a lot to get through. So I'm going to just break it down one headphone brand at a time.



Sennheiser
The classic HD650 sounds unbelievable when driven by HA-300. By that I mean it really is hard to fathom that this thing has been around for decades, and is often derided as a "mid-fi" headphone. With the HA-300 and a suitably high-quality source, the HD650 has rich, lifelike textures, superb treble finesse, and punchy bass. I'm not sure I've ever heard it sound better. Other headphones may surpass it in one area or another but it's hard not to love the total package the HD650 delivers - when driven by a world-class amplifier. Volume here is no problem at all, with plenty of usable range and a dead silent background.

The newer HD660S is a small but pretty clear step down compared to its predecessor. It just doesn't scale as well, and the highs have some etch which isn't present on the HD650. The more sensitive design means volume comes in a bit quicker than I'd like, and there is just a touch of background noise (sort of a hissing sound) if you listen very carefully for it. Switching to a smartphone turns the tables in favor of HD660S, which keeps most of the same quality while HD650 becomes limp and lifeless. But when using a superb amp like the HA-300, you'll want to go with HD650 every time.

The HD800 pairs exceedingly well with the Cayin. Direct heated triode designs are known for their purity and tonal "rightness" and the HA-300 certainly embodies those characteristics - a good thing with the HD800 for sure. The presentation ends up being vibrant and supremely detailed while (mostly) avoiding that overly clinical HD800 feeling we have all experienced. Imaging and soundstage are world-class - perhaps the best I've ever heard. Low end response is punchy and tight, though still lacking the authority of something like the Susvara or Empyrean... just no way around that with this particular headphone. That said, it's still a thoroughly engaging presentation, and I may actually prefer HD800 here than when driven by my more expensive Niimbus US4+. That's extremely high praise indeed.

Like HD650, there are no issues with volume, and HD800 has a completely inky background.



HiFiMAN
The Susvara plus HA-300 is makes for a $10K combination. Let that sink in for a moment. Back in the day, many of us got into the headphone scene precisely to avoid speaker/amplifier combinations with five-figure price tags... so this is a weird situation to be in. Having established that, I can't deny that this is a very charming combination. Whilst my Niimbus drives the stubborn Susvara with more dynamic gusto, the HA-300 veers more towards sweetness of tone, with a beguiling midrange that defies any sort of technical evaluation. It's a tad forward, lush, and wonderfully euphonic, with a particular synergy on female vocals. Treble shows a purity, much like HD800 but more flowing and organic, whilst bass is wonderfully articulate if not as authoritative as it is with my Niimbus. Yes, the price tag is absurd, but the overall sound is very rewarding if you like a slightly more romantic presentation.

Switching to the first iteration of HiFiMAN's HE1000 which I prefer over the V2 (at least with the particular example I have), we get similarly expressive sonics with a slight shift in focus. This time it goes more "by the book", with a more neutral response overall. The one exception is the top end which has a gentle downward slope without going overboard. The presentation remains airy and extended (my listening notes say "ethereal") but at the same time can forgive a bit of harshness on poor recordings. Low end slam actually feels more prominent here than with Susvara, possibly due to the HE1000 being easier to drive. Susvara, with this particular amplifier, is all about that midrange magic while HE1000 spreads the focus around more evenly, and has a bit more midbass to create the illusion of impact. Susvara does extend lower but it's not as obvious with the HA-300 compared to Niimbus or Violectric V281. Overall an excellent combination which, in its own way, matches or arguably surpasses the Susvara but for less money.



Focal
This was probably the biggest area where my volume complaints manifested themselves. I listened with Utopia, Elear, and Elex, and all three had roughly similar issues. The lowest volume setting was already moderately loud, and the background was not as black as I'm used to with my other reference amplifiers. The former could be mitigated by using a DAC with quality attenuation - turning it down a bit on that end made certain loud recordings more palatable. The background "grunge" was never truly solved but I learned to live with it after a while. Despite my complaints, Utopia and particularly Elex made some beautiful music with the Cayin amp. Wonderfully fleshed out soundstage (particularly with Utopia) and superb instrument placement (especially with Elex) made these enjoyable despite their flaws. Treble was, once again, delicately balanced between detail and control, much like I had already heard with HD800 and Susvara. My listening notes include the terms "pristine", "fluid", and "vivid", and I meant those in the best possible way.

Elear was less successful - it already has my least favorite tuning in the Focal stable, and the HA-300 put too much focus on the already over-emphasized low end. The resulting sound was just too warm and somewhat murky for my tastes. I'm sure someone out there would love this presentation but for me it was just over the top.



AKG
One of the biggest surprises was how great the venerable AKG K7XX sounded via the vastly more expensive HA-300. I admit to not listening to these very often lately, but as I broke them out for this review I was shocked at how much they scaled. I'm used to the HD650 showing huge improvements with quality gear but I didn't realize the K7XX could scale similarly high. While perhaps not quite reaching lofty heights like the latest flagship headphones (which cost far more), this veteran AKG impresses with huge dynamics, nuanced treble, precise imaging, and rich, full bodied lower mids which in my experience is something 300B tubes do like no other. The K7XX impresses with an inky black background and there is plenty of range for volume adjustment. This is truly a superb combination.



The K812 is slightly less successful, though still enjoyable in its own way. Like the Focal models, this is a more sensitive headphone and thus suffers from a bit of background noise. It also gets moderately loud even on the lowest volume setting. Looking beyond those issues (or compensating with DAC volume adjustment) reveals a punchy, articulate, incredibly fast sounding headphone that excels with classical, jazz, and acoustic material. I was actually surprised that I could set aside the background noise and fully enjoy these genres - sparse arrangements have less tendency to hide the noise like more complex music would. Note that my K812 has been modified, replacing AKG's cheesy flexible printed wire assembly (FPWA) which carries the signal from the headband to the drivers. Very thin pure silver wiring was routed through the headband, along with a corresponding balanced cable to fully balance the connection. The FPWA was unreliable which is the initial reason for the mod, but once I heard the finished result I was blown away by how much better it sounded. The HA-300 may not technically be a perfect match but the combo still ends up sounding pretty fantastic.


Fostex
Using an original TH-900 (stock) as well as the TH-X00 (in Mahogany, upgraded with Lawton Audio level 1 mods and angle pads), I was surprised by the blackness of the background. For some reason these low impedance, easy to drive cans, which I normally prefer via solid-state amplification, pair exceedingly well with the Cayin tube amp. As with some of the other pairings, it is perhaps not the most technically accurate sound, but there's a sense of emotion or soul in there which really speaks to me. Bass thump is killer, midrange very open and beautiful, highs controlled yet extended. The TH-900 gets a little loose downstairs but I think that is a property of the headphone itself more so than the amp. The modded TH-X00 actually sounds more balanced and just overall superior to the more expensive TH-900, which is tons of fun but not always something I'm in the mood for. Again, I would have never predicted this outcome, but it really works. Volume control is not an issue here whatsoever.


Audeze
I tried LCD-2, LCD-3, and a bit of LCD-4 with the HA-300, and all made excellent pairings. I gravitated most towards my LCD-2 (non Fazor), where the HA-300 imparted a stunning sense of top end energy. Not sure I've ever heard the LCD-2 sound more spacious and open. LCD-3 (also pre-Fazor) was excellent as well, with a more neutral, linear sound and beautiful tone. Instrument placement on the LCD-3 was particularly well done, though LCD-2 was no slouch either. In both cases I consider the resulting sound to be about the best I've heard from these particular cans. I spent a little time with a first-run LCD-4 and it was jaw dropping as well - crushing dynamic swings and thoroughly engaging midrange. I've always passed on the LCD-4 but this particular pair, matched with HA-300, made me want to own it.

Unfortunately my Meze Empyrean arrived just a short time after I had already shipped out the HA-300 review unit, so I never got to pair those two. I suspect (and have heard from multiple friends) that this is a killer setup, but I can't confirm for myself. I did try various Grado, Audio Technica, and beyerdynamic headphones with generally solid results, but since I don't love these headphones or use them on a regular basis, I'm not really confident in describing the result compared to other top-level amplifiers I have on hand. I'll just say that for 90% of the headphones I tried, the result was up there among the very best performance that particular headphone has ever displayed - which means the HA-300 is doing something right.


Final Thoughts
As you can tell, the Cayin HA-300 doesn't necessarily have a consistent sonic signature that applies to every headphone it drives. In some cases the result is a bit warmer, somewhat forgiving and "fun", while other times it seems exceedingly neutral and clean. And at times it imparts a lovely midrange focus that I end up falling in love with.

I initially thought it paired best with higher-impedance cans but after trying out a dozen models (give or take) I don't really find a specific pattern. I love it with the planar magnetic Audeze and HiFiMAN designs. It is excellent with 300 ohm Sennheisers. But it also somehow loves the Fostex TH-X00, so I don't really know what to make of it. The Focals and K812 were probably the weak points but even so, I found myself mesmerized regardless of their faults. The only time I actually disliked a pairing was the HD660S, which is definitely more a function of the headphone (and its value relative to the old HD650) than the amp itself.

Without rambling on even longer, I will point out that the HA-300 is very much impacted by the source pairing. While Cayin's other tube amp, the HA-1A mk2, is rather forgiving of source, the HA-300 demands as good a DAC or player as you can possibly find. I realize people in the market for a $4K headphone amp likely have the means for a really nice source, but it's just an observation I wanted to touch on.



In the end, Cayin's HA-300 is quite possibly one of my all time favorite amps. It took me on a wild adventure, exploring various headphone pairings and the sometimes unique soundscapes possible with each. I definitely rank it up there with top caliber amps from Eddie Current, Donald North Audio, and Apex, while noting that the Cayin is significantly less expensive than top models from those firms. Not that a massive $4K amplifier could ever be considered "affordable", but it is a relative bargain in comparison. I would have a very tough time choosing if I were in the market for a reference tube amp right now.

I did not tube roll, nor did I have a matching pair of speakers on hand which could be driven by the Cayin's 8 watts per channel. So I didn't even fully explore the capabilities of this beastly amp. But from what I did experience, this thing is clearly a winner, and absolutely worth auditioning if the opportunity presents itself.

Confession time: I don't really like reviewing cables. It's a controversial topic that can often be more trouble than it's worth. And even when dealing with a crowd of cable enthusiasts, most would generally agree that cable upgrades represent a much smaller change in sound compared to swapping headphones, DACs, or amplifiers. I tend to think of cable upgrades as a "last few percent" type thing - squeezing the absolute best performance out of the system I already love. With this in mind, I actually have done the occasional cable review, such as when I first discovered Effect Audio cables way back in 2013. This is opposed to something like Moon Audio, whose cables I use and enjoy but have never reviewed - they are already well-known enough to where I don't feel I have much to add.

Today's topic is Audio Art, a brand which has been around since 2005 but wasn't very familiar to me until recently. Owner Rob Fritz is extremely responsive and easy to work with - one of a small number of folks in the industry with whom I find myself veering off topic to discuss music and other gear aside from cables, because he's just so approachable. I've become so taken with his products over the past year that my reference system is now wired with the power1 ePlus AC cables... keep in mind that I've got many, many AC cables to choose from, including some very expensive/high profile names. But I just enjoy the Audio Art power1 ePlus more, regardless of price or brand recognition. I'm also using the superb Audio Art D-1SE digital cable whenever a DAC features BNC inputs (unfortunately not all of them do). At some point I'm likely to pick up a few interconnects as well, just based on my satisfaction with the rest of their offerings.

Thankfully Audio Art now has a couple headphone cables in their lineup, which I'm told is a fairly new development. The HPX-1 Classic (starting at $300) is the "base" model, and the HPX-1SE (from $380) offers step-up performance for what I consider a reasonable upcharge. These are available in various lengths from 5 feet to 15 feet, and with a variety of termination options on either end. If you end up with some obscure application which isn't covered by the standard ordering menu, I'm fairly sure the company can handle your request via email. So don't be afraid to ask.



Under review is an HPX-1SE terminated with a rhodium Eidolic 4-pin XLR. On the headphone end, I chose the Eidolic mini-XLR plugs because they work with various headphones from Audeze, Meze, ZMF, and Kennerton, as opposed to most other types which only accommodate one specific model. I also like how robust the connection method feels - much more confidence inspiring than either style used by HiFiMAN over the years, and easier to deal with than the MrSpeakers connectors. This means I can use the HPX-1SE with my Audeze LCD-3, LCD-2, and Meze Empyrean, and not be concerned about wear and tear with frequent switching (unlike my HE1000 and Susvara, where I do in fact worry).



In describing the cable, Audio Art tells us the following:

"Our HXP-1SE features 2 paralleled cables with twin 28 awg single crystal OCC Copper conductors in each PVC jacket, insulated with PE dielectric. Left and Right channels are independently shielded with a braided silver-plated OFHC copper shield . The intention of this design is for high resolution audio performance, multiple application compatibilities, and rugged durability. Hand built in San Diego, California."

On looks, the HPX-1SE is what I'd call understated and classy. It's not the thickest cable, nor does it feature any flashy accents. But it's nonetheless attractive in its own way, and feels robustly made, with enough flexibility to be comfortable in use (unlike the stock Meze Empyrean cable which is rather unwieldy).




Although my system fluctuates regularly for review purposes, the bulk of my listening was done using the following gear.

*Core Power Technology Equi=Core 1800 balanced power conditioner
*Nativ Vita streaming Roon, where Roon Server is run on an Asustor AS6404T NAS in another room
*Titans Audio Labs Helen reclocker
*Resonessence Labs Mirus Pro Signature DAC
*Niimbus Audio US4+ headphone amplifier
*Audio Art power1 ePlus AC cables with Furutech FI-28(R) Rhodium connectors for all components
*Cabledyne Silver Reference AES/EBU cable from transport to reclocker (only certain low profile cables fit with the Vita)
*Audio Art D-1SE BNC cable from reclocker to DAC
*BetterCables Blue Truth II XLR cables from DAC to amp
*Audeze LCD-2, LCD-3, and Meze Empyrean headphones


First up was the LCD-2, which is actually my most-used Audeze headphone. This particular set is what we used to call the LCD-2.2, meaning it the last iteration before Audeze rolled out their "Fazor" technology which fundamentally changed the sound, not necessarily for the better in my opinion. I've gone through numerous sets until I arrived at this "unicorn" example, which to my ears has visceral low end impact, creamy mids, and a tastefully smooth top end which is forgiving yet detailed enough for versatile listening enjoyment.



Switching back and forth between the stock Audeze "ribbon" style cable and the HPX-1SE, I notice a subtle but worthwhile improvement in treble clarity. Not necessarily more extension, as the tonal balance is unchanged. Yet I hear cymbals, triangles, and particularly trumpets come through with more realistic attack as well as more believable note decay. For example, on the excellent XRCD release of Tiger Okoshi's Color of Soil, I feel the "bite" of the trumpet more distinctly than with the stock cable, while at the same time finding it less fatiguing at higher volumes. When the sound comes through more clearly, it can be more dynamic and alive while simultaneously feeling less artificial, and thus not triggering the part of the brain that says "turn that down".

Midrange remains untouched by the HPX-1SE - in this case an absolutely perfect outcome. This particular LCD-2 is rich, creamy, almost saturated through the mids, and I would not want to change a single aspect of it.

Low-end extension and body seem very slightly improved. I notice a touch more tonal richness from the double bass on an SACD rip of Gary Karr Plays Bach, along with a similar feeling of increased authority on Yo-Yo Ma Plays Ennio Morricone. Again, these improvements are subtle, and by no means does the Audio Art cable change the fundamental character of the headphone. Yet between this low-end improvement and the superior treble clarity mentioned above, the result is a welcome fine tuning to an already exceptional headphone.



A reasonable argument might be to suggest the money would be better allocated to a higher-end Audeze, rather than using the LCD-2 with the added cost of a cable upgrade. Makes sense on paper, but I have already been through at least one example of every model in the LCD series. In fact I still own a very nice sounding LCD-3 (pre Fazor) that I settled on after going through about 4 different pairs. I quite enjoy it, and admit it is clearly the more neutral, resolving headphone compared to my LCD-2. It would be far more useful in evaluating gear, but I already have other headphones that perform that function even better. In terms of listening pleasure though, my LCD-2 is king.

Still, swapping in the HPX-1SE brings a more substantial improvement to my LCD-3. It sounds more liquid, more flowing than with the stock ribbon cable, while again bringing more clarity to the treble... which in this case feels more significant as the treble is more prominent with this headphone. While my preference remains for the LCD-2/HPX-1SE combo, I can't deny that the recabled LCD-3 sounds pretty amazing. The delta between them becomes smaller, and it really comes down to signature preference... I prefer the smooth, rich LCD-2 but could really make a case for the 3 as well.



Lastly, the Meze Empyrean, which has become my most used full-size headphone lately. I already dislike the stock cable for its stiffness so the Audio Art gets points on feel alone. As for sound, I didn't initially notice a huge difference... sure, treble was more clear, but it wasn't as significant as the Audeze models. After using the HPX-1SE for a week, I switched back to the stock cable (don't ask me why)... and immediately noticed a muffled, compressed sound signature. The difference was striking in this case - definitely on par with the change I heard with the LCD-3, or perhaps even greater. This change was nearly on par with switching between the two sets of bundled earpads that Meze includes with this headphone - a definite, clear change in signature.

The Audio Art cable opens up the top-end of the Empyrean in a very welcome way. It's not brighter per se, but more clear, open, and resolving. I'd call it treble "rightness" though I'm not sure that will make sense to all readers. The ample bass response of the Empyrean needs no further boosting and while the HPX-1SE does increase the sense of slam by a small amount, it seems to bring focus to the sub-bass region. So it's not initially obvious, nor does it significantly change the tonal balance of the headphone. The end result is a more open, clean sounding presentation with improved resolution, along with slightly more sub-bass impact if you really listen hard for it. Which for my taste ends up being pretty much ideal.



In the end, all three headphones benefit by a noticeable margin from adding the Audio Art HPX-1SE. Which makes it a worthy upgrade as-is. But I'd like to discuss one more thing - Audio Art makes an adapter cable system which allows the HPX-1 or HPX-1SE to be used with different headphones. I was able to use the adapters to make my cable compatible with both the Sony Z1R and the Focal Elex, while still being very low profile and comfortable.



Sony's Z1R is a weird headphone. It's certainly not neutral, and the experience brings to mind the boring-looking-yet-expensive Audio Note speakers - tactile, musical, but far from even-handed. Still, with the right supporting gear and the proper expectations they can be wonderfully engaging and fun to listen to. The Audio Art cable seems to make the Z1R a bit more linear and "correct" sounding, which to my ears takes away from the experience. I end up preferring my Moon Audio Silver Dragon which maintains the core signature while tightening everything up a bit.

Having said that, I can see how the HPX-1SE might be preferred by folks who don't necessarily love Sony's tuning on this headphone. It does end up sounding more neutral and controlled, which is something I've heard people wish for on more than one occasion. For my uses though, it's not an ideal match.



Focal's Elex, on the other hand, is a brilliant partner for the HPX-1SE. Again I hear improved low-end authority along with superior treble clarity, but this time midrange is a bit more fleshed out as well. Not enough to change the fundamental character of the Elex but it does make them feel slightly more tonally rich. I also notice improved staging and imaging accuracy. Elex is probably my favorite headphone in the sub-$1K space, and while the HPX-1SE brings it slightly above that number ($~1100 for the combo) I feel the result is definitely competitive with other headphones costing more (including Focal's own $1500 Clear).


Are cable upgrades worth it in all cases? Of course not. Your system has to be resolving enough to pick up on minor changes, and if it isn't then I'd recommend other upgrades first. Once you have everything else dialed in, quality cables can bring that exceptional system to an even higher level of performance.



Audio Art's HPX-1SE headphone cable is a very worthy entry into a crowded field of competitors. It did a superb job with 4 our of the 5 headphones I paired it with, and the 5th headphone was debatable based on listening preferences.

I also love the fact that Audio Art offers the adapter system, allowing users to get more bang for their cable buck. I've seen a few other brands offer conceptually similar adapters but the Audio Art system strikes a perfect balance between low profile and price - others I've seen are either too bulky or cost nearly as much as the entire cable.

Overall I can very easily recommend the Audio Art HPX-1SE. It's not as well-known as some of the big brands, but in my experience it offers better value and performance than many others in this price range. Definitely check it out if you are in the market for an upgrade.