• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Gustard X16 Balanced MQA DAC Review

I guess I'm a bit confused too as I see this common question about going from dac directly to amplifier or powered speakers.

I've considered running dac at full volume and use something like passive volume controller
https://www.amazon.com/Nobsound-Fully-Balanced-Single-Ended-Pre-Amplifier-Controller/dp/B01MXC9HHW

Or is that over complicating things and best to just use the digital volume control included on the DAC?

I'm always left wondering if one method of volume control is inherently worse for SINAD, or sound in general?
There's a difference between technically better and actually audibly better. Passive volume control is better but it's because it can attenuate the source's noise. If the source has low enough noise it's already inaudible so doesn't make much difference.
 
As a general input on the debate about volume control on a DAC:

ESS posted the following presentation several years ago, in which they make the case that digital volume control isn't as bad as its reputation when performed in the 32-bit space of the DAC chip AND when compared to the challenges of adding analog volume control either in the output stage or a preamp -- both of which have quality cost as well, depending on their implementation.

http://www.esstech.com/files/3014/4095/4308/digital-vs-analog-volume-control.pdf

If feeding an integrated or an amp with gain control, then obviously set that to keep DAC output as high as possible. But my take is that minor adjustments on the DAC (up to -10dB) should be acceptable and inaudible.

I am feeding a Bryston power amp directly from my Gustard A18. The amp has input sensitivity of merely 0.78v, and its gain pots are set pretty low to compensate for the very hot 6.3Vrms signal from the Gustard. I then use the Gustard remote to perform daily volume adjustments. The Bryston has separate L/R pots so it avoids the balance issue of some pots, but of course pots are not perfect either.

What I would NOT do are:

1) Use software volume control, since it operates in the 16 or 24-bit space and really cuts into the effective resolution.

2) Feed DAC through XLR directly into a power amp with no gain control where you have to adjust the the DAC way low. This has to hurt resolution. In that case using RCA out to cut voltage in half seems a better option, although it loses the balanced advantage. I don't have the experience to say if an XLR attenuator (as suggested) is a better solution. Could well be.

Please take this merely as input. But do read that presentation from ESS before concluding. It's not that long :)
That presentation is actually somewhat misleading.
Any 24bit or 32bit DAC currently are the same. If the dac is inherently 24bit or above there's no way the volume control is anything less. So as the DAC output stage is noisier than 24bit quantization noise, having 24bit 32bit or 64bit volume control doesn't make any difference. In the actual presentation it's technically correct saying ESS's performance is then limited by noise output stage. But so is others.
The presentation on the other hand should focus on that passive preamp is better because it's after the output stage. If you have low enough noise, adding an A to D converter then do the digital volume control then convert back to analogue is the same as the passive preamp. The key is the attenuation is after the DAC's output stage.
 
Fair point!

If we assume the Windows code was written with a modicum of intelligence -- performing calculations on 16/24-bit samples as floating point -- it would result in minimal loss of precision. AND if we assume that the DAC was developed with no intelligence --performing internal calculations as 32-bit integers, and gleefully performing divisions before multiplications -- it would result in maximum loss of precision.

Maybe I am naïve, but I think we can reasonably expect ESS and AKM engineers to be smarter than that when it comes to internal calculations. It doesn't cost much to use 64-bit registers to perform calculations on 32-bit samples. Certainly now, 10 years after the presentation was made, that is plausible silicon-wise, although I don't know if they do that.

As you mention in your post, while their PDF is correct in principle, their numbers are off. The entire presentation is a bit over-simplified -- in addition to being old. I assume it was a sales pitch for their new (at the time) 32-bit DAC.

But their main discussion is that of digital versus analog volume control once you go from a 16/24-bit to a 32-bit DAC. And even to this day the assumption remains that DVC is horribly flawed (which it IS if performed in 16-bit space) and AVC is better by definition (which isn't always true).

I am of the mind that I would rather leave the DAC to process my audio stream than the player software. And again, MQA necessitates it, which wasn't a consideration 10 years ago.

Still, fair point. Thank you for sharing!
 
Last edited:
That presentation is actually somewhat misleading.
Any 24bit or 32bit DAC currently are the same. If the dac is inherently 24bit or above there's no way the volume control is anything less. So as the DAC output stage is noisier than 24bit quantization noise, having 24bit 32bit or 64bit volume control doesn't make any difference. In the actual presentation it's technically correct saying ESS's performance is then limited by noise output stage. But so is others.
The presentation on the other hand should focus on that passive preamp is better because it's after the output stage. If you have low enough noise, adding an A to D converter then do the digital volume control then convert back to analogue is the same as the passive preamp. The key is the attenuation is after the DAC's output stage.
Thank you. Excellent clarification.

While ESS may have made that presentation to promote their Sabre DAC back then, I always saw it just as providing a more nuanced look at digital versus analog volume control in general!
 
Thank you. Excellent clarification.

While ESS may have made that presentation to promote their Sabre DAC back then, I always saw it just as providing a more nuanced look at digital versus analog volume control in general!
In that case, true. Passive preamp(good performing ones) is better in SNR than volume control in the DAC.
As the dac's noise is approaching or beating 1uV(1.3uV or so for okto dac8 stereo), analogue attenuators don't bring much benefit anymore. It's still around 2.3uV for X16/dual 9068 though so there's room for improvement.
 
I thought you got your replacement, wait for replacement I guess.
In my house, I don't remember I ever get any static shocks. Humidity is very high in my area, so I took calculated risk to examine x16 with 2009 L30, using low cost headphones, and for voltage measurements only. BTW, Apos placed order for new L30 for me, but I still got no shipping notice yet....

Anyway, if I typically listen at -13db with x16, would it be better to use L30 to do -10db? And thus I listen with -3db via x16? Or it doesn't matter?

Thanks!
 
In my house, I don't remember I ever get any static shocks. Humidity is very high in my area, so I took calculated risk to examine x16 with 2009 L30, using low cost headphones, and for voltage measurements only.

Anyway, if I typically listen at -13db with x16, would it be better to use L30 to do -10db? And thus I listen with -3db via x16? Or it doesn't matter?

Thanks!
Minus the DC part, difference is minimal. But you can do more attenuation on L30 for better technical performance that's for sure.
 
Does this output on XLR and RCA simultaneously?

I'm thinking of getting this and using it as DAC and preamp to a power amp. Now the power amp I'm looking to get doesn't have a pre-out. The speakers I'm looking to drive have an active sub but passive wide range. Now they do have a high level in so I could just feed the amped signal, but I was told that best would be to feed amped signal only to the wide range in and pre-amp signal to the sub in.

So I was thinking of going xlr to amp to wide range and RCA to sub?
 
Does this output on XLR and RCA simultaneously?

I'm thinking of getting this and using it as DAC and preamp to a power amp. Now the power amp I'm looking to get doesn't have a pre-out. The speakers I'm looking to drive have an active sub but passive wide range. Now they do have a high level in so I could just feed the amped signal, but I was told that best would be to feed amped signal only to the wide range in and pre-amp signal to the sub in.

So I was thinking of going xlr to amp to wide range and RCA to sub?
Yes, both simultaneously.
 
Thank you.
Any idea on my use-case? Would that make sense to feed the rca directly to the sub?
Use case is fine. The only issue is I can see is about the big pop sound when turning on and off x16. This is true for RCA, but xlr have no such issue.
 
Make sure they are not waiting for you to pay import fees, I only found out by clicking on "manage delivery" in a txt from DHL, when I entered the tracking number direct into DHL tracking it didnt mention the fee.

This is not how it works in Germany. Either you get a letter from customs, or the postman rings at your door and expects you to pay in cash (no, seriously). But the tracking number is not international, which means that it is going to be sent from Germany once they get it, so I think I will not have to pay dues.
 
Although I mentioned I prefer E30 for its sparkles, I found a type of music I like the most with x16 today: Jennifer Thomas' The Fire Within.

I like x16 the best in comparison to these two, based on my recollections of previous playbacks, not side by side comparisons:
Nod2i own DAC connect to Marantz 8801
Topping E30 direct to amp

I found x16 a bit more satisfying. Each instrument have more "feels" to it. I wanted to listen more and finish the whole album. The other two were a little thin sounding, less engaging, and I loss interest fast, maybe a song or two and I stopped.

That's all for today. : )
 
I am of the mind that I would rather leave the DAC to process my audio stream than the player software.
Perhaps you can read this as well, of course including the linked article and subsequent replies. Also, try the test files.
https://www.audiosciencereview.com/...y-of-software-volume-control.5922/post-184254
Pretty much the same way someone asked me to read that ESS pdf, so I did carefully read it, more carefully than most people I suppose. In this way, I also hope you can also carefully read what I wrote.

Also a little program I wrote, can be used to check your audio files with foobar2000 and SoX so that you can see how things change when you use digital volume control. A very detailed ReadMe is included with the program, so I hope you can also read it carefully as well.
https://www.audiosciencereview.com/...he-obsession-with-dr-meters.11297/post-649576

For the MQA part, @mansr is very familiar with it, perhaps you can also search his posts. Of course, it would even be better if he joins the discussion here.
 
Would this be overkill for a pair of Genelec 8030Cs and an SVS PB-1000 sub? Since this outputs to both XLR and RCA out simultaneously, I will just plug the 8030 to the XLR out and the PB-1000 to the RCA out. Is there something similar that is cheaper and not much of a downgrade in terms of performance? I don't really care about MQA.
 
To avoid customs and I had to have it quickly I bought last week at aufiophonics and got it already. Is there something different compared to those from china? I heard there is no need for a Firmware update.

you did not avoid customs. audiophonics already paid it before so you paid more overall.
 
Back
Top Bottom