Gustard X16 Balanced MQA DAC Review

[Atanasi]

It is advised it is short, though. Not a couple meters length. I squared S was never intended to cross long lengths of cable.

I2S over HDMI cable converts the internal signal to differential LVDS, so the range is longer than straight I2S.

 

[dan kolov]

ASIO bypasses the Windows Audio system entirely, which can yield lower latency. It was really meant for studios and others to connect their DAW (Digital Audio Workstation) to an interface, sound card or other audio hardware. It requires 3rd party (non-Microsoft) hardware drivers and software.

It is possible that DSD512 or PCM768 might work more reliably with ASIO. Also, I can imagine on a PC with lots of junk running in the background (say, a boatload of Windows update and telemetry services) that ASIO might help avoid some buffer underruns. But I don't know firsthand.

One thing ASIO will do: Guarantee bit-perfect. No Windows software volume etc.

My Windows 10 is not in English but I think you will understand.I am thus configuring my Windows and I have no problems with latency, asio and Gustard X16.The sound from the computer is now much better.Turning off a lot of unnecessary services and a minimum of background applications make the sound amazing.I use audiophile optimizer and fidelizer.And I manually turn off many services.

 

[MiLi]

It is advised it is short, though. Not a couple meters length. I squared S was never intended to cross long lengths of cable.

I'm personally running 6 meters long HDMI cable as interconnection between rooms without problem.

 

[Snert]

Just to remove anything BluRay player or HDMI related, I dug out an old Denon CD player from circa 1989 that I have not used in 15+ years. It still works apparently, and has a coax output. I don't get any issues with it connected to the X16. In fact is sounds good as I can expect from it. Playing at 44.1KHz. No dropouts, maybe some pop's but not as pronounced and could be the wax in my ears.

Of note, Bluetooth sounds fine. The phone does support MQA. So does the DAC, but not surprising the DAC does not unfold MQA over Bluetooth. In fact, most "master" tracks won't play at all over Bluetooth for some reason.

I have tried PCM setting on the Roku. Settings are pretty much automatic depending on what is connected.

The Roku Ultra does apparently support 24/192 over PCM, but depends on the app. Which... got me to thinking that the HDCP handshake to the TV is what is limiting and the older TV may be causing an issue. Since the TV sends out parameters of what it is, then the Roku will only send that compatible signal. If I had the Roku plugged into an AVR, then the ROKU would detect the parameters of the AVR. And maybe I would not be having these issues. The pops or cutouts might also so be caused by the return signal from the TV. I dunno just guessing.

Update.... I went ahead and ordered a Node 2i and connected it to the Gustard X16. Streaming Tidal, Amazon and Spotify sounds amazing. I didn't really have anything to stream music very well so wanted to try this before attempting to send it back to China. I did try the laptop without success briefly, but not leaving it connected to the DAC anyway - maybe I could have eventually made that work though. I also plugged in the optical from the TV (content from Roku) into the Node instead of the X16, and now TV contents sounds amazing as well. So appears the Gustard is kinda sensitive..., and happened to not like my other sources / set ups, or my other sources are just old and inferior (but worked with my older DACS). Before when I could get it to not pop or have dropouts with other sources, the sound was just flat, not robust. Like interesting coffee made without enough beans...

Now the sound connected to my amplifier and old Dynaudio speakers, has an incredible soundstage, and does not at all sound weak, but very robust. I am now happy with it. Thrilled actually. Still has the broken optical port door, but now I'm not using it anyway. Maybe I'll fix it someday, maybe I won't.

Happy for now... onward to the next projects..

 

[ivo.f.doma]

Hi. I want to ask if the Gustard X16 turns on and off according to the input signal, as the Topping D50S does? Respectively, does it automatically switch the inputs according to the incoming signal?

 

[Frio]

Nope just on and off via switch or the display off in the menue. Careful some of those auto turn off functions leave the device on all the time. My e30 had the same drain in stand by as a x16 under load.

 

[misterdog]

Anyone wishing to know what MQA is actually doing should watch this video.

In order to test what MQA is doing, the author had to publish his own music on Tidal..

Is he a relative of Amir ?

 

[AdamG]

Anyone wishing to know what MQA is actually doing should watch this video.

In order to test what MQA is doing, the author had to publish his own music on Tidal..

Is he a relative of Amir ?

There is a separate thread for this. I realize this specific product has MQA capabilities and disgusting this attribute is fine here. However, when we start down the MQA rabbit hole please Take those comments and conversation here please. Link: https://www.audiosciencereview.com/forum/index.php?threads/testing-mqa-is-it-worse-than-flac.21735/

 

[Viper Necklampy]

It is necessary insofar that a modern (delta-sigma) DAC will extrapolate the input data to a much higher rate, so it can output a smooth analog signal. Some explanation here https://www.audiosciencereview.com/...ing-differ-from-over-sampling.3440/post-83318
If you mean how the different oversampling filters differ, there are trade-offs, simply. Slower filters will roll off the last bit of upper frequency, but quite likely only very young people could hear -3dB at 22kHz for example. There's plenty more to read about it if you're really interested... but TL;DR oversampling filter will not make the sound suddenly different, most people will switch filters and hear absolutely zero change.
Just don't put the oversampling OFF, Gustard recommends trying OFF/NOS(no oversampling mode) only when feeding the DAC very high sample rates so that oversampling is not as necessary for a correct output. Otherwise it won't sound very good...

I don't get it a little, if it's called oversampling, why amir graphs shows it doesn't goes up to 24khz? Maybe just because he use measurements cut of 44khz?
Should oversampling go to 96khz or higher?

I would really love a dac that oversample my 44/48khz sources to go 96khz, especially with my young ears and planard headphones that do it well.. It the L-fast a oversample to 'fake' hi-res, or it cut until 24khz only like in graphs?

 

[Dannemand]

I would really love a dac that oversample my 44/48khz sources to go 96khz, especially with my young ears and planard headphones that do it well.. It the L-fast a oversample to 'fake' hi-res, or it cut until 24khz only like in graphs?

Any modern Delta-Sigma DAC will oversample, including X16 unless you choose its NOS filter. And it will oversample far higher than 96KHz. It's an integral part of their noise shaping and how they reconstruct a proper waveform from digital samples.

And you (not you personally) really shouldn't use the NOS filter unless you perform software upsampling with HQPlayer or similar software. Check Archimago's classic NOS vs. Digital Filtering DACs and admire those ugly waveforms from the NOS DAC. All that distortion along with the increased amplitude (over- and undershooting the wave) could make it sound better to some. Of course the fast impulse response too. But higher fidelity it is not.

I think many people haven't given thought to how sparse the samples are when you get close to the Nyquist frequency (1/2 the sample rate): Reconstructing a decent waveform from just 2 samples isn't trivial -- and arguably wasn't accomplished with music until decades after the Nyquist-Shannon theorem came out. But that's a whole other discussion

 

[Viper Necklampy]

Any modern Delta-Sigma DAC will oversample, including X16 unless you choose its NOS filter. And it will oversample far higher than 96KHz. It's an integral part of their noise shaping and how they reconstruct a proper waveform from digital samples.

And you (not you personally) really shouldn't use the NOS filter unless you perform software upsampling with HQPlayer or similar software. Check Archimago's classic NOS vs. Digital Filtering DACs and admire those ugly waveforms from the NOS DAC.

I think many people haven't given thought to how sparse the samples are when you get close to the Nyquist frequency (1/2 the sample rate): Reconstructing a decent waveform from just 2 samples isn't trivial -- and arguably wasn't accomplished with music until decades after the Nyquist-Shannon theorem came out. But that's a whole other discussion

Oh thanks, you are the first one to tell me it does, it's about 2 year i'm biased how to get at least 96khz hi-res bitrate for my Playstation 3 and 4 systems in optical 44khz, turns out with Topping D50 time ago I had hi-res and with X16 i will too, maybe even 384khz sample or more which every filter i choose Nyquist-scream

 

[Dannemand]

Oh thanks, you are the first one to tell me it does, it's about 2 year i'm biased how to get at least 96khz hi-res bitrate for my Playstation 3 and 4 systems in optical 44khz, turns out with Topping D50 time ago I had hi-res and with X16 i will too, maybe even 384khz sample or more which every filter i choose Nyquist-scream

You generally shouldn't worry about oversampling (or upsampling). Modern DAC chips do a great job of reducing noise and distortion and reconstructing the waveform, and oversampling is part of that process. Just worry about getting a bit-perfect digital stream to the DAC.

You can play with the filter choices. You may be able to hear a difference. With ESS chips I generally can (and my hearing is damaged from Meniere's). With AKM chips I can't, although my wife can. The effect on impulse response is most significant, affected by the amount of ringing. As for the speed of the roll-off, you probably won't hear that.

 

[Viper Necklampy]

You generally shouldn't worry about oversampling (or upsampling). Modern DAC chips do a great job of reducing noise and distortion and reconstructing the waveform, and oversampling is part of that process. Just worry about getting a bit-perfect digital stream to the DAC.

You can play with the filter choices. You may be able to hear a difference. With ESS chips I generally can (and my hearing is damaged from Meniere's). With AKM chips I can't, although my wife can. The effect on impulse response is most significant, affected by the amount of ringing. As for the speed of the roll-off, you probably won't hear that.

With Topping D50 a year ago i find differences, i'm 22 and i was using Linear phase fast roll-off, also because i heard it was the most correct by some experts here, but L-phase on Gustard X16 might not be linear-phase fast roll-off..
I'm a bit confused again, you mentioned 22khz roll-off, so the dac doesn't really oversample 44khz signal to much higher range of played frequencies like 40khz+, instead of usual 22khz?

 

[Veri]

I'm a bit confused again, you mentioned 22khz roll-off, so the dac doesn't really oversample 44khz signal to much higher range of played frequencies like 40khz+, instead of usual 22khz?

Note that 44kHz is sampling rate. So the rate at which the file is sampled. You are confusing it with the frequency response. 44kHz sampled files does not mean "played frequencies" to 44kHz. A common misconception.

 

[Dannemand]

With Topping D50 a year ago i find differences, i'm 22 and i was using Linear phase fast roll-off, also because i heard it was the most correct by some experts here, but L-phase on Gustard X16 might not be linear-phase fast roll-off..
I'm a bit confused again, you mentioned 22khz roll-off, so the dac doesn't really oversample 44khz signal to much higher range of played frequencies like 40khz+, instead of usual 22khz?

I edited my post immediately after posting it to avoid that ambiguity. No ultrasonics are going to magically show up in your 44.1Khz file by the DAC's oversampling. It'll be deliberately limited to 22.05Khz by the source. If you're playing a HiRes file, it may contain ultrasonics up to half the sample rate -- but often they don't.

 

[Veri]

I edited my post immediately after posting it to avoid that ambiguity. No ultrasonics are going to magically show up in your 44.1Khz file by the DAC's oversampling. It'll be deliberately limited to 22.05Khz by the source. If you're playing a HiRes file, it may contain ultrasonics up to half the sample rate -- but often they don't.

Yep but your more technical explanation requires proper understanding of Nyquist sampling theorem

 

[Dannemand]

Yep but your more technical explanation requires proper understanding of Nyquist sampling theorem

I know. I tried to keep it simple, but evidently I didn't succeed

 

[Viper Necklampy]

I edited my post immediately after posting it to avoid that ambiguity. No ultrasonics are going to magically show up in your 44.1Khz file by the DAC's oversampling. It'll be deliberately limited to 22.05Khz by the source. If you're playing a HiRes file, it may contain ultrasonics up to half the sample rate -- but often they don't.

So so so, at the end of the day the dac's oversampling, won't oversample "low res" 44khz or 48khz data in optical, into higher?

 

[Dannemand]

So so so, at the end of the day the dac's oversampling, won't oversample "low res" 44khz or 48khz data in optical, into higher?

It WILL oversample it, but that won't add any musical information not already present in the file. The oversampling is done merely as a technical step in the process of turning the digital signal into a proper analog waveform. It reduces rounding errors, if you will.

44 and 48KHz are actually very high quality if preserved intact all the way to the DAC. Particularly with a good DAC, such as the Gustard X16.

 

[Veri]

So so so, at the end of the day the dac's oversampling, won't oversample "low res" 44khz or 48khz data in optical, into higher?

Oversampling is a necessary step to get a correct (analog) output.