Review and Measurements of Asus Xonar U7 MKII ADC/DAC/HP

[DWPress]

Well I've been trying to get all 8 channels recognized now for 3 days with no luck. I tried the Asus driver with their crapware, the Microsoft USB audio device driver that gets downloaded otherwise, ASIO4ALL and FlexASIO - none of them will give me all outputs, consistently I have no rears and routing is messed up. Then I started messing around with Equalizer APO and Peace pluggin to route channels and still no luck (yes I know I can't use it in conjunction with WASAPI driver) but it seems EAPO won't route channels unless Windows knows they're there first. I thought, "how hard can it be?" - obviously Windows makes it hard. I'm using Jriver for playback and tried to route channels there too, no luck.

I'm a Mac guy, then Linux and as a last resort Windows but I'm at a last resort because while fiddling with my miniDSP 4x10 I dropped a final screw onto a live board and got magic smoke. This is the true disadvantage of a pure fully active 3.2 active XO that is hardware based like miniDSP, I just got the tracking number today for my replacement board on the slow boat from Hong Kong. I had picked up this MKII for just such an experiment, upmix from 2 channel to 8 to run active XO fully through computer software and ditch the miniDSP and move the music server onto a dedicated Windows box.

If anyone has actually gotten Windows 10 to recognize all channels somehow and would like to help out consider this a plead for mercy - I've no music otherwise in my studio until the new board arrives.

 

[ernestcarl]

If anyone has actually gotten Windows 10 to recognize all channels somehow and would like to help out consider this a plead for mercy - I've no music otherwise in my studio until the new board arrives.

Just uninstall the Asus software as you really don't need it in most cases. I'm not sure why you can't route audio to JRiver's WDM driver. I assume you've enabled it and set is as your default output playback device in Windows.

Within JRiver, whichever zone you've selected, set your audio device to

And set your output format to 7.1

Enable the room correction plugin (set as desired), and use the parametric plugin(s) to switch/route channels around in if you need to...

Test the audio signal levels sent to each channel with the audio calibration tool

Play some MCH test tracks/videos and check the analyzer window

 

[trl]

Perhaps it's worth measuring how USB audio performance varies?

Amir did tested several USB-powered DACs and most of them didn't benefit of any performance increase when powered via an external low-noise power adapter. This is valid for ASUS U7 too; I had it tested when powered from the USB port vs. powered from Schiit Wyrd and the internal noise and harmonic profile remained constant. I had also tested ODAC v1 and I couldn't find any benefits when powered from Wyrd vs. the regular power port (tested with ARTA).

I am sure that some DACs, depending on their internal design, might benefit from an external low-noise power supply, but this is not a general rule.

However, I usually like to use the lowest noise possible USB port when I connect my DAC.

 

[Robbo99999]

It worked perfectly fine with ASIO4ALL in Windows 8.1 for me, first time, no fiddling about. However, I haven't tried it in 24 bit.
I think PCs are generally an unreliable platform for testing USB powered DACs.
Audio performance varies from one PC to another because the quality of the PSU generally varies also. The power source is crucial for audio performance.
Electromagnetic interference varies too.
As does memory latency and background CPU activity, and probably USB audio latency.
Laptops generally do not have great latency but at least they have the choice of using it's battery as a clean power source.
Perhaps it's worth measuring how USB audio performance varies?

My viewpoint is that "clean energy sources" for DACs are not that important for USB DACs - I think it's generally been shown on this site that good DACs will be able to handle a noisy USB power supply without adverse effects. And I think in general, power supply conditioners have also been shown to be a waste of time & money when talking about good quality audio equipment.

 

[DWPress]

I'm not sure why you can't route audio to JRiver's WDM driver

Thank you for your thorough answer! I have progress with the information you provided though now I'm ashamed to admit I've blown an expensive ribbon tweeter because of a channel mismatch between sides and rears and not being familiar with what controls what with volume control in Win10.

Not a very good week for audio here but now I get to figure out how to replace a ribbon since the driver isn't currently available in the US. This is despite having a blocking cap inline with the tweets too, guess it didn't like full 350-3650Hz 100 watts at full system volume. I just recently took my MSB MVC-1 out of my chain, if I hadn't it likely wouldn't have happened. Happy happy joy joy....

 

[ernestcarl]

Thank you for your thorough answer! I have progress with the information you provided though now I'm ashamed to admit I've blown an expensive ribbon tweeter because of a channel mismatch between sides and rears and not being familiar with what controls what with volume control in Win10.

Not a very good week for audio here but now I get to figure out how to replace a ribbon since the driver isn't currently available in the US. This is despite having a blocking cap inline with the tweets too, guess it didn't like full 350-3650Hz 100 watts at full system volume. I just recently took my MSB MVC-1 out of my chain, if I hadn't it likely wouldn't have happened. Happy happy joy joy....

Sorry to hear about that... that’s one of the reasons why I prefer ready-made active speakers which already have built-in limiters/protection circuitry and amps in place. Before calibrating the volume levels, I used to overload my monitors quite a bit — including bottoming out my sub many times... it’s a miracle that I have not destroyed a driver yet.

 

[Tasslehoff]

My viewpoint is that "clean energy sources" for DACs are not that important for USB DACs - I think it's generally been shown on this site that good DACs will be able to handle a noisy USB power supply without adverse effects. And I think in general, power supply conditioners have also been shown to be a waste of time & money when talking about good quality audio equipment.

Hi there.
I have tested many PSUs on the Arcam RDAC and without doubt there is a noticable audible difference in the analog output so it depends on the DAC.
However, we digress because we are talking about the Asus U7 MKIi here.

 

[trl]

I have also made cheap DACs sounds amazing at a very low cost.

It's a bit off-topic this one, but you might want to open up a dedicated thread for this and show up what exact modifications and improvements you've done and to what DACs exactly.

Also, some sort of measurements are a must, otherwise the subjective audio quality is not important anymore. What's important is the DAC to introduce near zero analogue and digital artefacts (lowest jitter, THD+N and IMD distortions possible).

 

[Tasslehoff]

It's a bit off-topic this one, but you might want to open up a dedicated thread for this and show up what exact modifications and improvements you've done and to what DACs exactly.

Also, some sort of measurements are a must, otherwise the subjective audio quality is not important anymore. What's important is the DAC to introduce near zero analogue and digital artefacts (lowest jitter, THD+N and IMD distortions possible).

I believe the Arcam RDAC, Dragonfly, Dacmagic etc has better performance with a better PSU which has been measured, but may be not here.
You are right to say my comment is somewhat subjective. I will delete that comment which is subjective.
I will hopefully open up a new thread up at some point with some measurements if I can measure the differences.

 

[Geoff-AU]

I once owned this for 2.5 years. I was really disappointed that after 2.5 year(warranty is 2 years), this didn't start and just blinked the blue LED and cliking. The short solution was pull out the usb cable and put a usb cable which was connected to a usb power. I tried it on various pcs so it was not a undervoltage problem of my pc. Also you can find so many reviews with this problems on amazon...

PSA: This is a known fault with Xonar U7 cards. It has been well documented:

I bought 5 x U7 for a project. After 18 months, 2 of them were dead. I found that heating the CM6632 temporarily brought it to life (until next power cycle), so I replaced all 5 cards with new CM6632 ICs. After another 18 months, they are dying again. I contacted Asus, and Asus said there was no fix that can be done.

So, don't waste your time with these cards. They will just fail on you after a short time.

 

[Robbo99999]

I was lucky with Asus Xonar U7 Mki....I used it everyday from 2014 - end of 2019, no issues, I guess some are ok, but yes I've heard of these problems with U7 cards dying.

 

[MRP]

PSA: This is a known fault with Xonar U7 cards. It has been well documented:


I bought 5 x U7 for a project. After 18 months, 2 of them were dead. I found that heating the CM6632 temporarily brought it to life (until next power cycle), so I replaced all 5 cards with new CM6632 ICs. After another 18 months, they are dying again. I contacted Asus, and Asus said there was no fix that can be done.

So, don't waste your time with these cards. They will just fail on you after a short time.

I don't know if I'm part of a lucky minority but my Mk I is still going strong after 4 years. I use it for both my notebook and desktop PCs over the years. I even accidentally dropped it some time ago. The Mk I works just fine with default Windows driver. I think I'll keep using it as a DAC until it throws down the towel.

 

[JohnYang1997]

No issue after 3 years. Mk1

 

[DWPress]

I'm happy to report that I did indeed get this card working with WIN10 and JRiver as an active XO with convolution. In fact, I liked it so much that I've ordered a Motu Ultralite-MK3 so that I can do it on my Mac as well. I've since received my miniDSP 4x10 replacement board and when I got it all hooked up again and back in the system, subjectively, it didn't sound as good. I've always suspected the miniDSP to color the sound somewhat even though I send it a purely digital signal bypassing ADC, maybe it's their implementation of all processing at 24/92 I really don't know but since this $60 Asus card somehow (subjectively) sounded better and I've learned some new skills I'm going to remove the 4x10 and do software XO and correction. I'll be moving the mini PC with Asus card into the house where I'll reimplement it for an active system there.

 

[ernestcarl]

I'm happy to report that I did indeed get this card working with WIN10 and JRiver as an active XO with convolution. In fact, I liked it so much that I've ordered a Motu Ultralite-MK3 so that I can do it on my Mac as well. I've since received my miniDSP 4x10 replacement board and when I got it all hooked up again and back in the system, subjectively, it didn't sound as good. I've always suspected the miniDSP to color the sound somewhat even though I send it a purely digital signal bypassing ADC, maybe it's their implementation of all processing at 24/92 I really don't know but since this $60 Asus card somehow (subjectively) sounded better and I've learned some new skills I'm going to remove the 4x10 and do software XO and correction. I'll be moving the mini PC with Asus card into the house where I'll reimplement it for an active system there.

Can’t say I noticed an audible difference myself other than noise floor changing may be from the difference in output VRMS(?) — just comparing with the 2x4 HD

Can’t understand why miniDSP 4x10 or 10x10 HD does not allow the use of it's USB input for ASIO or WASAPI, instead it seems that it is rather just for programming/controlling the device, right?

*Oh, yeah, you mentioned that you use convolution (I don't) so maybe the difference lies there... BTW -- you probably know this already -- but JRiver's default Room Correction plugin has some serious limitations as to the crossovers it allows for one to use, which can be rather frustrating. This can be bypassed by not using it, and instead manually adding xo filters in the PEQ plugins' section -- it's a bit more work and can be rather confusing, though, since there are no 'visualized graphics' as a cue to all filters enabled -- as in the case with something like a miniDSP -- just a very, very long list of filters.

 

[DeepFried]

I had an original xonar U7. it just up and died after a few years, was never particularly fond of it for the price. it was useful while I had a 3.5mm microphone but once I got a USB mic it just went in a draw. No toslink in or out was also a fairly big negative for me.

Oh and now i think about it it had this odd issue where it would need to be plugged in for several hours to 'warm up' before it would work. it was fine if you always had it plugged in but if you let it get cold it would just appear dead. I guess it was on the way out for a while before it finally actually died.

 

[trl]

No toslink in or out was also a fairly big negative for me.

But it has TOSLINK and COAX out. U7 can wirk as an USB transport this way too.

 

[DWPress]

*Oh, yeah, you mentioned that you use convolution (I don't) so maybe the difference lies there... BTW -- you probably know this already -- but JRiver's default Room Correction plugin has some serious limitations as to the crossovers it allows for one to use, which can be rather frustrating. This can be bypassed by not using it, and instead manually adding xo filters in the PEQ plugins' section

I did set up the XO, channel routing and some EQ in the JRiver's PEQ section not using the Room Correction tab at all and yes, it did take a little figuring out to get the order correct. I then used the exact same convolution filter I had been using with JRiver on my Mac.

As I said, my observations are purely subjective as it would be considerable work to A/B test 8 channels. I've got my new miniDSP 4x10 board all tuned back in again and it sounds fine to my ears but the Motu is on it's way and I'll be happy to try the same experiment on the Mac version of JRiver to see (hear) what's going on. A couple differences are 1) I believe I read that JRiver has a slightly different "Q" equation and 2) I was also stacking two BW2 filters to achieve LR4 XO points rather than miniDSP's straight up LR4. I'll be using the Motu just for channel routing limiting any other hardware signal processing and see if I can replicate what I thought sounded so good.

 

[DeepFried]

But it has TOSLINK and COAX out. U7 can wirk as an USB transport this way too.

it has coax. where is the toslink?

 

[ernestcarl]

1) I believe I read that JRiver has a slightly different "Q" equation

Yeah, from what I understand, many equalizer hardware/software don't really exactly use the same methods of calculating the Q. Also, I get slight differences between the default LR4 xo in JRiver and in miniDSP -- probably in the phase (?)... dunno, but, my basic filters between the two don't 100% match so I can't exactly interchange the exact same filters I used between these two. Oh, well...