• 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!

Review and Measurements of Gustard U12 USB to S/PDIF Converter

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
46,033
Likes
256,916
Location
Seattle Area
This is a review of Gustard U12 which converts USB input to S/PDIF, Toslink, AES/EBU and I^2S. The U12 retails for $170 shipped from Amazon as of this writing. I will be comparing it to Audiophilleo 1 (about $500), Fiio E10K ($76) and Melodious-Audio MX-U8 ($269). Sadly my Berkeley Alpha USB is broken so I can't test against it even though it is at much higher price point.

Gustard U12.jpg
Overview
The Gustard U12 is for people who have existing DACs and would like to add USB connectivity to it. The device has very good functionality with both consumer S/PDIF and AES/EBU professional digital outputs. It also has I^2S output for the few DACs that support it. For this testing, I focused on S/PDIF output only. Let me know if you want me to test other outputs.

I tested the unit without any drivers. Source was Roon player which reports the following formats:

upload_2017-12-17_16-30-43.png


So as is, support is limited to 192Khz, not 384KHz as advertised.

Reliability was excellent in the few hours of testing so no concern there.

A front panel display shows the sampling rate which I like to see. It is a simply LED 7-segment but useful to figure out if resampling is in the pipeline without one knowing.

The box is brushed aluminum and hefty example of that seeing how it includes the power supply inside. That also makes the unit quite hefty for the little box that it is. Overall, it is still a low-budget affair but at the high-end of that spectrum. And fit for the price charged.

Measurements
Since this device is a digital input, digital output, I started my test with an "eye pattern." I played 24-bit/48 Khz J-Test inside Roon as usual but then measured the time domain response of the S/PDIF stream:

Gustard U12 vs Audiophilleo SPDIF eye pattern.png


The purpose of the eye pattern is to show much of safety margin we have in a serial communication stream to detect the "one and zeros." The red zone is the defined minimum by AES. This is for balanced output which I am not testing but gives you an idea. The Audiphilleo has a pretty strong output that is also clean, able to clearly get above the minimums standard. The Gustard U12 on the other hand, has anemic voltage output that was nearly half of Audiophilleo. As such longer cables could pose an issue for it.

Next we jump into our favorite test, J-Test in all cases driving the Topping D30 DAC. Here, I am zooming in both in frequency and amplitude to show the more interesting bits. First comparison is against the Audiophilleo:

J-Test vs audiophilleo.png


As we see, the Audiophilleo easily outperforms the Gustard U12. I should mention that my Audiophilleo is a first generation device that I bought nearly 10 years ago. There is now second generation version of it. Despite that, it has no problem producing the best the Topping D30 can through its S/PDI input (which I should say, is worse than its USB input). So good design trumps age!

Next, let's go to the other spectrum, testing against the much cheaper DAC+Headphone Amp Fiio E10K:

J-Test vs Fiio E10kpsd.png


Alright so something had to give for $79 and all this functionality and the Fiio E10K generates worse performance. Not an audible concern mind you but objectively so. Note that I am only using the digital S/PDIF output of Fiio E10 to drive the Topping D30. Not its DAC.

Rummaging through my bins, I find a loaned Melodious-Audio MX-U8 from member Mivera. It has identical functionality to Gustard U12 but in a bigger box.

Gustard U12 J-Test vs Melodious X-U8.png


In this zoomed display the Melodious-Audio MX-8 does better but when tested broadband (not shown) it had a spike that bothered me. So I say it is a toss up which one is engineered better.

So far we have been testing with a budget DAC. For the next set of tests, I used the Exasound E32 as the DAC (around $3,500).

Gustard U12 J-Test vs Audiophilleo Fiio E10K using Exasound E32 DAC.png


As we see here, there is no difference at all how you drive the S/PDIF input on the Exasound E32. It excellently cleans up whatever ills may be on S/PDIF and renders the same excellent output. So if you have a high-quality DAC, there is no reason to anguish over how you are driving it.

Summary
The Gustard U-12 is an "OK" engineered product. I like its display and connectivity so I am not opposed to someone buying it. But it doesn't get my recommendation.

As usual, comments, questions, corrections are welcome.
 
Last edited:
Have you been good for Santa???
Well I always left a minced pie and a drum of whiskey out for him.. that seemed enough to wipe out any past discretion’s .

Santa don’t come around no more, my mum tells me this is because I’m too old and he’s busy with the younger folk rather than any ill of my character but then mums would say that wouldn’t they.....
 
Thanks for these measurements Amir. Many people tout the Gustard. The Exasound shows how things should be handled.
 
Many thanks, Amir! I would only note that the Gustard U12 costs $170, not $260.

I use it because the drivers are bulletproof and never lock up, not because of any improvement in sound quality.
 
Many thanks, Amir! I would only note that the Gustard U12 costs $170, not $260.
My pleasure. I got the price right at the start but somehow wrong at the end :). I deleted that reference.
 
Houston, we have a problem!!! :D I don't have a DAC with AES/EBU digital input on my bench! My main DAC in the audio system does but it is not close to my bench to measure. I might rip it out to test later but for now, I thought we settle for the eye pattern:

AES Eye Pattern Gustard U-12 vs Melodious-Audio MX-U8.png


As we see here, the Gustard yet again puts out anemic output. The Melodious-Audio MX-U8 does much better.

AES3 spec requires 2 to 7 volt peak to peak. The Gustard U-12 comes at 1.76 volts peak to peak vs 3.4 volts for MX-U8. In that regard it is not compliant with the AES reconnections.
 
I thought you had a Tact still. Well maybe not on the bench. That had aes input.
 
I thought you had a Tact still.
The TacT is a doorstop. I tried to update its firmware a few years ago and it bricked. I asked what's his name for a fix and he said there is none. He has no parts and I shouldn't have tried to update its firmware!
 
OK, I ripped out my Mark Levinson No 360S DAC, circa 1999. In other words, it is 18 years old! In human years, that is like 108. :D

Let's compare the AES output while driving the Mark Levinson DAC comparing the Gustard U12 to Melodious-Audio MX-U8:

AES ML360S AES Input.png


This is a clear win for the Melodious-Audio. The Gustard has those two jitter peaks in yellow.

Next, let's compare the S/PDIF to AES output on the same device. First, is Gustard U-12:

AES vs SPDIF Gustard U12 ML360S AES Input.png


Using S/PDIF shifted the frequency of jitter components but otherwise they are there just the same as AES. It however also added random, low-frequency jitter which causes the "skirts" or broadening of the 12 Khz tone in J-Test. Given this, if you have a choice, I say go with AES output.

Let's run the same test on Melodious-Audio MX-U8:

AES vs SPDIF Melodious-Audio MX-U8 ML360S AES Input.png


Very clean and no sign of either the jitter or low-frequency random noise that Gustard U12 has.

As a side note, what do you guys think of my old, dusty Mark Levinson DAC??? Darn clean, no? Yes there is a component at 1 Khz but it is down 128 db or so (the peak is at 8 db). I will do more testing on it now that I have it out but for now, this shows that good engineering last a long, long time.
 
All makes me wonder if I should invest in a Singxer SU-1, which, at $400, has galvanic isolation, the Amanero USB chip (purportedly superior to the XMOS chip), and, according to audiophiles interested in USB devices, better sound than the Gustard. Without Amir's measurements, it's unknown whether the Singxer has appropriate output levels.

Once I take delivery of the Dutch & Dutch 8Cs, I'll just need a way to get USB audio from a computer into the AES ports of the 8Cs. With DSP in the Dutch & Dutch 8Cs, I'm not sure that the Gustard's jitter spikes will be audible, and with reasonably short cable runs, the low output might not be problematic.
 
I don't think my measurements show audible problems. It is just lack of good engineering relative to alternatives.

That said, I also got the feedback on another forum that Singxer SU-1 is the new poster child in this category. I asked the members there for loaning it to me and they were unwilling.
 
I don't think my measurements show audible problems. It is just lack of good engineering relative to alternatives.

That said, I also got the feedback on another forum that Singxer SU-1 is the new poster child in this category. I asked the members there for loaning it to me and they were unwilling.

Lol, maybe I'll have to buy one and loan it to you.
 
Sorry for the off-topic, but besides the great measurements, great audio engineering and 100% non-profit answers, you know what else I love to this forum?
Free speech, nice admins/moderators and the ability to post links from other forums. :) I almost forgot how great it is to post a link without being worried if I broke any of so-called "rulez" or I'll get a perm ban or anything...it's like the 90's are back here. :)
 
Back
Top Bottom