Re: RME UFX II weird bit crusher distortion problem

RME Support wrote:

It's odd, there used to be a time when one could avoid audio glitches on laptops with Windows XP by turning off battery surveillance. Apple is about 10 years behind on that "feature"... :-P


Regards
Daniel Fuchs
RME

As an ex Mac user and current win10 user it's pretty astounding to me what hoops you need to jump through to get a good system performance on a mac. I mean disable THAT many services and features? Wtf?

Re: RME UFX II weird bit crusher distortion problem

ramses wrote:

BTW My wife told me that iPhone doesn't charge anymore the batteries. For security reasons they screwed it up. Now you need to login to iPhone 1st before you can charge it !!!

What brilliant company with such a 1st class support. If the devices at least would be cheaper / adequate to such poor service ...

You can blame North Korea for this one. During the latest international conference with other nations, they handed out USB fans. Most of the people attending the conference used these on their laptops. This could have been a security problem, but it turned out the North Koreans didn't use the opportunity. It scared a lot of state security services to see that these users simply hadn't a clue about the risk.

Attacks over USB aren't new. It was first demonstrated over ten years ago and some infected USB devices showed up in the past.

Apple responded by blocking USB access unless the user approves it by logging in.

Strangely enough, it doesn't always seem to happen with my iphone...

Probably something I don't understand. But understanding is hard, since Apple doesn't care to document the behaviour at all.

Besides, Firewire and Thunderbolt (and USB-C) are even more dangerous, as they allow direct ram access and could fi read keys from memory without even accessing the storage.

MB Pro - 2 X FireFace 400, FF800 & DigiFace USB
ADAT gear: Korg, Behri, Fostex, Alesis...

103 (edited by ramses 2018-09-09 19:38:42)

Re: RME UFX II weird bit crusher distortion problem

Security good and nice.

But my wife didn't connect the iPhone to a computer for charging, it was the standard power supply from Apple.

From this everybody expects to charge the device regardless of being logged in or not.

All I want to say is that this company should at least offer better service if they fail on something if you take the very exclusive product price into consideration.

And rejecting support with the argument that a 3rd party device was connected is ridiculous.  The whole aim of USB is a multi purpose interface to connect all types of USB devices.

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: RME UFX II weird bit crusher distortion problem

You suppose a charger is just a charger. It's as potentially dangerous as every other USB device. The Apple charger has 2 main CPU's and an unknown number of embedded CPU's. Any one of those could be used to launch an attack. POC's abound.

Even a simple fan could contain something else, besides the fan motor.

Nobody likes the inconvenience or cost of security measures. But they all whine when it goes wrong. At least, Apple responded in the end. None of the others did. And Android is still an accident-wating-to-happen, security wise.

These days, you're lucky if all you get is a virus on your computer. It could also be a virus running on your ethernet card, harddisk or sound system. One of these is being exploited as we speak, used in a targeted industrial espionage attack against a few companies. No virus scanner can catch that thing and removing it is extremely hard. Even simply detecting it took over a year. How many others are out there that we haven't found yet? Hec, even you car could be infected. There are around a million cars out there that run the open source software curl. That's safe, but what other stuff is running?

MB Pro - 2 X FireFace 400, FF800 & DigiFace USB
ADAT gear: Korg, Behri, Fostex, Alesis...

Re: RME UFX II weird bit crusher distortion problem

I tried these steps to disable the battery manager however it doesn't work for me. The driver still loads. I am running OSX Mojave 10.14 Beta (18A384a)

$ csrutil status
System Integrity Protection status: disabled.

I renamed the folder, it's currently:

$ ls -la /System/Library/Extensions | grep AppleSmartBatteryManager
drwxr-xr-x@   3 root  wheel    96B  5 Aug 21:07 AppleSmartBatteryManager.xkext/

Cleared System cache:

$ sudo kextcache -system-caches

After a reboot:

$ kextstat | grep Smart
   49    0 0xffffff7f838c5000 0xb000     0xb000     com.apple.driver.AppleSmartBatteryManager (161.0.0) 0B03A9DC-B9D1-3ACC-830F-C7CCBFCA89D7 <44 11 5 4 3 1>

Driver is still running..  Something I am not doing correctly?



ewtwolf wrote:

Simple version of MBP battery manager hack:

  • Make sure you are plugged into the AC Adapter

  • Reboot and hold Command+R for recovery mode.

  • Open Utilities>Terminal and type: csrutil disable

  • Press return and then reboot

  • In OSX go to System/Library/Extensions/AppleSmartBatteryManager.kext

  • Right Click>Rename>Change .kext to .xkext

  • Enter your password when asked

  • Now open the OSX Terminal: Applications>Utilities>Terminal and type: sudo kextcache -system-caches

  • Hit Return and wait for the new terminal prompt showing it processed

  • Reboot

  • Check that the hack is working by opening Applications>Utilities>Terminal and typing: pmset -g batt (It should return "AC Power" and no battery info)

That's it. To turn it off, just rename the .xkext back to .kext, run the Terminal cache command again, sudo kextcache -system-caches, and Reboot. Check it is off by running Terminal command: pmset -g batt (It should return battery charge info)

Turn SIP back on by booting into recovery mode and type csrutil enable in the terminal and reboot.

*You will not get any battery readout with the hack. So best to run it with the AC Adapter.

106

Re: RME UFX II weird bit crusher distortion problem

That's the same as what I had. Didn't seem to work so just set everything back to normal.

I only get this problem when I'm recording or monitoring audio in through the soundcard. Playback is fine for me, even running heavy audio in Ableton.

From this morning, the pop and digital noise at the end of this https://soundcloud.com/user-633193613/rme-glitch

Re: RME UFX II weird bit crusher distortion problem

Did you restart the machine afterwards?

There's probably a copy of the driver in the kernel extension's cache.

You could download Northern Softworks' Cache cleaner and run the function to delete all caches. The demo will do that. Afterwards, restart. The OS will rebuild caches, so startup will be a little bit longer.

http://www.northernsoftworks.com/

MB Pro - 2 X FireFace 400, FF800 & DigiFace USB
ADAT gear: Korg, Behri, Fostex, Alesis...

108

Re: RME UFX II weird bit crusher distortion problem

Is it confirmed to make a difference, I thought people had tried but the problem persists? I use my MBP a lot at work without an adapter so it's not ideal for me.

109 (edited by danielnayberg 2018-09-12 13:30:41)

Re: RME UFX II weird bit crusher distortion problem

Hey everybody - just chiming in here since I just (seem to have) solved this issue in my particular context after a week of frustration.

I recently bought a second hand Fireface UC and eagerly hooked it up to my 2018 Macbook Pro 15" (6-core i9, 32GB Ram, CPU-throttling-update installed, High Sierra 10.13.6, running Ableton Live 10) along with my Push 2, Midi Fighter Twister and iPad through a high quality USB C - USB A hub, only to find exactly the type of glitches and dropouts described in this thread.

After trying firmware-updating the interface, installing the 3.08 driver, replacing the USB-hub with another one, replacing all USB-cables etc. I still had the issue, and almost decided to sell the interface on. After all I bought it for its famed driver stability...

BUT then I accidentally discovered that whenever my Push 2 weren't connected or on, the problem vanished. I don't know what is going on, but it seems that whenever the UC and the Push 2 is handled by the same USB controller problems occur.
I simply plug the Push 2 into a separate adapter/hub connected to a USB C port on the opposite side of the MacBook and everything is running as it should (for now anyway).

So for anyone still experiencing this problem while using other connected USB-gear - if you haven't already, try isolating the RME stuff to it's own USB port.

110

Re: RME UFX II weird bit crusher distortion problem

Are you monitoring audio through the inputs?

My problem is when monitoring audio. I can be using it all day and it's maybe once or twice, sometimes more...you never know when it will strike.

fyi I have UCX on a separate port using an apple adapter and Push 2 plugged into a Belkin Hub, I've also tried a USB A to C cable.

111 (edited by ewtwolf 2018-09-13 16:57:44)

Re: RME UFX II weird bit crusher distortion problem

dlogik wrote:

I tried these steps to disable the battery manager however it doesn't work for me. The driver still loads. I am running OSX Mojave 10.14 Beta (18A384a)

$ csrutil status
System Integrity Protection status: disabled.

I renamed the folder, it's currently:

$ ls -la /System/Library/Extensions | grep AppleSmartBatteryManager
drwxr-xr-x@   3 root  wheel    96B  5 Aug 21:07 AppleSmartBatteryManager.xkext/

Cleared System cache:

$ sudo kextcache -system-caches

After a reboot:

$ kextstat | grep Smart
   49    0 0xffffff7f838c5000 0xb000     0xb000     com.apple.driver.AppleSmartBatteryManager (161.0.0) 0B03A9DC-B9D1-3ACC-830F-C7CCBFCA89D7 <44 11 5 4 3 1>

Driver is still running..  Something I am not doing correctly?

Make sure you don't try and rename the driver file within Recovery Mode. It needs to be done in OSX.

ONLY this command is run in Recovery Mode to turn off SIP:

csrutil disable

After you disable SIP:

1) boot back into OSX
2) change the driver name
3) run this command in OSX terminal

sudo kextcache -system-caches

4) then reboot.

I recommend renaming the driver file by navigating to the folder /System/Library/Extensions and Right Click > Rename the  AppleSmartBatteryManager.kext file so you can be sure the name was changed.

I know this terminal command definitely works for renaming in the terminal but still double check regardless before you clear the cache:

sudo mv /System/Library/Extensions/AppleSmartBatteryManager.kext /System/Library/Extensions/AppleSmartBatteryManager.xkext

Tested and works fine in Mojave.

You can check it worked after rebooting by running this from the terminal in OSX:

pmset -g batt

It will say: Now drawing from 'AC Power'

MBP 2018 i9 32GB RAM 1TB SSD - RME BabyFace Pro - Logic Pro X - Ableton 10 - Cubase Pro 9.5

112 (edited by dlogik 2018-09-14 10:21:38)

Re: RME UFX II weird bit crusher distortion problem

Yeah I dunno, I've tried your steps, the battery manager still loads on my machine... ‘i shall not clear that {k}ernel-{ext}ension-{cache}’

Anyway, I've found this helps a bit:

* Disable wifi / bluetooth / any network interface / as has been suggested {years ago}..
* Disable the systemsoundserverd service (requires System Integrity Protection status: disabled)

$ sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.audio.systemsoundserverd.plist

BTW - I can pretty much trigger the ‘bit crusher issue’ every time by disabling/enabling my network interfaces, a few times….

Anyone have any more info on this:

HALS_IOA1Engine.cpp:365:EndWriting:  HALS_IOA1Engine::EndWriting: got an error from the kernel trap, Error: 0xE00002EE

113 (edited by ewtwolf 2018-09-14 11:56:24)

Re: RME UFX II weird bit crusher distortion problem

Well the hack has worked for me across three different machines, and several new installs of both high Sierra and Mojave. I suggest you look at the permissions on the .kext file to make sure you didn’t alter it. It needs to remain System Read-Write and Wheel Read, Everyone Read. The kextcache command is needed for it to work too. Otherwise your system just loads the old version. I made two shell scripts I can just double click on my desktop to turn it on and off. Zero issues with the commands. They do work.

MBP 2018 i9 32GB RAM 1TB SSD - RME BabyFace Pro - Logic Pro X - Ableton 10 - Cubase Pro 9.5

114 (edited by dlogik 2018-09-14 20:17:47)

Re: RME UFX II weird bit crusher distortion problem

Thanks Man! I got it working, when I set secure boot to 'no security' the battery manager no longer loads.

Also, I think the bit crusher issue might be triggered by the timed service. When it updates my system clock, I see these errors and my audio playback becomes distorted:

default    11:14:37.508179 -0700    kernel    outCallback (147) - returned status: e00002be
default    11:14:37.508191 -0700    kernel    outCallback (147) - frame 7 returned status: e00002be
default    11:14:37.509158 -0700    kernel    outCallback (148) - returned status: e00002be
default    11:14:37.509168 -0700    kernel    outCallback (148) - frame 0 returned status: e00002be
default    11:14:37.509171 -0700    kernel    outCallback (148) - frame 1 returned status: e00002be

So I have also created a script which I run before an audio session, seems to help:

$ sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.timed.plist
$ sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.audio.systemsoundserverd.plist

115 (edited by gozk 2018-09-17 10:27:49)

Re: RME UFX II weird bit crusher distortion problem

Thanks, I got it working too. So switching the Secure Boot Mode is also a way to turn it on and off I presume?

With the Battery Hack working: Is there a way of finding out the battery status using Terminal?

Re: RME UFX II weird bit crusher distortion problem

gozk wrote:

Thanks, I got it working too. So switching the Secure Boot Mode is also a way to turn it on and off I presume?

With the Battery Hack working: Is there a way of finding out the battery status using Terminal?

You need SIP turned off in order to run the hack terminal commands. Otherwise it prevents you from changing name of system files. If you turn SIP back on, OSX doesn’t replace what you changed. It just prevents you from making more changes.

There is no way to see battery status with the hack active. That’s what you loose. The driver is what communicated that info to OSX. As long as you are plugged into AC then the system hardware will keep the battery charged in the back ground. Just rename the .kext file and clear the caches with terminal command, reboot, to get the battery driver running again.

MBP 2018 i9 32GB RAM 1TB SSD - RME BabyFace Pro - Logic Pro X - Ableton 10 - Cubase Pro 9.5

Re: RME UFX II weird bit crusher distortion problem

So... no other solution than the hack? Someone tried the official Mojave release?

Re: RME UFX II weird bit crusher distortion problem

I am having similar problems on a Mac Pro5,1
It has no need of icloud or network as it is never online, thinking to disable a bunch of stuff
Does anyone have a good guide on how to disable all those kexts for better audio performance?

Re: RME UFX II weird bit crusher distortion problem

Hi there, me again.

I didn‘t really find any time to work on this issue, contact apple support and/or go to a shop yet...

Also, i haven‘t been able to disable the Smartbatterymanager on my MBP... (still running the „last“ High Sierra Update, 10.13.6 i think)...

As you can imagine, the issues still occur... Does anyone have any news? Has someone already heard „good“ news from Apple Support, a hotfix even?

https://www.gearslutz.com/board/music-c … bug-2.html

Looking at this other thread I‘m starting to worry this Smartbatterymanager issue will stay around for longer... I‘m starting to consider ditching the MBP and going the mac mini route: no battery, no problems. It has the T2 in it though...

120

Re: RME UFX II weird bit crusher distortion problem

Just posted to current macbooks thread, long story short: firewire connection with mbp 2018 and rme UFX - no glitches anymore.

Re: RME UFX II weird bit crusher distortion problem

We are also experiencing show-stopping glitches recording with a MADIface Pro and 2018 13" MBP via official Apple adapter.  However, we're also getting frequent glitches on playback through the ADI-2 Pro FS, which bypasses RME drivers altogether.

Are people experiencing this problem too?

It's hard to test something so sporadic but it seemed as if the first of the two 10.13.6 'supplemental updates' fixed the issue, at least for the recordings we did during this time -- anyone else remember?

The glitches then returned in the second 10.13.6 update, and subsequent versions of 10.12...

Eastwood Records
www.eastwoodrecords.co.uk

122 (edited by dlogik 2018-12-04 10:12:01)

Re: RME UFX II weird bit crusher distortion problem

Have you tried disabling the 'timed' daemon? This reduces dropouts/glitching significantly for me on a Macbook Pro 2018 running Mojave 10.41.1

Re: RME UFX II weird bit crusher distortion problem

Thanks for posting this info!

Finally a bit of light at the end of the tunnel.

https://arenjayasemarang.com

Re: RME UFX II weird bit crusher distortion problem

dlogik wrote:

Thanks Man! I got it working, when I set secure boot to 'no security' the battery manager no longer loads.

Also, I think the bit crusher issue might be triggered by the timed service. When it updates my system clock, I see these errors and my audio playback becomes distorted:

default    11:14:37.508179 -0700    kernel    outCallback (147) - returned status: e00002be
default    11:14:37.508191 -0700    kernel    outCallback (147) - frame 7 returned status: e00002be
default    11:14:37.509158 -0700    kernel    outCallback (148) - returned status: e00002be
default    11:14:37.509168 -0700    kernel    outCallback (148) - frame 0 returned status: e00002be
default    11:14:37.509171 -0700    kernel    outCallback (148) - frame 1 returned status: e00002be

So I have also created a script which I run before an audio session, seems to help:

$ sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.timed.plist
$ sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.audio.systemsoundserverd.plist

Thanks for the tip.  You say this 'seems to help' but does it ever glitch once you've run the script?

I'm on a 2018 13" MBP running 10.14.1 and the hourly glitches/dropouts on an ADI-2 Pro FS and MADIface Pro are driving me nuts.  Switching to Firewire is not an option.

Has anyone successfully used the 12" Macbook for extended recording/playback?  I believe this is the last remaining Apple notebook that has USB-C and no T2 chip...

Eastwood Records
www.eastwoodrecords.co.uk

125

Re: RME UFX II weird bit crusher distortion problem

Same problem !! 
-> Macbook pro 11.3 15", late 2013, osx 10.13.6
+ Madiface pro

And no workaround ...  :-(

Maybe RME can do some pressure to Apple to fix this ..?

Re: RME UFX II weird bit crusher distortion problem

frabo wrote:

Same problem !! 
-> Macbook pro 11.3 15", late 2013, osx 10.13.6
+ Madiface pro

And no workaround ...  :-(

Maybe RME can do some pressure to Apple to fix this ..?

I thought the latest audio issues were related to USB-C and the T2 chip, but doesn't the 2013 MBP have USB-A and no chip?

Have you identified the issue in Console?  And have you tried Mojave yet?

Eastwood Records
www.eastwoodrecords.co.uk

Re: RME UFX II weird bit crusher distortion problem

Those with dropout issues could try to test audio load without SSD access and with SSD access. The T2 chip is responsible for any data transfer from/to the flash chips. I am curious if this makes a difference or not?!

Re: RME UFX II weird bit crusher distortion problem

dlogik wrote:

Thanks Man! I got it working, when I set secure boot to 'no security' the battery manager no longer loads.

Also, I think the bit crusher issue might be triggered by the timed service. When it updates my system clock, I see these errors and my audio playback becomes distorted:

default    11:14:37.508179 -0700    kernel    outCallback (147) - returned status: e00002be
default    11:14:37.508191 -0700    kernel    outCallback (147) - frame 7 returned status: e00002be
default    11:14:37.509158 -0700    kernel    outCallback (148) - returned status: e00002be
default    11:14:37.509168 -0700    kernel    outCallback (148) - frame 0 returned status: e00002be
default    11:14:37.509171 -0700    kernel    outCallback (148) - frame 1 returned status: e00002be

So I have also created a script which I run before an audio session, seems to help:

$ sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.timed.plist
$ sudo launchctl unload -w /System/Library/LaunchDaemons/com.apple.audio.systemsoundserverd.plist

Terminal returns this message: "Operation not permitted while System Integrity Protection is engaged"

Eastwood Records
www.eastwoodrecords.co.uk

Re: RME UFX II weird bit crusher distortion problem

ewtwolf wrote:

Without the battery manager running, I no longer get the coreaudio overload bug every 60 seconds when running a low latency active live track in Logic.

I'm reluctant to start altering the OS but may have to try this given how prevalent the glitches are.  I'd like to know Battery Manager is definitely the cause before proceeding. What message do I look out for in Console?

Also, have you spoken to Apple technical about your findings?

And why doesn't the battery manager affect older MacBooks?

Eastwood Records
www.eastwoodrecords.co.uk

Re: RME UFX II weird bit crusher distortion problem

I have this problem, on a Mac mini 2018 with a digiface USB, so it's obviously not a battery problem, no response yet on the forum, anyway to get some support round here?.. RME always was the go-to for stability, now I can't get 30mins without a audio error..

www.yila.co.uk

Re: RME UFX II weird bit crusher distortion problem

Hey, same problem, also on a Mac mini 2018, with a Fireface 802 this must be a T2 core audio issue, meaning the only fix is if apple is so kind as to fix it....

132

Re: RME UFX II weird bit crusher distortion problem

mjfe87 wrote:
frabo wrote:

Same problem !! 
-> Macbook pro 11.3 15", late 2013, osx 10.13.6
+ Madiface pro

And no workaround ...  :-(

Maybe RME can do some pressure to Apple to fix this ..?

I thought the latest audio issues were related to USB-C and the T2 chip, but doesn't the 2013 MBP have USB-A and no chip?

Have you identified the issue in Console?  And have you tried Mojave yet?

Yes, sorry, but I'm infected :-(
I tried the other usb port, no success...
No, Mojave, I have to work with protools ...
Console, where I have to look at ? user, mac-analyse or system reports ?

I contacted RME over a year ago, they asked some questions about usb, madi cable and if the sound is also on the inputs of the madiface, and some other, .. but once, the contact broke up..

I had some shows, where I used my madiface pro usb powerd, just for stereo out without any problems.
could it be that this has to do with the madiconnection, or the power supply ?

Thanks for help

atb

133 (edited by van 2020-03-28 02:25:42)

Re: RME UFX II weird bit crusher distortion problem

so we have 2020/03/28 and the problem still exists...
wtf?

Re: RME UFX II weird bit crusher distortion problem

van wrote:

so we have 2020/03/28 and the problem still exists...
wtf?


I'm afraid to buy now...

135 (edited by ramses 2020-04-01 11:44:22)

Re: RME UFX II weird bit crusher distortion problem

deliad wrote:
van wrote:

so we have 2020/03/28 and the problem still exists...
wtf?


I'm afraid to buy now...

If memory serves me right this was an Apple / MacOS specific issue. Where apple broke things a) USB stack in MacOS and b) introduced a new self designed communication chip which also caused USB problems.

EDIT: Deliad, wink wink ... you have a PC ... this thread is about Apple ...

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

Re: RME UFX II weird bit crusher distortion problem

I was dismayed to discover this thread today while shopping for an audio interface for my 2018 MacBook Pro!  Since it seemed unbelievable that Apple would allow such a bug to persist for so long, I kept Googling, and found this post on MacRumors, dated 16-July-2020:

macOS Catalina 10.15.6 Successfully Fixes USB 2.0 Issues Experienced by 2020 MacBook Pro and MacBook Air Users

Can any of you who have experienced this problem confirm (or refute) that MacOS 10.15.6 fixes it?

Thanks,
Michael

Re: RME UFX II weird bit crusher distortion problem

Just to share my experience with the same bug of my RME UC & Windows 10. It started one day that I got frequently this "bit-crushed" sound on playback. It remained for a while then normal playback came back and sometimes soon, sometimes later the distorted audio playback was back again. I got really mad about it and was thinking that it may be a defect of the interface.
In parallel I noticed a Windows error message everytime I shut down my computer via the normal process. In the event log I also found messages with errors pointing to areas I could not relate to RME drivers or applications.
But then I found the information by accident of a newer TotalMix FX version (1.66 vs 1.62 of my driver package). It fixes an issue in shutting down the application. Once I replaced the TotalMix FX EXE there was no longer the error message during shutdown and also I never had the distorted audio!

Re: RME UFX II weird bit crusher distortion problem

This same issue has been plaguing me for quite some time. I actually bought a new RME Babyface Pro FS to try to solve the issue - no luck. It really does seem to be the way that Apple handles USB audio devices.

I'm on a 2015 MBP, 15". So it's not a T2/2018 MBP issue.

It specifically arises on the input side - the "glitches" are sporadic and only apparent on incoming audio. You can hear it when playing back the recorded audio with glitches written to the file.

As mentioned, the problem is way beyond just RME - MOTU, Blue, UA, and seemingly any other manufacturer's USB2.0 devices are experiencing these problems with what seems to be increased frequency.

https://cdm.link/2019/02/apple-2018-glitch/
https://cdm.link/2019/03/apple-usb-audio-2018-fix/

These articles seem related, if I'm not mistaken. But they don't address the issues that affect the broader range of Apple computers that folks on this thread have reported experiencing problems.

139 (edited by gl 2021-10-26 08:45:15)

Re: RME UFX II weird bit crusher distortion problem

I have the same bitcrusher effect with the Babyface Pro on Windows 10 with latest firmware/drivers and TotalMix 1.75.  This happens on my DeskMini X300 system (https://www.asrock.com/nettop/AMD/DeskM … /index.asp) with a Ryzen 5600G APU.

The Babyface Pro is connected to the only external native USB 2.0 port, but there are an additional 2 USB 2.0 ports from a bracket connected to the only 2.0 internal 9pin header.

When the BabyfaceP is the only USB2.0 device connected, or when only a USB2 keyboard is also connected, audio is flawless at the lowest latency @ 44.1k.  However as soon as I plug in another USB 2.0 device into the header bracket (tested with a couple of pen drives), while the device is being accessed, the bitcrushing kicks in, gets worse over a few seconds to where the (crushed) audio then gradually fades out, until it's completely silent - either permanently (until ASIO playback is stopped and restarted), or sometimes it briefly comes back before crushing out again.  This happens even at high ASIO buffer sizes - it's like the other devices unfairly hog the 2.0 bandwidth and choke the ASIO stream (educated guess), or maybe they affect transfer timing somehow?

I know that AMD has had overload issues with their Ryzen USB controllers, and those have been patched in various BIOSes.  Is this a remaining side effect?

Happy to try anything - is the debug driver still valuable, or too out-of-date?

140 (edited by ramses 2021-10-26 10:21:26)

Re: RME UFX II weird bit crusher distortion problem

gl wrote:

I have the same bitcrusher effect with the Babyface Pro on Windows 10 with latest firmware/drivers and TotalMix 1.75.  This happens on my DeskMini X300 system (https://www.asrock.com/nettop/AMD/DeskM … /index.asp) with a Ryzen 5600G APU.

The Babyface Pro is connected to the only external native USB 2.0 port, but there are an additional 2 USB 2.0 ports from a bracket connected to the only 2.0 internal 9pin header.

When the BabyfaceP is the only USB2.0 device connected, or when only a USB2 keyboard is also connected, audio is flawless at the lowest latency @ 44.1k.  However as soon as I plug in another USB 2.0 device into the header bracket (tested with a couple of pen drives), while the device is being accessed, the bitcrushing kicks in, gets worse over a few seconds to where the (crushed) audio then gradually fades out, until it's completely silent - either permanently (until ASIO playback is stopped and restarted), or sometimes it briefly comes back before crushing out again.  This happens even at high ASIO buffer sizes - it's like the other devices unfairly hog the 2.0 bandwidth and choke the ASIO stream (educated guess), or maybe they affect transfer timing somehow?

I know that AMD has had overload issues with their Ryzen USB controllers, and those have been patched in various BIOSes.  Is this a remaining side effect?

Happy to try anything - is the debug driver still valuable, or too out-of-date?

Seems that your system has no PCIe expansion slots. Then it would have been possible to add an dedicated USB3 PCIe card to your system to isolate the recording interface from the rest of your systems USB infrastructure. As nice as it is to have a very compact system. If it's too small then you can e.g. not expand like in this case to get a potentially better solution than having to wait for BIOS updates, where you never know, whether a) it can be fixed in the current chipset and b) this fix is still intact in further (potentially required) upgrades.

My server mainboard has a good Intel Chipset C612 and I could initially connect two UFX+ (with a lot of audio channels!!) and an ADI-2 Pro. Issues arised once I started connecting additional USB devices for Bluetooth and additional stuff, then I got audio drops at lowest ASIO buffersizes, which were not there before.

I could fix that by adding an USB card with supported FL1100 USB3 controllers. In this case an expensive card, because it had four USB3 controller on board, so that each connected device can be fully isolated behind such an USB3 controller and also has the full USB3 bandwidth.

If you look for such a card then I check that it also has support for UASP, because this enables for faster and more efficient data transfers, which is very interesting to be able to provide fast backups and restores and especially if you have to deal with a lot of small files. An external USB based backup device you can turn off as long as you use a recording interface. Then you can take a good card to a more affordable price, where 1-2 USB3 controller are shared for four USB connectors.
Sonnet builds good card, but also Startech has good offerings and has also matching USB housings which also need to support UASP if you want to make use of this technology.
https://www.startech.com/en-us/blog/all … about-uasp

I tell you this so that you have a chance to draw a good decision, whether to stay with this small system or to make some thoughts how to expand. Typicall you could keep CPU and cooler (would get a good and silent tower cooler instead), DRAM and then get a case with good PSU (e.g. hybrid where blower only turn on when needed under load) and generally a better mainboard, but look that you get a supported one for your CPU and with e.g. passive cooling for the chipset and the features that you need. The board needs to provide connectors for monitors if you want to keep using your CPUs internal GPU. It can be an advantage to get a not too expensive separate GPU. Also have a look at the PCIe slot layout, maybe this is also a good occasion to get a faster SSD (M.2, ..).

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

141 (edited by gl 2021-10-26 22:36:27)

Re: RME UFX II weird bit crusher distortion problem

The small form factor is a deliberate choice as this is a live audio rig built into a camera-style case (it's a cool but powerful 8core/16thread rig):

https://r-i-l.net/_pics/dB/dB_live_rig.jpg


My understanding is that this motherboard doesn't actually have a chipset, so the APU is providing everything (including the USB controller, PCIe lanes etc) itself.  That means that this is raw AMD Ryzen that AMD is fully in charge of, which reduces potential places of failure, and should work (otherwise potentially all Ryzen platforms could be affected, which RME should want to avoid).

Also these aren't the usual dropouts or glitches, but a weird bitcrusher effect that I've never seen in decades with various audio interfaces.  It sounds more like transfer timing is gradually drifting on a microscopic level, a sort of de-synchronization, rather than buffer overruns (ie. audio glitch or skip).

I don't think this is normal, confirmed by larger ASIO buffer sizes not helping.  In fact the system is capable of perfect latency, the lowest settings work perfectly otherwise.

So this is some USB incompatibility, most likely to do with recent Ryzen USB controller updates in the AMD BIOS, that RME should look at, or work with AMD on.

Re: RME UFX II weird bit crusher distortion problem

Chipset is AMD X300, but seems to be integrated into the CPU.

https://www.computerbase.de/2020-09/asr … x300-test/ "ASRock has fully adapted the X300 chipset, which is actually not a chipset at all because all connections are realized via the CPU/APU, to Renoir. First and foremost, the manufacturer has provided a completely revised BIOS that now also allows overclocking."

On the download page I noticed different chipset driver for Win10 and Win11.
Which chipset driver are you using ? Maybe worth an upgrade...
https://www.asrock.com/nettop/AMD/DeskM … p#Download
For Win10: AMD chipset driver ver:2.11.26.106 2021/3/11
For Win11: AMD chipset driver ver:3.08.06.148 2021/10/21

BR Ramses - UFX III, 12Mic, XTC, ADI-2 Pro FS R BE, RayDAT, X10SRi-F, E5-1680v4, Win10Pro22H2, Cub13

143 (edited by gl 2021-11-02 01:30:08)

Re: RME UFX II weird bit crusher distortion problem

As I said, it's not a real chipset, CPUs always supply a certain amount of PCIe lanes etc.  On this motherboard it just doesn't add any more with a 2ndary chipset.

Thanks for the drivers link but I was already using more recent ones, including the latest BIOS.

The issue seems specific to USB2.0 ports, it works fine on in-built USB3.0 ports.  Also when I tried it again, the bitcrusher effect was less common, it mostly just had weird gaps and dropouts, still dropping quickly to total silence though as before.