Minidsp U-DIO8 USB to AES Converter Review

[Parametric]

Anyone know if this device works with the ps5?

 

[Kal Rubinson]

Do do what?

 

[Parametric]

Do do what?

To output 5.1/7.1 digital audio via the ps5s usb port trough this device. I would like to go U-DIO8 - ps5 - Genelec SAM.

 

[Kal Rubinson]

I doubt it. I have no idea what the PS5's USB output capabilities are but the U-DIO8 input requirements are:
Driverless interface for Mac OS X v10.6.4 and up
Thesycon Windows driver provided for WDM/ASIO support
Linux compliant when used with Alsa 2.0 driver

 

[zym1010]

To output 5.1/7.1 digital audio via the ps5s usb port trough this device. I would like to go U-DIO8 - ps5 - Genelec SAM.

for true 5.1/7.1 audio using Genelecs, I believe you need something like https://www.genelec.com/9301a to do manage the crossover properly. Too much of an investment for me, so I kinda gave up on the idea of using multichannel genelecs.

 

[zym1010]

The jitter is almost certainly because the XMOS implementation MiniDSP uses synthesizes the audio clock from the XMOS master clock, and doesn't do it particularly well. It's probably the same issue that other people have noted with the MiniDSP USBstreamer.

I wonder what are the references for "same issue that other people have noted with the MiniDSP USBstreamer"? I'm currently using USBstreamer B to feed my Genelecs and I've also recently placed an order of expensive interface Matrix X-SPDIF 2, looking forward to checking if I can hear any difference. If the issues of USBstreamer is audible, then probably I will have a better chance of keeping the Matrix instead of returning it...

 

[gondorff]

The RME Digiface is cheaper (at least where I live) and you get RME drivers including TotalMix. It has more features (headphone out, and you can record with it), and probably less jitter as well.

The difficulty with this alternative is that one has to somehow convert ADAT to AES/EBU to get the same functionality as the U-Dio8. Only solutions I found for this (converter) need an investment as high as the box itself for the 8 channels. Have I missed something?

 

[madakuska]

I wonder if the unit works with Raspberry Pi + Kodi usb output. Anybody has a direct experience with this setup?

 

[mdsimon2]

I wonder if the unit works with Raspberry Pi + Kodi usb output. Anybody has a direct experience with this setup?

It is USB class compliant, and they advertise linux compatibility, pretty safe to assume it works. I don't have a U-DIO8 but do have a USBstreamer which is the same underlying hardware with slightly different firmware and it works fine on a raspberry pi.

Michael

 

[Henry Pootel]

I also have their MCHStreamer Lite and it works with a Raspberry Pi and Ubuntu (didn't try Kodi). I used @mdsimon2 's guide to get camilladsp running with it... Thanks Michael! Worked great!

I just ordered a U-DIO8 (RCA SPDIF) and have been testing it, however, and am not able to get it to work at 192 KHz. It works at 48 and 96 KHz, but not at 192 KHz. The MCHStreamer worked great at 192 KHz. Anyone using a U-DIO8 at 192 KHz?

By not working, I mean that the SPDIF signal only intermittently gets through. The test signal path is:

PC -> (USB) -> Topping D10 -> (SPDIF) -> U-DIO8 -> (USB) -> Ubuntu -> (USB) -> U-DIO8 -> (SPDIF) -> Topping E50 -> (Analog) -> E1DA -> (USB) -> PC

I'm using REW on the PC to generate and view the analog signals. I simplified the Ubuntu system by using Jack to connect the input to the output for each channel. This worked great with the MCHStreamer, but not with the U-DIO8 unit that I received. I am hoping that it is simply a defective U-DIO8 and that it actually works. Here is a screen shot of REW and sine wave with channel 1 (in and out) of the U-DIO8:

and here is a screen shot of REW and sine wave for channel 5 (in and out) of the U-DIO8:

I made no changes to the PC or Ubuntu system for these two tests -- started the test and left it running... and then just swapped between the U-DIO8 cables. So, more noise on some channels than others.... I have a support email in and and waiting for their response. Hopefully, it is a defective (noisy) cable or defective U-DIO8 because I really like their products and would like to use this in a system I am building.

 

[mdsimon2]

192 kHz often doesn't work well with SPDIF, although usually you see this more with optical than coax.

With that setup you have a lot of potential culprits which makes it difficult to say which device is at fault.

I assume when you are saying the U-DIO8 only works at 96 kHz and below you are referring to the U-DIO8 sample rate? And you are keeping the D10 sample rate constant while varying the U-DIO8 sample rate?

The Topping DACs are notoriously bad with SPDIF input and from the ASR measurements the U-DIO8 doesn’t have the lowest jitter. I would test the U-DIO8 without the D10 (I think that should be possible) in to the E50 into the Cosmos. If you do not get see hiccups in REW that suggests a D10 output / U-DIO8 input issue.

If you think the issue is the D10 output / U-DIO8 input, I’d test the D10 in to the U-DIO8. Use the U-DIO8 as a capture device and see if you get any hiccups in REW.

I might be missing something but the D10 seems redundant in your setup.

All that being said, if it works at 96 kHz I’d probably just run with that.

Michael

 

[Henry Pootel]

Thanks for your help! Still working my way through the differential diagnosis...

When I run a test, I have -- all -- components in the signal path (even the ED1A ADC) at the same sample rate. I also verify that the D10 and E50 are displaying the rate under test.

Of course, I first tested with a loopback: PC -> (USB) -> Topping D10 -> (SPDIF) -> Topping E50 -> (Analog) -> E1DA -> (USB) -> PC to make sure that worked.
I inserted the MCHStreamer and Ubuntu into the path and it worked.
I then replaced the MCHStreamer with the U-DIO8 and it failed at 192 KHz.
I then generated a signal on the Ubuntu system and played it back through the system using these commands:

$ sox -V -r 192000 -n -b 32 -c 8 square1k.wav synth 30 square 1000
$ aplay -D hw:1,0 square1k.wav

and that test was successfully. This means the second half of the processing chain works -- I can get a signal from Ubuntu through U-DIO8 at 192 KHz. So, the U-DIO8 output and E50 input SPDIF is working.

So, this suggests that the problem is the SPDIF input -- either the cable, or the unit, or some incompatibility between the D10 and U-DIO8.

Next, I did something similar to what you suggest -- I sent a signal from the PC and recorded it on Ubuntu, and then was going to copy that file onto the PC to analysis. This will check the first half of the path, but I haven't copied the file off and analyzed it. I paused testing and was waiting for their response to my support request. Maybe they have seen this before and know the problem.

(I had also used an oscilloscope to look at the SPDIF waveforms between the D10 and UDIO8 and I'll say that the cable provided with the U-DIO8 isn't helping (the D10 signal at the end of a 1m generic cable looks better than that same D10 signal at the end of the U-DIO8 cable (have to measure at the D25 connector), but I think it is good enough.)

The D10 is simply an SPDIF output for the PC for test purposes. In the actual system I will be building, the SPDIF source will be a blue sound Node and there will be 3 Topping E50s for six channels on to the amplifiers... the signal path will be:

Node -> (SPDIF) -> U-DIO8 -> (USB) -> Ubuntu (CamillaDSP) -> (USB) -> U-DIO8 -> (SPDIF) -> Topping E50 (x3) -> Amplifiers -> speaker drivers

This is replacing a decade old system that used a heavily modified DCX2496 for the crossover, equalization and time delay. Now I will have a much better DAC (with the E50s) and will have much more DSP power with CamillaDSP. Well, assuming I can get it to work!

 

[Henry Pootel]

As an update, the minidsp team responded and asked me to simplify the processing chain... so I went as simple as possible with the following:

Bluesound Node -> U-DIO8 -> PC

I loaded a wav file generated by REW (1KHz sine wave at 192 KHz) onto the Node for playback, and here is what I get:

Similar distortion as what I was getting with the D10 as the SPDIF input into the U-DIO8.

As a sanity test, I used the following processing chain:

Bluesound Node -> Topping E50 -> E1DA -> PC

And got this perfect waveform (of course, verified that the E50 was locked at 192):

So, I am thinking that either the U-DIO8 is really finicky about SPDIF input (and is incompatible with both the D10 and Bluesound Node at 192 KHz) or I have a defective cable or U-DIO8.

Has anyone tried a U-DIO8 at 192KHz with either a Topping D10 or Bluesound Node? Thanks!

 

[Henry Pootel]

I came up with a very simple test for the U-DIO8. I connected the “OUT 1-2” SPDIF RCA connector to the “IN 1-2” SPDIF RCA connector and send a 192 KHz test signal via USB from a Windows computer through this loopback. Just the U-DIO8 connected to the PC, its all digital (no DAC or ADC) and for the SPDIF portion of the signal path the U-DIO8 is talking to itself. Here is a photo of the setup:

I used a RCA connector to join the OUT and IN cables. I used REW on the Windows PC to generate a 1 KHz sine wave and send through this loopback test and here is the received signal:

The SPDIF signal gets so distorted, that the U-DIO8 can rarely decode and extract the sine wave data. So, there appear to be some significant signal integrity issues in this Minidsp cable and U-DIO8. It can't even talk to itself.

Looking back at @amirm 's original review, I see that he tested at 48 KHz and was testing the AES/EBU variant. The balanced signal pair and differential receiver might help hide some of the signal integrity issues of their design (and the APX555 may be very forgiving), but with a SPDIF RCA unbalanced connection you need to step up your game.

I am hoping that this is just a bad cable or unit, so don't take these test results as representing minidsp products. I really like their products and I think the team there is the best -- they have been very helpful and nice.

 

[312elements]

I came up with a very simple test for the U-DIO8. I connected the “OUT 1-2” SPDIF RCA connector to the “IN 1-2” SPDIF RCA connector and send a 192 KHz test signal via USB from a Windows computer through this loopback. Just the U-DIO8 connected to the PC, its all digital (no DAC or ADC) and for the SPDIF portion of the signal path the U-DIO8 is talking to itself. Here is a photo of the setup:

View attachment 362381
I used a RCA connector to join the OUT and IN cables. I used REW on the Windows PC to generate a 1 KHz sine wave and send through this loopback test and here is the received signal:

View attachment 362385
The SPDIF signal gets so distorted, that the U-DIO8 can rarely decode and extract the sine wave data. So, there appear to be some significant signal integrity issues in this Minidsp cable and U-DIO8. It can't even talk to itself.

Looking back at @amirm 's original review, I see that he tested at 48 KHz and was testing the AES/EBU variant. The balanced signal pair and differential receiver might help hide some of the signal integrity issues of their design (and the APX555 may be very forgiving), but with a SPDIF RCA unbalanced connection you need to step up your game.

I am hoping that this is just a bad cable or unit, so don't take these test results as representing minidsp products. I really like their products and I think the team there is the best -- they have been very helpful and nice.

Getting ready to place my order. I have no interest in upsampling, but curious if you ever found a resolution to this issue.

 

[Hayabusa]

I came up with a very simple test for the U-DIO8. I connected the “OUT 1-2” SPDIF RCA connector to the “IN 1-2” SPDIF RCA connector and send a 192 KHz test signal via USB from a Windows computer through this loopback. Just the U-DIO8 connected to the PC, its all digital (no DAC or ADC) and for the SPDIF portion of the signal path the U-DIO8 is talking to itself. Here is a photo of the setup:

View attachment 362381
I used a RCA connector to join the OUT and IN cables. I used REW on the Windows PC to generate a 1 KHz sine wave and send through this loopback test and here is the received signal:

View attachment 362385
The SPDIF signal gets so distorted, that the U-DIO8 can rarely decode and extract the sine wave data. So, there appear to be some significant signal integrity issues in this Minidsp cable and U-DIO8. It can't even talk to itself.

Looking back at @amirm 's original review, I see that he tested at 48 KHz and was testing the AES/EBU variant. The balanced signal pair and differential receiver might help hide some of the signal integrity issues of their design (and the APX555 may be very forgiving), but with a SPDIF RCA unbalanced connection you need to step up your game.

I am hoping that this is just a bad cable or unit, so don't take these test results as representing minidsp products. I really like their products and I think the team there is the best -- they have been very helpful and nice.

could be the unit is not designed to make digital loopback like this, it needs the input to be the master clock.
in this case you try to lock on a signal that tries to lock on itself which could could give weird side effects.

Edit: I checked the specs of the U-DIO8 and it has sample rate converters on the inputs so it should be able to handle a digital loop.