Gustard U12 USB Interface 8 Core XMOS chip

[amirm]

My pleasure. What is the best way to test that? Is there a low cost DAC with I^2S?

 

[DACLadder]

The U12 is in the ballpark on SPDIF output. +/-200mV is the SPDIF minimum spec level. Nomimal level is +/- 500mV into a 75 ohm load. With a short cable should work OK. Driving a really long cable though may become a problem due to normal cable attenuation at the end point termination.

The biggest issue with the U12 I2S output is the phase of some of the signals on the cable pairs are backward from the much more popular "PS Audio" I2S wiring method using HDMI hardware. The U12 will play but the left and right channels are swapped plus the absolute phase of the audio signal is 180 deg out if connected to "PS Audio" wired DACs.

The U12 cost is attractive but today there are much better options for a DDC. The U12 XMOS USB chip is dated plus the clock oscillators aren't the best. It is a minimalist design sold for a good price. I used a modified U12 for a long time until I purchased the SIngxer SU-1. The SU-1 is more versatile as it has switches that allow you to configure the HDMI I2S output signals for your specific DAC. Cost though is about 2.5 times as much but the SU-1 uses a newer XMOS chip (faster/ more memory) and employs the expensive, highly regarded, and low-jitter Crystek 957 temperature compensated oscillators.

 

[amirm]

Thanks for the info. I am hesitant to spend $400 on yet another device to test. Would you be willing/interested in loaning me your SU-1 to be measured?

 

[DACLadder]

I use my SU-1 daily for hours on end so I can't part with it. But forum member @bimmer100 sells the SU-1s (Kitsune HiFi) and lives in the Seattle area (I believe). He may have a loaner.

 

[Wynnytsky]

I use my SU-1 daily for hours on end so I can't part with it. But forum member @bimmer100 sells the SU-1s (Kitsune HiFi) and lives in the Seattle area (I believe). He may have a loaner.

I lost interest in the U12 long ago, but I was really looking forward to see how you'd phrase your reply! I'm a generous guy but I'd sooner loan my cat than the SU-1.

I recall the PUC2 Lite being the best sounding non-I2S interface. I wonder how the LKS 100 + a 5v TeraDak would stack up against the SU-1.

 

[amirm]

I recall the PUC2 Lite being the best sounding non-I2S interface. I wonder how the LKS 100 + a 5v TeraDak would stack up against the SU-1.

All of these questions would be answered if the people who have them would loan them out for measurements. It takes me a day to measure them and I can send then them back priority mail at my expense. Such data would help the entire community. A number of people have done so. And I keep buying gear myself but more needs to be done if we want real data to use for our decisions as opposed to anecdotes.

 

[DACLadder]

What are you measuring besides the SPDIF levels on the cable? Clock jitter is probably the most important spec of all. That seems to separate the really great products from the average in digital audio.

 

[amirm]

Clock jitter only matters once it is internally filtered by the DAC (chip). Raw clock output on S/PDIF is not relevant. Proof of this is when I fed the multiple different S/PDIF bridges to Exasound E32:

The internal PLLs in the Exasound E32 do an excellent job here and hence, make the result the same regardless of which USB bridge is driving it.

Ultimately what you hear is the analog output of the DAC, not the clock input to the S/PDIF port. And that is what I measure.

Clock jitter that doesn't make it to the analog output of the DAC is immaterial.

 

[amirm]

Note also that low frequency clock jitter is immaterial as that is massively masked by our hearing system. So to the extent low clock jitter is advertised as being so at low frequencies, that too is improper.

 

[riffer]

Nice to see some actual measurements of these devices. I've had a bunch here (Including U12, SU-1, Off-ramp 5) and I still prefer the sound of the U12 from my Squeezebox Touch into my Audio-gd Master 7 BNC input. I'm going to leave it in the system for now as I have other upgrades to do - namely speakers.

 

[xevman]

I'm probably gonna invest in either the u12 or su-1. The optical out of my pc has serious clock issues no matter what DAC i pair it with. I get audible clicks and pops whilst my music plays using the optical out. Switched from the optical out of the motherboard to the optical out of my soundcard to my various Dacs and it fixed the issue. Should pair nicely with my original Benchmark DAC1.

 

[Nu3nO]

I'm probably gonna invest in either the u12 or su-1. The optical out of my pc has serious clock issues no matter what DAC i pair it with. I get audible clicks and pops whilst my music plays using the optical out. Switched from the optical out of the motherboard to the optical out of my soundcard to my various Dacs and it fixed the issue. Should pair nicely with my original Benchmark DAC1.

If you are running Windows, you might want to check your DPC latency. If the latency is above 1ms then clicks/pops will surely occur & most of the time it's the network adapter driver causing this issue. My older Killer network driver caused this problem & I had to disable the Advanced Stream Detect feature to resolve it. The tools to check latency can be found here:

http://www.thesycon.de/eng/latency_check.shtml
http://www.resplendence.com/latencymon

Cheers!

 

[xevman]

I'll investigate but it's strange how switching to the sound cards optical out fixed the issue. If it was latency related wouldn't I have issues here aswell?

 

[xevman]

Highest I recorded with the latency checker was 1003 μs so I dont think its a latency issue here

 

[ginetto61]

Clock jitter only matters once it is internally filtered by the DAC (chip). Raw clock output on S/PDIF is not relevant. Proof of this is when I fed the multiple different S/PDIF bridges to Exasound E32: ... the internal PLLs in the Exasound E32 do an excellent job here and hence, make the result the same regardless of which USB bridge is driving it. Ultimately what you hear is the analog output of the DAC, not the clock input to the S/PDIF port. And that is what I measure.
Clock jitter that doesn't make it to the analog output of the DAC is immaterial.

Hi and thanks a lot for the very interesting investigation. If i understand well it looks like the spdif input of the Exasound is very unsensitive to jitter in the incoming signal ? (sorry for the bad english). I think that the availability of an AES digital out in the Gustard U12 is a nice thing. I have been using it with an old but nice apogee rosetta 200 and the sound was the best i have heard on my system. Maybe the real problem is the system