Got a tip for us?
Let us know
Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
MP 1,1-5,1 Strange Audio Issue on MP 4,1>5,1 Mojave 10.14.4
- Thread starter rocksarntpeople
- Start date
- Sort by reaction score
I can CONFIRM that with X5680s the Audio issue is no longer present on the consumer version of 10.14.4...will update the first post with this info as well
[doublepost=1554937563][/doublepost]Also as an aside, kinda cool to see my first post ever on this site make it into a sticky thread...thanks guys for all the help. Not sure if the core issue will be resolved in 10.14.5, but for now anyone running the E5520s at least should probably stay away from 10.14.4 if they care about audio
[doublepost=1554937563][/doublepost]Also as an aside, kinda cool to see my first post ever on this site make it into a sticky thread...thanks guys for all the help. Not sure if the core issue will be resolved in 10.14.5, but for now anyone running the E5520s at least should probably stay away from 10.14.4 if they care about audio
I am having this same issue. I have a MacPro 5,1. Intel Xeon W5590 Dual CPU 3.33 GHz processors. Graphics card: Radeon RX 580 8 GB.
Everything was great on Mojave 10.14.3. Within minutes of updating to 10.14.4 my issues started.
I am in the Apple development program. I installed the 10.14.5 public beta 2 and the problem still exists. I have submitted a bug report, and will refer them to this thread as well.
Last edited:
I updated macOS to 10.15.5 public beta 5 and first it seems there no this strange audio issue at all, but in headphones i heard some issue only 4-5 times per song, it is much less than on 10.14.4
On 10.14.4 playback stuttering each 2-3 seconds and no need in headphones to hear it, even from built in audio sound was terrible.
New statistics:
2 users with E5520 rocksarntpeople, Riff_Al - have this issue on 10.14.4
1 user with X5550 [jsl] - have this issue on 10.14.4
1 user with E5520 stpmusic - do not have this issue, but he has 10.14.5 public beta 5
1 user MIKX with Dual Xeon 2.93 do not have this issue, but we do not know what exactly CPU he use.
1 user with W5590 Choppy23 - have this issue on 10.14.4 and 10.14.5 public beta 5
1 user with E5520 Riff_Al (it is me) - still have this issue on 10.14.5 public beta 5 but more less than in 10.14.4 I had to listen with headphones to hear the issue.
1 user rocksarntpeople upgraded his E5520 to X5680s and got rid of this issue.
i think user stpmusic with E5520 still have this issue on 10.14.5 public beta 5, but he will have to listen carefully, may be in headphones, and he will hear the issue. As i said, in 10.14.5 public beta 5 the issue is much less than on 10.14.4.
Common:
- all users with this issue have cMP 4.1 converted to 5.1
- all users have this issue on Mojave starting from 10.14.4
On 10.14.4 playback stuttering each 2-3 seconds and no need in headphones to hear it, even from built in audio sound was terrible.
New statistics:
2 users with E5520 rocksarntpeople, Riff_Al - have this issue on 10.14.4
1 user with X5550 [jsl] - have this issue on 10.14.4
1 user with E5520 stpmusic - do not have this issue, but he has 10.14.5 public beta 5
1 user MIKX with Dual Xeon 2.93 do not have this issue, but we do not know what exactly CPU he use.
1 user with W5590 Choppy23 - have this issue on 10.14.4 and 10.14.5 public beta 5
1 user with E5520 Riff_Al (it is me) - still have this issue on 10.14.5 public beta 5 but more less than in 10.14.4 I had to listen with headphones to hear the issue.
1 user rocksarntpeople upgraded his E5520 to X5680s and got rid of this issue.
i think user stpmusic with E5520 still have this issue on 10.14.5 public beta 5, but he will have to listen carefully, may be in headphones, and he will hear the issue. As i said, in 10.14.5 public beta 5 the issue is much less than on 10.14.4.
Common:
- all users with this issue have cMP 4.1 converted to 5.1
- all users have this issue on Mojave starting from 10.14.4
Last edited:
RIFF_AI -- I can confirm that I also felt the audio glitches were improved when I updated to 10.14.5 public beta 2. I was excited at first, but they still happen upon further testing, unfortunately.
For me, I hear glitches when playing notes from my midi controller that trigger audio samples (after a few seconds of holding a note). iTunes, Spotify, movies, everything else is okay. I used to be able to play 80-track midi symphonic pieces without an issue. Now playing 1 track glitches. Changing the Latency in my DAW from 256 to 2048 samples has a positive effect, but yet still glitches. Just much less.
For me, I hear glitches when playing notes from my midi controller that trigger audio samples (after a few seconds of holding a note). iTunes, Spotify, movies, everything else is okay. I used to be able to play 80-track midi symphonic pieces without an issue. Now playing 1 track glitches. Changing the Latency in my DAW from 256 to 2048 samples has a positive effect, but yet still glitches. Just much less.
Do not believe W5590 was ever an Apple-issued CPU. It is also a Gainestown processor. Is your machine a 4,1>5,1 or authentic 5,1?
Have not seen reports of this issue impacting Bloomfield, Gulftown, and Westmere-EP (which were all officially issued in MP5,1). If anyone with those processors has this issue, please report.
[doublepost=1554980451][/doublepost]
GAINSTOWN processors. So far that is the one common theme.
Reports of issue being resolved when moving to Bloomfield, Gulftown, or Westmere-EP processors.
Yes I agree.
But how can processor affect the sound?
There is no audio chip in these processors, not even integrated graphics.
But I thought, maybe it caused with bootrom update.
10.14.4 installation is associated with update to 141.0.0.0.0 bootrom.
Are there any microcode of the chipset in bootrom? Maybe new chipset microcode is not good for Gainestown, or totally not for Gainestown?
But if so, why 10.14.3 with 141.0.0.0.0 bootrom do not have the audio issue and 10.14.5 have much less than 10.14.4?
P.S. I don’t remember any of the previous macOS to cause so many different difficulties in MacPro. Need to replace the graphics card to Metal supported (and no nVidia drivers till now); Mojave do not recognise nVidia GTX680 like Metal suppotred when clean install from stick (but clean install available when run the installer from another macOS volume), need to replace Wi-Fi card, because Apple removed fully functional Atheros from Mojave (easily solved by adding a driver); the inability to use BootCamp assistant for Windows installation (assistant is necessary to install macOS and Windows on a single drive); various reported problems with the new APFS file system; and now the strange audio issue has also been added in this list.
I edited my post above.
Last edited:
GAINESTOWN processors were/are EXCLUSIVE to MacPro4,1 release.
MacPro4,1 is NOT officially supported for Mojave. A native MP4,1 will not get past El Cap without modification/hacks. You can make as many arguments that you want, saying the machine is basically the same as MP5,1 but there absolutely is different hardware in regards to the processors (and trays and SMC).
Many users have hacked their machines to "pretend" to be a MacPro5,1 (which is officially supported for Mojave). Doing this type of hack and not expecting ANY sort of issues down the road is naive. You do that hack at your own risk. You've discovered one of the risks.
It may be time to realize you've hit the end of the road for this particular hack with these particular processors, if you care about this particular issue. If you want your machine to ID as a MP5,1 without updating (some of) the hardware, you may have to accept the fact these issues exist. Again, you'd be stuck on El Cap without hacks. However, the fix is relatively simple - UPDATE TO MP5,1 PROCESSORS.
Apple does not disclose their changes/updates for security very clearly. You can argue that is good or bad at another time. There is a good chance there were/are software-based security changes made to the OS that (obviously) impacts GAINESTOWN processors. I would assume this is a patched security vulnerability fix, but is just an assumption. There is no reason any of the Apple engineers/developers should be testing against GAINESTOWN processors in Mojave at this time. Again, NONE are officially supported or issued with officially compatible Mojave machines.
MP5,1 was likely never planned for Mojave. That's another topic for another day/thread. However, Apple has VERY CLEAR hardware requirements for Mojave that they've clearly laid out on a single page - https://support.apple.com/en-us/HT208898. Either abide by them or stay on High Sierra (or switch to Windows). It's not any more difficult than that.
The other stuff you are rattling off are pretty much special-use case scenarios that likely impact a fraction of Mac Pro users, which are a fraction of Apple/macOS users overall. You can rant and vent as much as you want, but that is not going to change the situation that currently exists. Knowing the rules/lines/boundaries/lanes that Apple has slotted these machines into is what has enabled so many users to take full advantage of these MP5,1 workstations and eek additional performance out of them. Couldn't ask for anything else from a 7-9 year old based machine.
If you are an authentic Mac Pro 5,1 owner running Mojave you can register your complaints to Apple via their feedback form:
https://www.apple.com/feedback/mac-pro.html
If you are a registered developer, add SPECIFIC feedback that this issue exists with GAINESTOWN processors:
https://developer.apple.com/bug-reporting/
I personally cannot see Apple devoting a ton of resources to fix/patch an issue that only impacts unsupported machines, but if the issue is minor/quick to fix maybe it will happen.
MacPro4,1 is NOT officially supported for Mojave. A native MP4,1 will not get past El Cap without modification/hacks. You can make as many arguments that you want, saying the machine is basically the same as MP5,1 but there absolutely is different hardware in regards to the processors (and trays and SMC).
Many users have hacked their machines to "pretend" to be a MacPro5,1 (which is officially supported for Mojave). Doing this type of hack and not expecting ANY sort of issues down the road is naive. You do that hack at your own risk. You've discovered one of the risks.
It may be time to realize you've hit the end of the road for this particular hack with these particular processors, if you care about this particular issue. If you want your machine to ID as a MP5,1 without updating (some of) the hardware, you may have to accept the fact these issues exist. Again, you'd be stuck on El Cap without hacks. However, the fix is relatively simple - UPDATE TO MP5,1 PROCESSORS.
Apple does not disclose their changes/updates for security very clearly. You can argue that is good or bad at another time. There is a good chance there were/are software-based security changes made to the OS that (obviously) impacts GAINESTOWN processors. I would assume this is a patched security vulnerability fix, but is just an assumption. There is no reason any of the Apple engineers/developers should be testing against GAINESTOWN processors in Mojave at this time. Again, NONE are officially supported or issued with officially compatible Mojave machines.
MP5,1 was likely never planned for Mojave. That's another topic for another day/thread. However, Apple has VERY CLEAR hardware requirements for Mojave that they've clearly laid out on a single page - https://support.apple.com/en-us/HT208898. Either abide by them or stay on High Sierra (or switch to Windows). It's not any more difficult than that.
The other stuff you are rattling off are pretty much special-use case scenarios that likely impact a fraction of Mac Pro users, which are a fraction of Apple/macOS users overall. You can rant and vent as much as you want, but that is not going to change the situation that currently exists. Knowing the rules/lines/boundaries/lanes that Apple has slotted these machines into is what has enabled so many users to take full advantage of these MP5,1 workstations and eek additional performance out of them. Couldn't ask for anything else from a 7-9 year old based machine.
If you are an authentic Mac Pro 5,1 owner running Mojave you can register your complaints to Apple via their feedback form:
https://www.apple.com/feedback/mac-pro.html
If you are a registered developer, add SPECIFIC feedback that this issue exists with GAINESTOWN processors:
https://developer.apple.com/bug-reporting/
I personally cannot see Apple devoting a ton of resources to fix/patch an issue that only impacts unsupported machines, but if the issue is minor/quick to fix maybe it will happen.
I agree with all of this. As the original poster I knew there could be issues with the modded 4,1 setup but I took my chances (definitely not a stranger to issues like this regardless of platform).
I would have to say the definite answer to avoid unnecessary complications down the road (assuming you are in the 4,1>5,1 camp) is just make your machine as close to OEM 5,1 specs as possible.
We know, generally speaking that the hardware (backplane, psu, fans, chipset, etc) is the same and that one of the few differences is what processors they came with. As stated above, the simple solution seems to be swap those Gainstown chips for 5,1 Westmeres (which are ridiculously cheap btw, mine were $130 for a matched pair already delidded with a guarantee) and profit. I will continue to test to verify as much as I can that the issue is gone, but running various audio sources for about 5-6 hours yesterday produced no discernible abnormalities even via DP (which was a tragedy before the processor upgrade, even worse than the line-out)
I don't have authentic 5,1. It was a 4,1 flashed to 5,1. Then sent to OWC for a turnkey processor upgrade:
https://eshop.macsales.com/shop/turnkey/MacPro/2009_2010_Xeon_Processor/Apple_Mac_Pro_2009_2
Basically confirms this is another 4,1>5,1 with GAINSTOWN processors experiencing the issue.
Depending on when you had this upgrade performed, you may have a gripe with OWC. At least reach out to their customer service since you paid a premium to have them perform this upgrade.
They have a 1 year warranty, but unsure it covers incompatibility due to software vs. just hardware failure:
https://eshop.macsales.com/Service/warranty/owc-processorup-mp09.cfm
As stated a few times, the issues seem to unanimously point to 4,1 to 5,1 flashed macs running Gainstown processors, which the W5590 is.
I would say your options are as follows:
-Downgrade the OS (and stop there if you care about audio clarity, cuz its very unlikely to be fixed via a software patch as these chips were never offered in the 5,1s)
-Upgrade the processors to a compatible Westmere-EP chip (might as well go with this since it was the last ones issued in these machines meaning it could have support for longer)
-Contact OWC and explain the issue, perhaps they can compensate you depending on how long ago the purchase was made
As far as I see it, these are really your only options to maintain audio clarity
No authentic MP5,1 shipped with GAINESTOWN processors but I would 100% assume any authentic MP5,1 with processor(s) swapped out to GAINESTOWN would also experience this issue. 4,1>5,1 is a common factor but not the root cause.
Unlikely you'll see many/any issues with MP5,1 unless a user upgraded CPUs and went with 5500-series for some reason. Maybe could happen for a single processor to dual processor machine upgrade? Even then, does not make a lot of sense unless it was a tremendous cost savings/difference. (Likely a fractional percentage of MP5,1 users at best.)
Unlikely you'll see many/any issues with MP5,1 unless a user upgraded CPUs and went with 5500-series for some reason. Maybe could happen for a single processor to dual processor machine upgrade? Even then, does not make a lot of sense unless it was a tremendous cost savings/difference. (Likely a fractional percentage of MP5,1 users at best.)
I think its safe to say at this point that the issue is almost positively related to the Gainstown chips in MPs running Mojave 10.14.4. Suggest updating the post on the compatibility sticky to reflect this new information
Already done. This is likely GAINESTOWN in general on MP, regardless of 4,1>5,1 or 5,1. Just don't see many (any?) MP5,1 with GAINESTOWN upgrades this issue would impact.
Hahaha yeah I actually edited the reply to reflect that as youre probably right. Makes sense since that chip never made it into officially supported Mojave compatible models anyway.
[doublepost=1554998364][/doublepost]First post edited to reflect all findings. Thanks again guys, and will update if any further developments come to light regarding this!
Last edited:
I have an open bug report with Apple, and have notified them of the issue with GAINESTOWN processors. I also listed the specific processors mentioned in this thread.
Since I'm also in the Developer program, I have access to the beta Mojave releases and will report back here with any updates/fixes, etc.
Since I'm also in the Developer program, I have access to the beta Mojave releases and will report back here with any updates/fixes, etc.
Howdy did you get the 5690 3.46ghz 6 core chips yet? If so did the audio go away?
Anybody with Westmere 5690 6 core CPUs in there Mac Pro 2009 with FW 5,1 with no audio problems after upgrading to 10.14.4? Thanks DonH
I just tested a clean install of 10.14.4 with a dual E5520 tray. No audio problems with:
Anyone has a link for something that for sure triggers the problem?
[doublepost=1555029429][/doublepost]
[doublepost=1555029646][/doublepost]This is not a Gainestown problem…
Edit: can't disprove that this are Gainestown related.
- Netflix + Firefox
- Youtube + Safari or Firefox
- IINA + 1080P h264 videos
- iTunes.
Anyone has a link for something that for sure triggers the problem?
[doublepost=1555029429][/doublepost]
[doublepost=1555029646][/doublepost]
Edit: can't disprove that this are Gainestown related.
Last edited: