EVGA Nu Audio

[rickca]

 

5 hours ago, Samuel T Cogley said:

 

Yes, "device disappears" is what I'm calling a card crash.

I've opened a support ticket with EVGA.  I tried to change the HQPlayer settings after playing part of a track converted from redbook to DSD256 and the Nu Audio disappeared from Control Panel playback devices.  I'll test some more but I think maybe just stopping HQPlayer (without trying to change the settings) might be sufficient to cause this problem.

 

This is a real bummer because I was planning to evaluate HQPlayer Desktop 4 with this card.  Obviously, needing to reboot after changing settings makes it a very unproductive exercise.  I guess I'm going to miss the Cyber Monday 15% discount.  I don't have another DAC that does DSD.

[Samuel T Cogley]

59 minutes ago, rickca said:

 

I've opened a support ticket with EVGA.  I tried to change the HQPlayer settings after playing part of a track converted from redbook to DSD256 and the Nu Audio disappeared from Control Panel playback devices.  I'll test some more but I think maybe just stopping HQPlayer (without trying to change the settings) might be sufficient to cause this problem.

 

This is a real bummer because I was planning to evaluate HQPlayer Desktop 4 with this card.  Obviously, needing to reboot after changing settings makes it a very unproductive exercise.  I guess I'm going to miss the Cyber Monday 15% discount.  I don't have another DAC that does DSD.

 

This might be a touch off-topic, but you can pick up a Topping D50S for $249.99 on Amazon.  I saw a lower price, but it looked like shipping from China that was going to take weeks.

 

I had a great time trying different HQPlayer settings with an iFI iDAC2 (via USB).  As much as I think the Nu Audio is a great PC sound card, it doesn't rise to the level of audiophile kit IMHO.  But having said all this, thank you for opening that support ticket.  Hopefully @EVGA_Lee is watching the thread and can offer some feedback on how they're testing for smooth transition from PCM ->DSD -> PCM modes.

 

I hope they haven't forgotten about this card because of the NU Audio Pro 7.1.  I'd like to see this get fixed.

[rickca]

OK I have some better input for @EVGA_Lee and @Miska.

 

I rebooted, then brought up HQPlayer with default output mode PCM.  Then I simply changed the default output mode to DSD and without even starting to play a track the audio icon in the lower right corner of the taskbar got a red X and Nu Audio disappears from Control Panel playback devices.  Interestingly, the Nu Audio GUI also immediately changed to ASIO active (DSD mode) even though I wasn't playing a track!  With PCM, I don't think the Nu Audio GUI shows ASIO active until something is playing.

 

So now I can't even play a single track in DSD mode.  Stopping HQPlayer or quitting HQPlayer doesn't bring Nu Audio back to life.  I have to reboot.

 

After a reboot, if I bring up HQPlayer with default output mode DSD, the same things happen and I can't bring the Nu Audio back to life by changing HQPlayer default output mode to PCM.  I have to reboot.

 

@Samuel T Cogley I wasn't looking for audiophile nirvana with this card, just a simple way to try out HQPlayer 4 Desktop and see whether my machine could handle the new EC modulators.  I have a fantastic DAC on my main system, but it is PCM only.

 

Something is really broken.

[Samuel T Cogley]

16 hours ago, rickca said:

 

@Samuel T Cogley I wasn't looking for audiophile nirvana with this card, just a simple way to try out HQPlayer 4 Desktop and see whether my machine could handle the new EC modulators.  I have a fantastic DAC on my main system, but it is PCM only.

 

 

My primary DAC is a first gen Yiggy.  But HQPlayer first got my attention when I tried it with an iFI iDSD Nano.  The primary reason I wanted this card is that I have a pair of gaming headphones that are just HD598's with an inline microphone.  The NU Audio card drives them very well.  I had another issue with in-game audio that's so specific, I find it difficult to believe it's the card.  I hope EVGA sorts the card crash issues.

[EVGA_Lee]

Thanks everyone, I'll forward this to our Audio team.  As you might imagine, both I and our Audio team were on break over the holiday.  Generally, I'd suggest everyone report these kinds of issues to me (of the more complicated variety, that is), as our Customer Service will probably just forward them to me anyway.

 

Yes, we will continue to support the original NU Audio card.  There is probably a higher focus on tuning the drivers for the Pro cards prior to shipping the cards at this moment, but we are still working on other bugs/issues with the current NU Audio card.

 

For Miska, I'll see if they can possibly work on a firmware updater for Linux, but Linux really isn't our strong suit when it comes to driver support.

[desponder]

18 minutes ago, EVGA_Lee said:

 

 

For Miska, I'll see if they can possibly work on a firmware updater for Linux, but Linux really isn't our strong suit when it comes to driver support.

 Oh, PLEASE, please do get a Linux firmware updater!!!!

[the_bat]

As a complete aside, the designer has been measuring various different op-amps with the nu-audio.  He's given me permission to post them here, but they may not stay up indefinitely.

 

https://drive.google.com/open?id=15fqZMLE6-zQNmPdvqXZNm6_hm8Osl3bq

 

Alan

 

 

[EVGA_Lee]

7 hours ago, the_bat said:

As a complete aside, the designer has been measuring various different op-amps with the nu-audio.  He's given me permission to post them here, but they may not stay up indefinitely.

 

https://drive.google.com/open?id=15fqZMLE6-zQNmPdvqXZNm6_hm8Osl3bq

 

Alan

 

 

I can confirm this.  I plan to put an FAQ up at the EVGA forums for op-amps, but time has not been friendly, lately.  I have these downloaded if the drive link no longer works.

 

Please note that these measurements were taken with the NU Audio Pro card, not the original NU Audio Card.

[rickca]

On 12/2/2019 at 6:34 PM, rickca said:

Something is really broken.

EVGA has been able to reproduce the DSD issue I described in post #103 and they're working on a fix.

[EVGA_Lee]

Official driver posted today that should fix the crash issue after DSD playback:  https://forums.evga.com/NU-Audio-Driver-Release-v0215-Win10-Fixes-Crash-after-DSD-Playback-m3015237.aspx

 

There is a firmware update, so you'll have to switch to a Windows PC briefly, if you're using Linux.

 

With regards to Linux firmware updating, I'm afraid it's not possible at this time. The short answer involves custom drivers from Thesycon and related firmware, which (as a side-effect) prevents the card from being recognized in Linux.  I don't know if that means it can never be done, but right now it isn't something on the roadmap, unfortunately.

[desponder]

That seems strange; particularly as Thesyon lists themselves as being a LINUX device/driver development ¨shop¨.

https://www.thesycon.de/eng/company_overview.shtml

 

 Alas, Win-blows stuff likely allows better mining of the monetary mint, as it were! 😉

Sort of like RME etc....

But we may have thoughts to hope ....

[rickca]

On 12/11/2019 at 7:33 AM, rickca said:

EVGA has been able to reproduce the DSD issue I described in post #103 and they're working on a fix.

@EVGA_Lee and @Miska

 

OK now we're making some progress.  With the latest driver and firmware 0.31, the Nu Audio card is released if you are playing DSD and bring up the HQPlayer settings.

 

The curious thing is that pressing Stop does not release the card, but bring up Settings does release it.  @Miska can you explain why this happens?

[Miska]

16 minutes ago, rickca said:

@EVGA_Lee and @Miska

 

OK now we're making some progress.  With the latest driver and firmware 0.31, the Nu Audio card is released if you are playing DSD and bring up the HQPlayer settings.

 

The curious thing is that pressing Stop does not release the card, but bring up Settings does release it.  @Miska can you explain why this happens?

 

This is working exactly as designed. HQPlayer keeps the card reserved as long as it is running, except for the settings dialog where you are selecting the device. So no other applications can come and steal the device or mess with it's state. ASIO and WASAPI Exclusive mode do exactly this, they allow exclusive control of the audio device.

 

[Samuel T Cogley]

It seems this latest update has resolved the "card crash" issue.  The only issue I'm still having is that one HQPlayer stutter stops playback.  With a USB device in the same stutter scenario, playback continues after a stutter.

[Miska]

5 hours ago, Samuel T Cogley said:

It seems this latest update has resolved the "card crash" issue.  The only issue I'm still having is that one HQPlayer stutter stops playback.  With a USB device in the same stutter scenario, playback continues after a stutter.

 

Did the driver control panel have buffer size adjustment? Setting it for bigger/safer buffers likely resolves that. If it doesn't you can try to adjust it from HQPlayer side instead, by setting "Buffer time" to 100 ms for example.

 

Stutter should not be visible to the driver and high CPU load shouldn't be causing dropouts in data flow to the driver. HQPlayer is always providing data immediately to the driver when asked. This sounds like the hardware buffers are having underflow under such heavy load circumstances, which is driver's internal business. Bigger hardwarebuffers should help on this.

 

I have not seen such behavior with other ASIO drivers, if the driver is from Thesycon I'm a little surprised such would happen because Thesycon drivers are widely used.

 

You could check that HQPlayer log file is enabled, restart HQPlayer, reproduce this issue and then send the log file to me over email. There are likely some hints about the problems...

 

[Samuel T Cogley]

1 minute ago, Miska said:

 

Did the driver control panel have buffer size adjustment? Setting it for bigger/safer buffers likely resolves that. If it doesn't you can try to adjust it from HQPlayer side instead, by setting "Buffer time" to 100 ms for example.

 

Stutter should not be visible to the driver and high CPU load shouldn't be causing dropouts in data flow to the driver. HQPlayer is always providing data immediately to the driver when asked. This sounds like the hardware buffers are having underflow under such heavy load circumstances, which is driver's internal business. Bigger hardwarebuffers should help on this.

 

I have not seen such behavior with other ASIO drivers, if the driver is from Thesycon I'm a little surprised such would happen because Thesycon drivers are widely used.

 

You could check that HQPlayer log file is enabled, restart HQPlayer, reproduce this issue and then send the log file to me over email. There are likely some hints about the problems...

 

 

Thank you.  I will investigate the things you highlight.

[MhtLion]

On 12/13/2019 at 1:41 PM, EVGA_Lee said:

Official driver posted today that should fix the crash issue after DSD playback:  https://forums.evga.com/NU-Audio-Driver-Release-v0215-Win10-Fixes-Crash-after-DSD-Playback-m3015237.aspx

 

There is a firmware update, so you'll have to switch to a Windows PC briefly, if you're using Linux.

 

With regards to Linux firmware updating, I'm afraid it's not possible at this time. The short answer involves custom drivers from Thesycon and related firmware, which (as a side-effect) prevents the card from being recognized in Linux.  I don't know if that means it can never be done, but right now it isn't something on the roadmap, unfortunately.

Great to see official EVGA support is here.  Here is what I have observed, which I hope you can rely to the development team.  The card clearly targeted for the audiophile community, but it was tested by none-audiophile people or someone with a very limited audiophile usage pattern of the card. Perhaps, the card was tested by non-audiophile engineers who do not have any knowledge about various usage pattern of usage of an audiophile sound card. 

 

I hope QA process going forward will include the audiophile like people in this forum - well they are the best, especially Miska, aren't they?  Or, just send the card to Miska prior to release under certain arrangement. I wholeheartedly recommend it. I work with IT dept, so I can 'guess' at least 3 major problems in EVGA's QA process with this card.

[desponder]

2 hours ago, MhtLion said:

... but it was tested by none-audiophile people or someone with a very limited audiophile usage pattern of the card. Perhaps, the card was tested by non-audiophile engineers who do not have any knowledge about various usage pattern of usage of an audiophile sound card.

 

 Just as a point of curiosity and to ¨clear the air¨ ; what exactly IS ¨audiophile usage pattern¨? Similarly, what specific verifiable knowledge of music do these august folk exhibit? I of course do not refer to electronics acumen, rather one intends knowledge of real live breathing non-electronic music; as none less can remotely be considered valid aurally.

 One does assume of course that Audio Note designers did indeed ¨eat their own cooking¨ before passing the project on for EVGA to produce. ;-)! 

 

 

 

[MhtLion]

1 hour ago, desponder said:

One does assume of course that Audio Note designers did indeed ¨eat their own cooking¨ before passing the project on for EVGA to produce. ;-)! 

I believe not. If they did test thoroughly, it would have played nicely with all major players using various music format. It's safe to assume they did not test it thoroughly - otherwise it won't had all these crashes and bugs.

 

Give it to an engineer who only used itune and WMP in MP3/wav files entire his life, he would said - Yes it's working perfectly fine. Give to this community like here, where we use Roon, HQP, Qobuz, Nirvana, Jriver, Linux, ASIO, WASAPI, DSD 64, 128, 256, 512, they would had ironed out the bugs before the release.

 

Also, Audio Note has nothing to do here. Audio Note sold transistors to EVGA and let EVGA use its brand name - nothing else.

[the_bat]

Wrong on just about every count. 

[MhtLion]

What's wrong?

[the_bat]

What's wrong is that you're stating as facts things on which you have manifestly no knowledge, that are completely untrue but that could damage the reputations of both companies.  As examples

 

On 12/18/2019 at 7:36 PM, MhtLion said:

Also, Audio Note has nothing to do here. Audio Note sold transistors to EVGA and let EVGA use its brand name - nothing else.

 

You couldn't be more wrong.

1. Audio Note were heavily involved in the design, component selection and testing; and

2. They don't make transistors.

[MhtLion]

59 minutes ago, the_bat said:

What's wrong is that you're stating as facts things on which you have manifestly no knowledge, that are completely untrue but that could damage the reputations of both companies.  As examples

 

 

You couldn't be more wrong.

1. Audio Note were heavily involved in the design, component selection and testing; and

2. They don't make transistors.

Transistor vs resistor - it's the same thing. Per EVGA website, they used Audio Note resistors and caps. https://www.evga.com/products/product.aspx?pn=712-P1-AN01-KR

 

Do you work for EVGA or Audio Note?  Audio Note - however awesome they are, they are not responsible for this card crashing. Sure, they are to have provided input on the sound and analog components such as the transistor - but we are not talking about the sound here. We are talking about the bugginess of the driver/firmware, which is solely responsible of EVGA in my humble opinion. So, will you please tell me what else I 'couldn't be more wrong' and what you know about their 'heavy involvement'?

 

[Samuel T Cogley]

On 12/16/2019 at 2:15 PM, Miska said:

You could check that HQPlayer log file is enabled, restart HQPlayer, reproduce this issue and then send the log file to me over email. There are likely some hints about the problems...

 

Just sent you a PM with the log file attached.  One stutter can halt HQPlayer.  But USB DACs don't seem to suffer from this issue.

[tigerdog]

I've read through the forum and found several mentions of a Linux driver, but I cannot find a link to the driver itself.  Would someone be so kind as to re-post here? 
 

Many thanks in advance!