MP 1,1-5,1 Strange Audio Issue on MP 4,1>5,1 Mojave 10.14.4

[bsbeamer]

Did you try line out and headphones? That seemed to be where people were complaining the most.

I have not personally been able to trigger on my 5,1 machine, or with single tray swapped back in. Asked clients with 4,1 (don’t know processors off hand) and none had this issue. Highly doubt they’re on E5520 however.

 

[tsialex]

Did you try line out and headphones? That seemed to be where people were complaining the most.

I have not personally been able to trigger on my 5,1 machine, or with single tray swapped back in. Asked clients with 4,1 (don’t know processors off hand) and none had this issue. Highly doubt they’re on E5520 however.

I use line-out to a Creative speaker set with subwoofer.

 

[bsbeamer]

Maybe the mismatched SMC or 142 firmware addresses somehow?

 

[tsialex]

Maybe the mismatched SMC or 142 firmware addresses somehow?

I doubt. I can downgrade to 141.0.0.0.0 and check again. I was going to downgrade to test my W3540. Just tested with iTunes Movies, used Hugo and Harry Potter 1, no audio problems.

 

[bsbeamer]

Everyone who have reported issues: please confirm your BootROM version and both SMC versions. Screenshots would be great.
[doublepost=1555030654][/doublepost]Issue was only reported with 5500-series so assume W3540 fine (same series in official 5,1).

 

[tsialex]

Maybe the mismatched SMC or 142 firmware addresses somehow?

Everyone who have reported issues: please confirm your BootROM version and both SMC versions. Screenshots would be great.
[doublepost=1555030654][/doublepost]Issue was only reported with 5500-series so assume W3540 fine (same series in official 5,1).

I have a X/L/E55xx something in my Dell R410, but I’m not going to take it down on a weekday.

 

[bsbeamer]

From screenshots that machine looks to be authentic 5,1 with 4,1 tray... in case it matters at all, the only reports of issues are with 4,1>5,1 hack/update to ID as 5,1 with 5500-series processors. I 100% assume they are 1.39f5 for both system and tray. Also assume they’re on 141 firmware at most. No one should be experimenting with 142 except people like you.

 

[tsialex]

From screenshots that machine looks to be authentic 5,1 with 4,1 tray... in case it matters at all, the only reports of issues are with 4,1>5,1 hack/update to ID as 5,1 with 5500-series processors. I 100% assume they are 1.39f5 for both system and tray. Also assume they’re on 141 firmware at most. No one should be experimenting with 142 except people like you.

I’m disabling SIP to downgrade to 141.0.0.0.0. I’ll check everything again with 141.0.0.0.0.

I can check with a 1.39f5 Mac Pro later.

 

[tsialex]

Downgraded to 141.0.0.0.0, still no audio problem.



[doublepost=1555032842][/doublepost]No problem with IINA playing .flac:


[doublepost=1555033155][/doublepost]Or with iTunes playing MP3.

 

[bsbeamer]

Since it appears no one who reported issues is checking in right now, have you tried with local (non streaming) audio and video files? Assume QT or iTunes as media player?

I tested with 6K RED R3D on MP5,1 dual X5690 with embedded audio and no issues. Same with HEVC UHD/4K and same with 4K MP4 same with 1080HD and straight audio varieties. Concluded no issue on my machine.
[doublepost=1555033373][/doublepost]Wouldn’t pursue the mismatched SMC testing further. Clearly there is no issue. If there is anything, it’s more likely a bad file or bad codec issue. You’ve exhausted this and have not observed the reported issue.

 

[Riff_Al]

Anyone has a link for something that for sure triggers the problem?

In my case playback of any audio file (mp3 or audio cd, i didn't check any on-line audio, YouTube or videofiles) is accompanied by stuttering sound on 10.14.4
Dozens clean install of 10.14.4 (installer run from another macOS volume after erasing the target volume - because of GTX680... you know what i mean) on different SSD or HDD, different sound sources - nothing has a positive effect. Sound is stuttering on 10.14.4
On 10.14.5 pb5 stuttering is much less, but still heard if listen attentively.
On 10.14.3 - everything is great as it was on previous macOS.

 

[tsialex]

In my case playback of any audio file (mp3 or audio cd, i didn't check any on-line audio, YouTube or videofiles) is accompanied by stuttering sound on 10.14.4
Dozens clean install of 10.14.4 (installer run from another macOS volume after erasing the target volume - because of GTX680... you know what i mean) on different SSD or HDD, different sound sources - nothing has a positive effect. Sound is stuttering on 10.14.4
On 10.14.5 pb5 stuttering is much less, but still heard if listen attentively.
On 10.14.3 - everything is great as it was on previous macOS.

What's the frequency that you hear the stuttering? It's constant, happens at frequent intervals or it's something rare?

 

[Riff_Al]

What's the frequency that you hear the stuttering? It's constant, happens at frequent intervals or it's something rare?

On 10.14.4 stuttering is unevenly. Could be every 3 or 4 sec, then 10-15 seconds without issue, and then every 1 second...
On 10.14.5 much less, about one a minute.
I will try to install 10.14.4 again, play file and record the sound.
Also, sound stuttering is terrible when play in iTunes. With QuickPlayer or Vox Player sound issue is less.
And with iTunes (10.14.4) i observed a strange thing - if i try to close it, sometimes it is impossible with regular way, and have to force quit application.
On 10.14.3 and 10.14.5 everything is fine with it.

 

[Choppy23]

I just tested a clean install of 10.14.4 with a dual E5520 tray. No audio problems with:

• Netflix + Firefox
• Youtube + Safari or Firefox
• IINA + 1080P h264 videos
  
• iTunes.

View attachment 831705

Anyone has a link for something that for sure triggers the problem?
[doublepost=1555029429][/doublepost] View attachment 831706
[doublepost=1555029646][/doublepost]This is not a Gainestown problem…

Hi tsiale,

I'm having the issue, and can confirm what you said: everything works great with Netflix, youtube, iTunes, Spotify, etc. The issue (for me) is with midi triggering audio samples.

I have Xeon W5590 dual processors on a 4,1 > 5,1.
Boot ROM Version: 140.0.0.0.0
SMC Version: 1.39f5

The easiest way to reproduce this on a problem machine with software that everyone has or can get for free:
1. Launch Garage Band
2. Select the Factory Template named “Keyboard Collection”
3. Select the instrument “80’s Sync Lead”
4. Press a note on midi controller and hold for 10 seconds. You will hear glitches every few seconds.

 

[dkboh]

Following this thread. I’m having this exact issue.

Mac Pro 4,1>5,1 with dual X5570

Boot Rom: 141.0.0.0.0
SMC Version: 1.39f5

EDIT: After watching this thread, really seems like upgrading the processor is the way to go, just to be future-proof. I’m ordering a pair of X5680’s.

 

[Kaspin]

I got my X5690s installed and it seems the audio issue is fixed

I made a video to demonstrate the issue as I heard it, using a Loop demo in Logic Pro X:

 

[Sg335man]

I just tested a clean install of 10.14.4 with a dual E5520 tray. No audio problems with:

• Netflix + Firefox
• Youtube + Safari or Firefox
• IINA + 1080P h264 videos
  
• iTunes.

View attachment 831705

Anyone has a link for something that for sure triggers the problem?
[doublepost=1555029429][/doublepost] View attachment 831706
[doublepost=1555029646][/doublepost]This is not a Gainestown problem…

Yeah I installed 2 x 5690 3.46 chips on my 2009 MP 4,1 now 5,1 the audio problems have gone?

 

[bsbeamer]

Yeah I installed 2 x 5690 3.46 chips on my 2009 MP 4,1 now 5,1 the audio problems have gone?

I assume you mean dual X5690 (2 x X5690). If so, these are not impacted by this bug/issue. Only reported impacts on Xeon 5500-series GAINESTOWN processors.
[doublepost=1555073284][/doublepost]

The easiest way to reproduce this on a problem machine with software that everyone has or can get for free:
1. Launch Garage Band
2. Select the Factory Template named “Keyboard Collection”
3. Select the instrument “80’s Sync Lead”
4. Press a note on midi controller and hold for 10 seconds. You will hear glitches every few seconds.

Step 3B - Window > Show Keyboard
(most here likely do not have MIDI controller)

No problems with this test on authentic MP5,1 with X5690 and 141 firmware. Also no issues with MP3 in iTunes or Quicktime. Previous tests were fine with Adobe Audition CC 2019 and Premiere Pro CC 2019 with both video and audio files of various formats.

 

[stpmusic]

So my first check was with iTunes, Youtube - had no issues on E5520 @10.14.5 beta 2, 142 bootrom

Them I've added my MOTU Firewire and installed a couple of audio-apps to test.

So now I can confirm:
For me everything is stuttering @44,1kHz when Garageband is running. Garageband switches the audio-driver to 44,1kHz
If I shut down Garageband it stays at 44,1kHz and even iTunes is stuttering.
When I manually switch back my sound card to 48kHz and hit play in iTunes, stuttering is gone.
If I switch MOTU-Audio to 96kHz stuttering gets worst.

Youtube on MOTU always stuttering
Youtube on internal sound no stuttering

Tried the same in Steinbergs Wavelab on 44,1kHz an 48kHz - other kind of stuttering - sounds different to Garageband

Over all it sounds for me like wrong word clock on audio busses, but I cannot imagine how this should depend on a processor

 

[bsbeamer]

@stpmusic issue you are having/experiencing has nothing to do with a source file's sample rate, format or bit depth - correct?

If that is the case, easiest test for most would be Applications > Utilities > Audio MIDI Setup and change format for output devices. Can you precisely say which output setup exactly you are using and experiencing the issue? 16-bit, 20-bit and 24-bit or has no impact?



Basically trying to have multiple people recreate this issue independently. If it's reproducible there is an issue. The semi-sporadic nature combined with audio terms many do not understand generally makes this a more difficult issue to troubleshoot.

 

[Choppy23]

So my first check was with iTunes, Youtube - had no issues on E5520 @10.14.5 beta 2, 142 bootrom

Them I've added my MOTU Firewire and installed a couple of audio-apps to test.

So now I can confirm:
For me everything is stuttering @44,1kHz when Garageband is running. Garageband switches the audio-driver to 44,1kHz
If I shut down Garageband it stays at 44,1kHz and even iTunes is stuttering.
When I manually switch back my sound card to 48kHz and hit play in iTunes, stuttering is gone.
If I switch MOTU-Audio to 96kHz stuttering gets worst.

Youtube on MOTU always stuttering
Youtube on internal sound no stuttering

Tried the same in Steinbergs Wavelab on 44,1kHz an 48kHz - other kind of stuttering - sounds different to Garageband

Over all it sounds for me like wrong word clock on audio busses, but I cannot imagine how this should depend on a processor

To add to this, if I switch the latency from 256 samples to 2048 samples the issue is much improved. But you can't work at 2048, and there's still glitches. Just less of them.

Along side your report about the @96 kHz getting worse, it makes me feel like the processor can't keep up and would still indeed be related to the 5500 series processors.
[doublepost=1555082287][/doublepost]

@stpmusic issue you are having/experiencing has nothing to do with a source file's sample rate, format or bit depth - correct?

If that is the case, easiest test for most would be Applications > Utilities > Audio MIDI Setup and change format for output devices. Can you precisely say which output setup exactly you are using and experiencing the issue? 16-bit, 20-bit and 24-bit or has no impact?

View attachment 831778

Basically trying to have multiple people recreate this issue independently. If it's reproducible there is an issue. The semi-sporadic nature combined with audio terms many do not understand generally makes this a more difficult issue to troubleshoot.

I have the issue on all of these settings. It's terrible and unlistenable, glitching multiple times per second at 96 kHz, while glitching only every few seconds at 44.1 kHz. Basically the more work the processor has to do the worse it gets. Which is why increasing latency to a max setting of 2048 samples helps. At lower latency, the CPU has to cycle many more times to play the audio.

I should also add that in Apple's release notes for Mojave 10.14.4, they fixed another audio issue, which is when these problems started for us:
"Improves the reliability of USB audio devices when used with the MacBook Air, MacBook Pro and Mac mini models introduced in 2018"
https://www.macrumors.com/2019/03/25/apple-releases-macos-mojave-10-14-4/

 

[rocksarntpeople]

I dont have anymore info to add other than my particular scenario: Mac pro 4,1 > 5,1 convert, original E5520 chips had discernible clarity issues on all channels (line-out, DP, bluetooth and all tested frequencies and applications to some degree). After upgrading to X5680's the problem is no longer present. No application or system changes (other than processor replacement and PRAM/SMC reset which was tried with E5520 chips with no effect).

Base system was tested on 2 different drives with 2 fresh installations with no changes.

Changing the processors was the only thing that had ANY effect on the issue whatsoever

**also I can attest to different sample rates changing the severity of the issue when the 5520's were installed but never eliminating it as well**

Also as requested, Bootrom and SMC remained the same for both processor sets at:

Boot ROM Version: 140.0.0.0.0
SMC Version: 1.39f5

 

[stpmusic]

@stpmusic issue you are having/experiencing has nothing to do with a source file's sample rate, format or bit depth - correct?

If that is the case, easiest test for most would be Applications > Utilities > Audio MIDI Setup and change format for output devices. Can you precisely say which output setup exactly you are using and experiencing the issue? 16-bit, 20-bit and 24-bit or has no impact?

bitrate has no impact - I´ve checked that before...

Along side your report about the @96 kHz getting worse, it makes me feel like the processor can't keep up and would still indeed be related to the 5500 series processors.

Doesn´t make ME feel that. You don't need much processing to get an audio stream working. I've build audio workstations since 1996 and issues like that I've never had related to the processor. Most of the time something like that was related to bad IRQ-handling or bad drivers.

Maybe here is a key: 10.14.4: Apple „Improves the reliability of USB audio devices when used with MacBook Air, MacBook Pro and Mac mini models introduced in 2018”. And something got lost...

 

[bsbeamer]

Apple also addressed a blown speaker issue on MBP models in updates recently. Apple also added DP and HDMI audio output, including for many AMD GPUs used in Mojave for Mac Pro 5,1. Also improved audio quality of audio recordings (basically voice notes) and moved to a better quality format. And as always, "bug fixes and improvements, improves the stability and reliability of your Mac" which usually translates to security fixes/patches.

So for anyone "stress-testing" to attempt to force the issue, change Audio MIDI Setup to 96kHz output to (possibly) experience more frequently/reliably.

With the reports of changing from GAINESTOWN Xeon 5500-series to anything in the 5600-series fixing the issue, it's pretty clear the processor is somehow involved here...

 

[rocksarntpeople]

Apple also addressed a blown speaker issue on MBP models in updates recently. Apple also added DP and HDMI audio output, including for many AMD GPUs used in Mojave for Mac Pro 5,1. Also improved audio quality of audio recordings (basically voice notes) and moved to a better quality format. And as always, "bug fixes and improvements, improves the stability and reliability of your Mac" which usually translates to security fixes/patches.

So for anyone "stress-testing" to attempt to force the issue, change Audio MIDI Setup to 96kHz output to (possibly) experience more frequently/reliably.

With the reports of changing from GAINESTOWN Xeon 5500-series to anything in the 5600-series fixing the issue, it's pretty clear the processor is somehow involved here...

I had held off on moving my crucial SSD to an OWC PCIE blade until I was sure this issue was taken care of, but I will be doing that today so the machine will be getting more testing. If there are any KNOWN scenarios that heavily cause the issue on the older chips I can put that exact scenario to the test on the new chips this afternoon.