[Dannemand]

Thanks @Dannemand, your great advice on here helped sway my mind to pull the trigger. The sound improvement over the Topping D50 is more than I had envisaged. Had a quick play with the filters and they actually change the sound.

I am very glad to hear that. I continue to be amazed with how well this DAC sounds. I always said about my V30 that it reveals poor recordings. And that is still true, it is very analytical. But the A18 is just incredibly musical to my ears (which admittedly aren't what they used to be) and it has a way of effortlessly making everything sound great -- or at least as great as it can given the material.

My only gripe is the small display and Gustard's refusal to show volume in large letters where it can be seen from a distance. They are very adamant that you should never use that volume adjustment, but instead play everything at 0dB through a preamp or headphone amp (preferably a Gustard, of course). Of course they are right that the DAC performs better at max volume (or rather at -3dB according to ASR measurements). But unless you have an excellent passive preamp, I still think you'll get better sound quality going directly from A18 to a power amp -- as long as you can keep the A18 volume adjustment to a minimum, say no more than -10db.

Gustard pointed out in the X16 thread on ASR that there is no preamp/attenuation in the output stage. I imagine it is true for A18 as well. Volume adjustment is performed entirely in the DAC chip. Digitally of course, hence their recommendation to avoid it. OTOH I prefer that to an added digital preamp/attenuation stage which in DACs is usually of mediocre quality.

But A18 outputs a very hot 6.3Vrms through XLR, so the amp must have gain controls. Otherwise you'd have to run the A18 at -30dB (or something). And that's not good for sure!

I am including a quote from my own ASR post (which I linked earlier) containing suggested test material:

Nice test tracks there. Great imaging, particular tracks 2-8. I made a playlist with them:

https://tidal.com/browse/playlist/06394610-9906-45f6-b271-8a008700994f

For excellent imaging tests, I recommend downloading the free David Elias MQA Track Sampler. They are not my music style at all, and yes it's MQA, but that is besides the point. They are just very well made.

Of course there is also the good old 2L Test Bench with some excellent recordings in different genres available for free download in a number of formats from RedBook to DXD (352.8KHz) and MQA and DSD. Great for comparing the formats and sample rates. They are all recorded in DXD. The classical ones tend to be mastered a little hot, but still excellent.

2L also has these tracks (and others) in Tidal playlists and albums (some links below) but those are all MQA which I know some will want to avoid. The Test Bench web page has whichever format you want.

https://tidal.com/browse/playlist/3f6a06c7-0dfc-4170-bac3-dce62aa6c42e
https://tidal.com/browse/album/2400318
https://tidal.com/browse/album/59883218

For excellent recordings in general, I like almost everything from Reference Recordings (Keith Johnson & Co). I can share a list of Tidal links if anybody is interested. No MQA, all RedBook

 

[Taz777]

With respect to the filters on the AK4499 chip, here is a mapping from the A18 display to the AK4499 data sheet:

Source: https://media.digikey.com/pdf/Data Sheets/AKM Semiconductor Inc. PDFs/AK4499_Feb2019.pdf

Format is A18 display = AK4499 Data Sheet

32-bit 8x Digital Filter
SHORT SHARP = Short Delay Sharp Roll-off, GD = 6.0/fs
SHORT SLOW - Short Delay Slow Roll-off, GD = 5.0/fs
SHARP = Sharp Roll-off
SLOW = Slow Roll-off
SUPER SLOW = Super Slow Roll-off
LOW-d SHOT = Low Dispersion Short Delay Filter

The default for the AK449 chip is the Short Delay Sharp Roll-off filter. I'm not sure what the GD represents.

 

[Dannemand]

I made a list of SMSL M200 filters vs Gustard A18 filters when I was comparing the two (they are both AKM, but number the filters differently).

As part of that I found the following blog post from AKM's Audio Meister Tomonori Sato. It touched on a geographic/linguistic reason why people prefer different filters.

https://velvetsound.akm.com/global/en/stories/meister/meister-tutorial01/

 

[Taz777]

My only gripe is the small display and Gustard's refusal to show volume in large letters where it can be seen from a distance. They are very adamant that you should never use that volume adjustment, but instead play everything at 0dB through a preamp or headphone amp (preferably a Gustard, of course). Of course they are right that the DAC performs better at max volume (or rather at -3dB according to ASR measurements). But unless you have an excellent passive preamp, I still think you'll get better sound quality going directly from A18 to a power amp -- as long as you can keep the A18 volume adjustment to a minimum, say no lower than -10db.

Gustard pointed out in the X16 thread on ASR that there is no preamp/attenuation in the output stage. I imagine it is true for A18 as well. Volume adjustment is performed entirely in the DAC chip. Digitally of course, hence their recommendation to avoid it. OTOH I prefer that to an added digital preamp/attenuation stage which in DACs are usually only mediocre quality.

But A18 outputs a very hot 6.3Vrms through XLR, so the amp must have gain controls. Otherwise you'd have to run the A18 at -30dB (or something). And that's not good for sure!

Good info!

Getting the A18 was a no-brainer as the AK4499 is a magical DAC chip and I'm familiar with the sound based on my D90. Another D90 was out of my price range as both versions of the D90 (MQA and regular) have rocketed in price recently.

Currently the A18 feeds Ruark Audio MR1 MkII active speakers on my desk in my study. It's set to 0.0dB and the speakers are not clipping (they did with my D50 at 0.0dB volume). I just checked the specs for the Arcam rHead headphone amp and XLR can take up to 8V! Hopefully it'll be fine. I think the D90 that's connected to my first rHead outputs 4V on XLR and the rHead sounds very good with that. My second rHead (for the A18) arrives in a couple of days.

I may sign up on ASR soon just for the techie discussions. Is there an A18 thread on there too?

 

[Dannemand]

Currently the A18 feeds Ruark Audio MR1 MkII active speakers on my desk in my study. It's set to 0.0dB and the speakers are not clipping (they did with my D50 at 0.0dB volume). I just checked the specs for the Arcam rHead headphone amp and XLR can take up to 8V! Hopefully it'll be fine. I think the D90 that's connected to my first rHead outputs 4V on XLR and the rHead sounds very good with that. My second rHead (for the A18) arrives in a couple of days.

I may sign up on ASR soon just for the techie discussions. Is there an A18 thread on there too?

I think negative 2-3dB is the sweet spot for A18. The official measurements posted on ShenzhenAudio were made at 4.9Vrms (see below) and max output was measured on ASR to be 6.3Vrms.

There are several A18 threads on ASR, although they are not very active as everybody is busy in the X16 thread, which is the latest measuremaster DAC.

Gustard A18 Balanced Desktop DAC Review (Amir measurements)
Wolf's Measurements of Gustard A18 DAC
Gustard A18 (older announcement thread)

Note that all measurements on ASR were performed on the original A18. We know that at least the XMOS controller was updated to XU216 in the 2nd Edition and Toslink added. Who knows what else was changed.

If you do sign up there, then be careful not to post listening impressions, particularly not ones comparing DACs. You're likely to be booed if you do.

 

[Gravitas McBB]

@crabdog - Thanks for the encouragement! Yes, I can wait.

It's not a critical upgrade for me at the moment. I've also got my eye on another headphone amp on an auction site that accepts balanced XLR inputs which I will be able to use with the A18. Currently, my desktop system is speakers only.

It will be interesting to see how the A18 stacks up against my non-MQA D90. I think it will be very similar in sound.

This would be a killer compassion. I have been waiting for this for awhile now. Eager to hear as well once you post.

 

[Taz777]

Note that all measurements on ASR were performed on the original A18. We know that at least the XMOS controller was updated to XU216 in the 2nd Edition and Toslink added. Who knows what else was changed.

If you do sign up there, then be careful not to post listening impressions, particularly not ones comparing DACs. You're likely to be booed if you do.

Ah. I'm ultimately an 'ears' person! I aim to get the very best out of audio equipment at any price point (in the past I've returned/sold more expensive items as they didn't sound good to my ears). I tend to end up with my 'end game' systems through a process of experimentation, rather than a graph of a component in isolation. Having said that, I do have an interest in the technical side of things.

Specifically with the A18, after experiencing my D90 which has the AK4499 too, I knew I wanted an AK4499 DAC for my desktop system. It was a simple as that. I knew the sound profile intimately but it was based on actual listening and synergy within an entire audio stack.

 

[Taz777]

Here's my initial thoughts after two days with the Gustard A18 in my desktop system - around 12 hours of music playback.

Audio chain: iMac --> Audirvana and TIDAL desktop apps --> USB --> Gustard A18 --> Ruark Audio MR1 MkII active speakers / BK Electronics XLS200-FF subwoofer

My 2.1 speaker system is an all-British affair with Ruark Audio and BK Electronics being great suppliers of HiFi equipment in the UK. Here's a link to the subwoofer I use in case anyone is interested in it:

https://www.bkelec.com/HiFi/Sub_Woofers/XLS200.htm

Gustard A18 configuration is out of the box apart from filter is set to SUPER SLOW (for now) and the display is set to auto-off.

Disclaimer: No charts or graphs or numbers here. I listen to music for emotional involvement. Nothing else. This is purely how the music affects me at an emotional level - how it draws me in and evokes memories. How it helps me to forget the crappy times that we are currently in!

Song selection: A selection of soul, R&B and smooth jazz local lossless FLACs (44.1/16-bit), a few Qobuz CD-quality and Hi-Res, and a few TIDAL CD-quality and Masters.

Sound profile: Very, very airy and effortless. The A18 (in my audio chain and to my ears) is highly suited for all-day listening. I work from home and can have my music playong all day long apart from when having online Teams meetings for work. The biggest improvement over my old, and sadly missed, Topping D50 is how the music engages me. I find myself almost distracted with the amount of information retrieval by the A18. There's more dynamic range compared to the D50. The clarity and separation of the intricate layers of songs is a step up. Similarly, the tightness and cohesion of the music is very appealing (to my ears). Instruments sound just right. It's the subtlety that is really brought out by the A18. I often find myself pausing from my work and just taking in a few seconds of music and hearing things that I haven't heard before.

Aesthetics: The Gustard A18 is quite a large desktop DAC (way bigger than a Topping D50). Hence, it does occupy quite some space on my desk. Here's a picture of it in place on my desk.

The subwoofer is hidden away under my desk.

Still to come: Headphones test (using an Arcam rHead headphone amplifier) and comparison with Topping D90 (non-MQA).

 

[Dannemand]

@Taz777:

Awesome, thank you for sharing this!

Your findings match my own: Airy, detailed, but NOT fatiguing in the slightest, and just very engaging. Once again congrats on your new DAC!

 

[Taz777]

Gustard A18 owners, I have a questions about DSD file playback. This is mainly out of interest rather than need. The specs suggest DSD512 capability. However, my Audirvana desktop app on my Mac shows this:

I am a little confused by this. It's quite possible that Audirvana is at fault here. Has anyone tested DSD file playback on the Gustard A18?

 

[Taz777]

Gustard A18 owners, I have a questions about DSD file playback. This is mainly out of interest rather than need. The specs suggest DSD512 capability. However, my Audirvana desktop app on my Mac shows this:

I am a little confused by this. It's quite possible that Audirvana is at fault here. Has anyone tested DSD file playback on the Gustard A18?

It was a setting in Audirvana. You need to drop down the Native DSD streaming method list and select DSD over PCM 1.1 and then Audirvana can read the DSD capability of the A18 DAC.

I downloaded some DSD-256 test tracks from http://www.2l.no/hires/index.html? and was completely blown away by the quality of the recordings. If you really want to max out the quality of the A18 then those DSD-256 test tracks are well worth a listen.

 

[Taz777]

Just a thing I've discovered if you're using Audirvana on a Mac with the A18 connected via USB and you are interested in DSD playback. Whilst your Mac/Audirvana will support Native DSD and send Native DSD to the A18, due to the way it is implemented before it reaches the A18, the maximum DSD support is DSD-256, even though the A18 itself can support up to DSD-512. The maximum rate of the DAC is halved when using a Mac and Audirvana for DSD playback.

In the real world this won't be an issue as there's hardly anything available in DSD-512, but it's worth bearing in mind. You can see on one of my screenshots a few posts above that Audirvana reads DSD-256 as the maximum supported rate.

 

[Taz777]

I'm really happy to report that IIS (IIS-LVDS or I2S) works like a charm on the Gustard A18 DAC! I used my Cayin N6ii DAP running UAPP with TIDAL as a digital transport using an HDMI cable (although I2S is nothing to do with HDMI other than it physically uses an HDMI cable to connect two devices together) to connect to the IIS input on the A18 DAC. It worked straight out of the box!

I use I2S with my non-MQA Topping D90 too. In fact, I struggle to tell any difference between the Topping D90 and the Gustard A18 when using IIS.

With IIS, the Gustard A18 (just like the Topping D90) smooths the audio out a little. Harsh edginess to the sound is reduced. There's a touch more top-end sparkle through my active speakers. If anything, the sound is a little more mellow, but still with lots of detail. It's more laid-back and a little further away.

Note that MQA decoding in the A18 does not work via IIS - this is documented on the Shenzhen Audio site. This is not an issue for me as my Cayin N6ii will natively decode MQA and send a PCM datastream over IIS to the A18 which then shows PCM.

I would say that, on balance, the USB input sounds a touch more dynamic and engaging to me and songs have a little more weight to them, which is something that some people may prefer. However, for ultimate clarity and subtle details, the IIS interface seems to be more revealing.

With the Topping D90, there's a clear sound difference between USB and IIS. IIS wins on all counts. With the A18, I think it's down to personal preference. You get a more intimate and involving presentation with the USB input, and you get a slightly smoother, clearer sound using the IIS input. My personal preference with the A18 is the USB input. On my non-MQA D90, it's the IIS input.

The USB interface on the A18 uses the XMOS XU216 chip. On my non-MQA, the USB chip used is the XU208. This may account for the sonic differences in the A18 sounding better when using USB. The MQA version of the D90 uses the XU216 so that would probably sound similar to the A18, with similar sonic results.

 

[Terriero]

@Taz777 do you think the A18 performance is more or less equal of the Topping D90? (obviously for less price). I'm planning to buy a Flux Lab AMP (FA 12 or Volot). If I buy the Volot, maybe I won't have budget for a expensive DAC and will try with Gustard or Topping, at first...

 

[Taz777]

@Terriero - I'd say it's very similar. The D90 has more and better features though. Remember that there is a non-MQA D90 and an MQA D90. I have the non-MQA D90 and am very happy with it. I'm also very happy with my MQA A18. My audio gear is mid-range and my hearing is very average, so there are perhaps things that I don't hear to differentiate the two even more.