Hugo2 - A Desktop DAC Mini Marvel
Daniel Mance




INTRODUCTION

Every marketing professional knows that the best way to influence a buying decision is through word of mouth – so Chord’s use of ordinary Head-Fi posters to review a product is a stroke of genius. These reviews are grass-roots opinions from community members who want to be honest, technically accurate and write an enjoyable read. And we’re doing so without monetary or in-kind compensation.

My review of Hugo2 is to focus on its use as a desktop DAC driving a loudspeaker system. I will not be commenting on headphones at all. Nor will I be discussing mobility or nice leather cases – nor emphasizing features. I will focus on sound quality in absolute terms – not coached by ‘value for money’ or ‘based on portability’. I want the best the H2 can deliver so I will be working to extract the best sound quality I can by reconfiguring and tweaking my system. And then I’ll do some comparisons against other desktop DACs.

Thanks to the many posts of Rob Watts and HeadFi’ers, I have come to appreciate what ‘transparency’ can mean, what to listen for, what to avoid and to wean myself off false euphonics. Loudspeaker setups, generally speaking, are more complex than headphones, much more expensive and require more effort to better. At some level, you reach a plateau. Yes …some loudspeakers are more revealing in detail or tighter bass extension and some amplifiers can improve musicality or low end and, of course, cables and power matter. However, I hope I can say my review system was as resolving as I could get it within a reasonable budget – leaving an objective reference to report only on the impact of the H2 and comparison DACs.

These are the system components used:



SETUP

A high-end DAC deserves a place of its own on the top shelf of an audio rack but given the H2’s diminutive size it looked lonely by itself so it shared space with an objet d’art.


Hugo2 and Friend

I was not enamored with the H2’s disco lights display but once you get used to the color scheme it’s a distinctive part of the product experience …and it produced quite nice ambient lighting. I faced the row of H2’s lights toward my seating position and controlled volume and filter settings from the remote.

Getting things running properly was not a problem. Drivers loaded on my Windows10 laptop, the DAC was detected and my smoke-test of a low-volume playback all worked fine. I am pretty determined to ensure noise-free/bit-perfect USB transport of music to my DACs and although I understand why people buy dedicated streamers and fuss with digital cables I’ve had great success producing phenomenal sound with ordinary windows running optimized Process Lasso and JRiver. I know that my USB cleanup (MP-U1 + Intona) isolates and removes most any vestige of source analog noise – and this was used throughout the test.



USB to MP-U1 to Intona to Hugo2


My H2 was directly hooked up to my ZOTL40 MkII amp and to my Voxativ AF-2.6 DIY, Finale Audio Vivace Mini or Coherent GR12 speakers. I used single driver speakers in this test to improve imaging and reduce (or eliminate) any cross-over artifacts. My JRiver playback volume was set to ‘disabled’ and I let everything up to PCM 768khz sampling and DSD512 pass directly. The H2’s chassis RCA openings were too small to allow me to secure the connector locks on my Analysis-Plus interconnects – so I undid them. I did not use a subwoofer but if you wanted to its easy to connect the parallel 3.5mm jack to an RCA splitter cable.



RCA openings still not large enough for my interconnects


I prepared my curated playlist of excellent audiophile tracks obtained from both legitimate and dubious sources. I used about 25 tracks for this review – a good mix of Redbook, Hi-res and DSD, including a handful of DSD256 and several tracks in various lossless and lossy formats. When I am critically evaluating for sound quality, I look for specific points in a track that highlight the differences I hear and A/B those – over and over. I also prefer live in-studio or in-venue or in-concert performances over electronic or over-produced content. A partial list of my tracks is provided at the end of this review.



Finale Vivace Mini, Coherent GR-12 and Voxativ DIY


IMPROVING SOUND QUALITY

My first order of business was to maximize the H2’s sound quality and try out the ideas and comments raised in several Head-Fi posts.

Burn-In
Initially, I felt the H2 sounded very bright through my speaker setup - too bright. I was almost ready to return the unit …but after a few days of regular usage it just settled down. Was this burn-in? I was the last of the tour reviewers and the unit most certainly had many hours on it. Was it the connection to my power amp that removed the last vestiges of cobwebs in the output stage? Unsure …but the transformation was real and not a situation of ‘brain burn-in’. The H2 was screechy for a few days then became sublime. So, buyer beware, a longer home audition may be needed for your H2 to become best friends with your speaker setup.

Filters and EQ
Maybe I’m sensitive to these things but I found the H2’s crazy detail to be almost the detriment of a pleasing presentation and dream weaver listenability. I expected the filter settings to dramatically change things but I found it very hard to hear any differences. Every setting still had huge amounts of high frequency, although it seemed like Red sounded the smoothest. With H2 you’re paying for taps and I didn’t want to lose out on the 256fs up-sampling so, in the end, I went with Green and dialed in a touch of JRiver EQ high shelf to tame down my Voxativs. This setting worked very well for all music genre’s and all bitrates.

I noted that H2 revealed so much new information in my reference tracks that they sounded like different recordings. It’s like the H2’s FPGA snuck a few more singers and performers in the back door. I was expecting ‘more of the same only better’ but instead I got doses of unfamiliar depth, hidden voices and instrument nuance – all beautiful. H2 really was transformative to the music and reaffirmed how important accurate sources are when determining what other aspects of your system may need attention.

H2 thru Pre-Amp – Yes or No
Audio pre-amps buffer input/output impedance and provide gain with a volume control. The H2 incorporates a digital volume control more accurately than any pre-amp and has a well-designed output section to allow direct input to a power amplifier. So, easy-peasy, no need for a pre-amp with H2, right? Well yes and no. If you really need a pre-amp to support phono or analog inputs, be aware that I found the H2’s ‘line-level’ setting of 3V RMS to be too high for my microZOTL. I obtained a much more pleasing sound when I dialed my pre’s volume gain to near maximum and set H2’s sound volume to below 2V. I suppose this is something to do with the interplay of DAC/Amp gain/impedance curves. Direct to my ZOTL40 amp the H2 sounded better as confirmed with several back-and-forth tests. I was after transparency so I ran the rest of my review with DAC direct to amp.



LTA microZOTL Pre-Amp



LTA ZOTL40 MkII Amplifier


USB Input – Conditioning Needed?
The H2 comes with a factory USB cable – nothing special – but it does work quite well directly connected to a computer source. Rob Watts gallant attempt with increased filtering in lieu of proper galvanic isolation has really helped. I compared this with my external isolation and listened intently for distortion signatures I know are due to analog noise intrusion to the DAC. I was struggling with confirmation bias but I will say that my external USB isolation may have brought forth a tad more clarity, snap and a smoother sound. So, I’d suggest that after buying an H2, go shell out for your trusted USB isolation solution …just to be sure.

Power & Battery
The importance of clean power feeding a DAC cannot be overstated – as any negative impact on the small signal D/A processing becomes quite audible when amplified. H2’s battery operation definitely gives it a huge advantage here – obviously making it immune to AC noise and ground loops. I got 6-7 hours battery operation and while charging I could discern no difference in the sound. The brilliant engineering by Rob Watts really works to isolate the delicate bits of the DAC from the terrible world of household power. This feature alone has me rethinking the value and importance of floating my entire source chain off the grid: laptop, USB cleanup and DAC. Then I can be assured that my RCA interconnects carry the DAC’s absolute best quality analog signal to my amplifier. Thank-you Hugo2.

My other review DACs all relied on AC power and definitely needed power conditioning to sound best. To offset the H2’s battery-only advantage over other DACs, I tried using a UPS (APC SmartUPS 1500) to provide a maximum of about 90 minutes of power to each of them. This may have helped - although the H2 still embarrassed the lot by dominating in sound quality. More below…

I was fortunate to have demo Nordost QKore ground units during my review. These work to silence spurious noise on the AC mains as well as component signal ground – like sticking a proverbial thumb on the buzzy spot. The H2, being battery powered, was not bettered by improved signal grounding. Other AC powered DAC’s, however, did marginally improve (deeper silence I felt) and were thus configured during my comparisons. Mostly, QKore delivered the best benefit to my AC powered ZOTL40 amplifier by grounding the RCA inputs (and by extension, the DAC’s RCA outputs and analog section).



Nordost Power Conditioning



QKore Ground Units


DSD vs PCM
Rob Watts has made his opinion on DSD well known – and emphasized his PCM-first approach for his DACs. However, somewhere during the development of the H2, the decision was made to give DSD content the same high regard as PCM. Exact details are unclear but I understand it’s a parallel implementation to his PCM approach and the results are very good. DSD playback (64x,128x,256x) plays flawlessly and with the same fidelity as PCM frankly. I looped several tracks for which I have both versions and loved them equally. I’d probably need better ears or a more resolving system to discern the differences.

Lossless vs Redbook vs Hi-Res
Any song played from my Google Play Music subscription (at an “effective” 320kbps resolution) was almost as satisfying as the lossless 16/44.1 version from JRiver or Tidal. Using the H2 to stream background music while trying to answer emails or preparing dinner proved highly distracting to the task at hand. Even at muted volume levels, the bass has wonderful impact and even whisper vocals carried great emotion. High resolution lossless versions sounded impressively more smooth and real …but Rob Watts WTA filters really work well across the source quality spectrum. Still, in quiet, intimate listening sessions (like after 1am with a glass of bourbon in hand), there is no comparison to the joy of 96khz music.

Direct to Loudspeakers
It has been noted that a Chord DAVE can directly drive high efficiency speakers. DAVE provides about 2 Watts of power into 8Ω to deliver the ultimate in transparency – or so go the comments from those few souls who have this kind of setup. I felt I needed to try it with H2 and even though it’s power output into 8Ω is only about 1 Watt it would be enough to drive my ~99dB efficient Voxativ speakers to listenable volume levels. To connect the RCA plugs on H2 to my speaker cable, I used adapters purchased from Amazon - although directly wiring an RCA end on the cable would probably reduce losses.



Speaker wire connected directly to Hugo2’s RCA outputs



UXCELL BNC to Dual Banana Binding Adapter with matching RCA connectors

And the sound is …Holy Transparency Batman! I thought my ZOTL amp was the final word but the H2 directly controlling Voxativ neodymium drivers sounded so much more realistic with authentic echoes from a deep soundstage that transported me to the recording venue. I heard visceral lows, incredibly detailed voices; a beautiful mid-range and effortless highs that just floated off the speakers. The nearly distortion-free linearity of the H2 amp was quite apparent – I noticed a purity in the mids that I found very enjoyable. This was the sound of headphones through loudspeakers. I was transfixed and will join the chorus of persons who say that this approaches the ultimate. For me, it’s the best I have ever heard.

That being said, I did notice a certain thinness in the delivery: less authoritative drum slam and slightly diminished body in general. Also, the digital nature of the H2 was laid bare without any interconnects or tubes to mellow the sound. So, it was gloriously transparent but at the same time …uncomfortably so. I returned my ZOTL to the mix and got back the ‘meat on the bones’ – as they say - but immediately sensed the loss of definition. Hmm. I wanted to go back to H2’s amp but its battery-based power supply just couldn’t provide the power (current) needed to move loudspeaker magnets with aplomb. Oh boy, what to suggest? Just add a subwoofer and call it a day? Upgrade to DAVE? Wait for Chord’s anticipated Digital Amps? For sure, once you have heard this kind of full transparency, there is no going back.

Software Up-Sampling
The H2’s ability to accept PCM768 and DSD512 means it can handle software up-sampled music and only the final conversion to analog is done by the H2’s pulse array hardware. The mediocre SoX up-sampler within JRiver easily ran on my laptop but produced vastly inferior audio. HQPlayer is a well-known media player with a very powerful up-sampling engine – arguably the best in the world. The most recent version includes ‘xtr’ filter options to mimic the ‘Chord Sound’ - says designer Jussi Laako. Despite Chord’s claims to the contrary, modern CPU’s are well capable of filtering at high tap counts and HQPlayer’s can run into the millions. Does this mean we can duplicate WTA filters and turn the H2 into a DAVE, or better? My laptop was not the most performant so I limited my tests to the PCM filters ‘poly-sync- xtr-mp’ with standard noise shaping – and these ran in real-time with no problem.

The results were interesting. I’d say that HQPlayer produced files almost as smooth as DAVE and there was a bit more lushness to everything. I found the instruments sounded more natural and I believe I also heard a deeper and wider soundstage. But overall, the HQPlayer up-sampling was not as magical as the H2 on its own. This certainly warrants more experimentation. So, you can consider H2 to be both a fantastic up-sampling DAC voiced by Rob Watts or an excellent D/A output stage that can accept software up-sampled PCM or DSD.


DAC COMPARISONS

There is nothing double-blind or independently-verified about the comparison below. This is not a shoot-out but rather my considered opinions being as objective as I can.

My first comparison was with the AudioQuest Dragonfly Red. Direct to amplifier the DFR sounded great and is a phenomenal implementation of the Sabre architecture. Say what you will about this el-cheapo, mass-market device, I’d rate the DFR as being about 70% of the H2 for sheer enjoyment. Although the DFR approaches the H2’s musicality, the H2 just dominates it with more and better details and convinces you of a much wider and deeper soundstage. H2 accurately portrays singers, instruments and layers of performers whereas the DFR was muted in comparison. The H2 just sounded more effortless and real. I don’t think I experienced listening fatigue with the DFR but the H2 certainly provided more pleasure and many more ‘suspension of disbelief’ moments.

Next came the Antelope Pure-2. This is a professional grade Burr-Brown R2R DAC with the claim of accurate clocking by heating its clock crystal in a tiny on-board oven. The result was a very good sounding presentation with excellent details and timbre. However, on a busy choral or an orchestral flourish the Pure-2 was not as convincing as the H2 which just stood the music up as the real thing. The Pure-2 is a noble attempt to push the limits of classical audio engineering however I just did not want to listen to it after hearing the H2. Rob Watts has expounded on the challenges of R2R designs – that they have terrible noise-floor modulation and are very prone to jitter. These are words on paper but you can hear for yourself how Rob Watts design prowess allows the diminutive H2 to do more with less. Way more.

PSAudio’s DirectStream Jr is an excellent desktop DAC – made more so with regular (and free) updates to their core FPGA - most recently a ‘Huron’ release – where Ted Smith (DSJ designer) has found, like Rob Watts, that he gets better transparency by removing any and all forms of signal correlated noise and jitter. The DSJ is very close to the H2 and after nearly two hours of back and forth listening it was basically a tie. The DSJ matched the H2 from the lows to the highs and through the details. I’ll raise two comments: (i) the DSJ presented itself as being every-so-slightly harsher with voices and wind instruments and was not as smoothly delineating of layers - whereas H2 perhaps had less low bass. And (ii), my sensitive speakers made the low level digital noise inherent in Ted’s design quite obvious. This is a known and natural side-effect of DSD DACs and perhaps made worse by my ZOTL’s lack of input coupling capacitor. I tried the DSJ gain adjustment but could not get rid of the hiss, hum and swishy sound I heard with my ear to the driver – even at zero volume with no inputs. It took a pre-amp and playing with the combined volume/gain controls to reduce the noise to a manageable level. In contrast, the H2 was absolutely quiet.

The Copland DAC 215 is an well engineered Sabre 9018 implementation with a tube-based headphone amplifier. This is a very nice DAC and the tubes added some nice euphonics to the presentation. The Sabre32 architecture in the 215 is likely as good as it can be made to sound and I’d say that it matched the H2’s FPGA design for extracting detail and staging information. Where it falls down is in just not matching the musicality of the H2. The 215 certainly made a piano solo sound beautiful – however the H2 rendered a real piano. Its uncanny how the H2’s superior timing is able to find subtle nuances in the notes and draw you into the instrument and performer. I would love the 215 if I had not heard the H2.

The Chord Dave was launched in 2015 as a revelation. Rob Watts pioneering work on WTA and noise shaping demonstrated that our ear/brain is capable of incredible discernment. All his knowledge was implemented in the expensive DAVE and I hoped H2, although a significant cost reduced version, had all the best bits. I did not have a DAVE for A/B listening tests but in my recent demo I played the same material through the same source and USB chain. The result? The H2 has many of the same traits but comes up frustratingly short of being in the same league. DAVE is so much smoother, richer and presents a much more expansive soundstage. And perhaps I should add the word ‘addictive’. Occasionally, on certain passages of music, I felt that the H2 might have matched what I heard on the DAVE …but overall, sorry, no, the H2 isn’t a DAVE. In defense of H2, I will say that the DAVE system I heard was probably twenty times the dollar value.


CONCLUSION

I don’t think audiophiles realize how amazing an achievement the Hugo2 is. It’s a marvel of digital audio engineering in miniature. A transportable cousin to the fantastic DAVE that delivers industry leading sound quality and reveals so much promise for the future of affordable high-end audio. As a desktop DAC replacement, Hugo2 presents a quandary for the interested buyer. Its not a traditional full-size ‘hand-built’ DAC with oversize capacitors and a toroid power supply. It’s a hyper-oversampling DAC in a world where purist NOS seems to be gaining favor. The Hugo2 is impossibly small and its simple glowing balls user interface an affront to the norm. Yet it’s a gloriously good sounding desktop DAC that squashes all and that should alarm the competition – who perhaps will never be able to attain the same levels of performance.

The Hugo2 is not without faults. Even after two weeks of use I feel it’s still a touch bright and on my high efficiency resolving speakers the Hugo2 needed EQ and filter help to sound best. I know the DAVE has wonderfully natural highs and so did my previous 2Qute. And the Hugo2 can sound much less bright when its fed up-sampled content so I think there is a compromise in the design – perhaps to accommodate headphone characteristics. Also, the filter settings really don’t sound different enough – although, yes, the Red was somewhat less bright than the White. Perhaps this is another situation where a headphone electro-acoustical environment reveals more differences.

I like listening to music through headphones but I love the ‘in-room’ loudspeaker experience. Its always seemed that the gulf between the two was characterized by huge differences in space requirements, costs and complications burdening the guy like me. My experiment with software up-sampling an Hugo2 driving loudspeakers directly to sonic nirvana gives me high hopes that my personal system can be simple and sound as good as the very best for a small fraction of the cost. The future of music is streaming and with a product like Hugo2 plus the imminent companion streamer then we’ve nearly reached a point of end game – for the industry and consumer.

-- Daniel Mance