Robotic Static Sounds USB Headset

Not for everyone. I've tried this fix before (I found it somewhere online; maybe earlier in this thread, can't remember) and it doesn't work for me. It must be device dependent, or there is some other factor complicating things.

It may well be device dependent, which is why I posted my device I'd. I would guess the settings might also vary as well. So what works for one may not work for another.

Also I'm not entirely sure the change persists if you restart or unplug etc. I have had the problem return and have to go into midi settings again and set it again.

I'm using 10.6 as not been convinced to upgrade further ATM. So again that might effect it too???

I have several pairs of Logitech USB headsets that I've encountered the stacatto-robot-sound problem with. I use Mac OS X 10.6.8. The problem doesn't occur with analog headphones--with those, I have a different problem, which is that there is audible hissing during and a few seconds after each audio event, after which there's a click and the hissing goes away. But the robot sound problem does occur with all of the Logitech USB headsets I have. I'm working on a MacBook Pro from 2009.

I have noticed that the robot sounds only start when Mac OS X is out of free memory. Generally that means I have a lot of browser tabs open and the Flash plug-in running in Firefox, but it's not specific to any program or any time period--just the amount of memory being used. Whenever I notice the issue, running Activity Monitor in Applications -> Utilities will reveal something around 20MB free or less, which is basically nothing on a system with 4GB of RAM.

One solution is to close programs until there's more free memory available, but what's bizarre is that built-in speaker output is never affected no mater how much memory is in use or free, nor does the built-in speaker exhibit the hissing problem that analog headphones tend to have.

It would be great if the developers at Mozilla, Adobe, and Parallels (who indirectly most contribute to this problem through their respective memory leaks) would fix their bugs. But in the meantime, it would be even better if Apple could take a look at whatever driver/memory management interactions are causing this and just fix it at the OS level already.

This problem seems to be present even still in OS X Lion. I have 10.7.3 and still get robotic voice problem.

It's not always there, and while using Skype, hanging up and redialing usually solves the problem, but still very annoying.

But really sad that a bug like this is still around after years and years....

The problem is very much existent in Mountain Lion, but as mentioned here, it's not up to Apple fix it. Skype is using some old calls or something and the problem lies with them, which they acknowledge. Sort of, they claim they fix the bug in version 5.3 and won't really listen to claims it isn't fixed.

The problem is not Skype or other software, Apple should NOT have released a new kernel module that simply blows itself up if called the wrong way. It should handle those calls and be stable, or just remove the old API and be done with it.

This IS an APPLE problem.

Just to enphasize, after rebooting the computer, without loading any third party software:

- iChat with audio calls produced the problem

- FaceTime reproduced the problem during a video call

- QuickTime recording the incoming audio reproduced the problem

As far as I know, those are also Apple products, or not? =p

The whole thing about "it is a third party problem" is a big fat lie. Or better, lets go back to the old version of the kernel module and see if ANY of the products present the issue.

If you have several variables, it is hard to find the cause, ok. Just change the module, does it work? Yes. People in this thread did so for months, until you could not replace the module anymore due to compatibility with the new kernel.

This is one of the worst and most revolting issues I've ever had with a Mac, where Apple simply refuses to fix it and accept that they created the problem, while misleading people to blame other companies.

It's funny that you should mention it, but I can't reproduce the problem in either of the Apple apps you list. In fact, using Soundflower and LineIn, I've tried recording my voice with Quicktime while chatting on Skype. And when I got told my voice had gone robotic, I listened to the voice on QT and it was fine.

Yes, the problem started with the kernel change, and while I can't comment for why Apple did it, I'm sure they didn't to it just to annoy people.

MrJeaz wrote:

Yes, the problem started with the kernel change, and while I can't comment for why Apple did it, I'm sure they didn't to it just to annoy people.

Oh no, there was a good reason. If I remember right, they wanted to increase the channels and allow more bandwidth through the module. In other words, they did so to improve professional audio applications, which is great.

However, my last test with Lion was quite simple:

- reboot, yes, please.

- login holding "shift", just to make sure you don't load skype or any other "third party" tool. If you're sure you don't have those in your login, feel free to skip this step

- open quicktime player

- Menu File -> New Audio Recording

- on the triangle on the right of QT window, select your USB mic

- I then just pull the volume slider in QT window to the right a bit, just enough to hear some of my voice

- play a song, start to sing, whatever, just use the mic

- you don't even need to press "record", just keep talking, singing, whatever

- after a few minutes, 10, 15, 20, it varies, I hear my "robot self" talking to me... freaks me out... (lol)

That is what I do, and the problem is there. Don't know if I'm missing something. =)

USB headset still buzzing in Mountain Lion, and not just with third-party apps. It covers all sound, including iTunes playback and computer alerts.

Playing with Audio MIDI Setup may reduce the buzz, but didn't eliminate it for me.

I was not having this problem with my Logitech USB headset on my old 2008 Macbook Pro with Lion but now that I have bought a new Macbook Pro with Mountain Lion, this has suddenly become a huge issue. I was so happy to have found a relatively cheap headset and would not like to have to shell out more because of this issue.

Also, personally I have only had people complain when using Mumble and Ventrilo, not when using Skype. But I haven't tried iChat yet either.

Yep, this issue is one of those that will take ten or more years to be fixed as not everybody is affected and Apple already knows about the bug due to several bug reports but is not doing much apparently.

In the mean time, this is my solution:

- Logitech Stereo Headset H230

- http://www.amazon.com/Logitech-981-000018-Stereo-Headset-H230/dp/B000OAHOVO/ref= sr_1_6?s=pc&ie=UTF8&qid=1349061882&sr=1-6&keywords=logitech+headset

and one of these for new Macs:

- StarTech.com MUYHSMFF 3.5mm 4-Pin to 2x 3-Pin 3.5mm Headset Splitter Adapter - M/F

- http://www.amazon.com/StarTech-com-MUYHSMFF-Headset-Splitter-Adapter/dp/B004SP0W AQ/ref=sr_1_2?ie=UTF8&qid=1349061581&sr=8-2&keywords=4pin+3.5+to+splitter

Apple doesn't tell anyone that the headphone plug since mid 2009 I guess is a 4 pin, or per say the same as an iPhone. With the adapter, I simply use that Logitech on my iMac and it is PERFECT. I love it.

If you don't, you can have any 3.5mm heaphone + mic to work with it.

I hope it helps.

Seems like the 10.7.5 upgrade fixed this. I downloaded the 10.5 kext's on Lion but they would not load. I left them there, made sure the permissions and ownerships were correct, did a fix permissions on the main disk by booting in the root partition and then rebooting and updating to the latest 10.7.5 . The Info.plist shows the version of the driver as

<string>AppleUSBAudio</string>

<key>CFBundleGetInfoString</key>

<string>2.9.0f5, Copyright Apple Inc. 2000-2012</string>

After almost 2 years. finally

Hi Fakeh, No the 10.5 (Leopard) kext would not even load on my Lion (10.7.4). I updated the system yesterday to 10.7.5 and it fixed the problem. I was listening to audio the whole day and also recorded and skyped with colleagues. Before i had updated the system i had left the non working 10.5.x kext in the /System/Library/Extensions directory. Made sure to chown -R root:wheel the kext directory. Made sure all directorys were 755 perms and all files were 644. I also booted into rescue and ran disk repair and repair permissions on the disk.

Here is the size and md5sum of my 10.7.5 driver

-rw-r--r-- 1 root wheel 672K Aug 23 16:31 Contents/MacOS/AppleUSBAudio

5a46c3862e1f59b33c33c93dd22ac838 AppleUSBAudio

Here is a diff of the Info.plist file of the current 10.7.5 kext and the 10.7.4 one that i had. (+ is newer - is older)

@@ -3,13 +3,13 @@

<plist version="1.0">

<dict>

@@ -19,25 +19,25 @@

@@ -103,19 +103,21 @@

@@ -128,7 +130,7 @@

Sorry to say but this is not the case in 10.8.2. The AppleUSBAudio.kext has indeed been updated for Mountain Lion as well, bringing us to version 2.9.0f6, but the problem is still very much present. But I guess it's good that Apple are at least working on it.

I checked my headset recording using quicktime on my retina 10.8.2 with 2.9.0ft kext and its working fine.

I am attaching permissions and md5 of the files i have if that helps

drwxr-xr-x 7 root wheel 238 Sep 24 16:37 Contents

-rw-r--r-- 1 root wheel 4033 Aug 25 00:58 Info.plist

drwxr-xr-x 3 root wheel 102 Sep 24 16:37 MacOS

drwxr-xr-x 3 root wheel 102 Jun 24 23:07 Resources

drwxr-xr-x 6 root wheel 204 Sep 24 16:37 _CodeSignature

-rw-r--r-- 1 root wheel 466 Aug 25 00:58 version.plist

-rw-r--r-- 1 root wheel 159 Aug 25 00:58 CodeDirectory

-rw-r--r-- 1 root wheel 80 Jun 24 23:07 CodeRequirements

-rw-r--r-- 1 root wheel 2104 Aug 25 00:58 CodeResources

-rw-r--r-- 1 root wheel 4064 Aug 25 00:58 CodeSignature

-rw-r--r-- 1 root wheel 159 Aug 25 00:58 AppleUSBAudio.kext/Contents/_CodeSignature/CodeDirectory

-rw-r--r-- 1 root wheel 80 Jun 24 23:07 AppleUSBAudio.kext/Contents/_CodeSignature/CodeRequirements

-rw-r--r-- 1 root wheel 2104 Aug 25 00:58 AppleUSBAudio.kext/Contents/_CodeSignature/CodeResources

-rw-r--r-- 1 root wheel 4064 Aug 25 00:58 AppleUSBAudio.kext/Contents/_CodeSignature/CodeSignature

-rw-r--r-- 1 root wheel 4033 Aug 25 00:58 AppleUSBAudio.kext/Contents/Info.plist

-rwxr-xr-x 1 root wheel 338936 Sep 18 19:18 AppleUSBAudio

-rwxr-xr-x 1 root wheel 338936 Sep 18 19:18 AppleUSBAudio.kext/Contents/MacOS/AppleUSBAudio

drwxr-xr-x 3 root wheel 102 Jul 21 19:19 English.lproj

-rw-r--r-- 1 root wheel 73 Jun 24 23:07 InfoPlist.strings

-rw-r--r-- 1 root wheel 73 Jun 24 23:07 AppleUSBAudio.kext/Contents/Resources/English.lproj/InfoPlist.strings

-rw-r--r-- 1 root wheel 466 Aug 25 00:58 AppleUSBAudio.kext/Contents/version.plist

=====

MD5

======

MD5 (AppleUSBAudio.kext/Contents/_CodeSignature/CodeDirectory) = 6efcd3d75fb19686edd3e507e8ec3540

MD5 (AppleUSBAudio.kext/Contents/_CodeSignature/CodeRequirements) = f753651d2c79ec1619b7ea1ee57dc031

MD5 (AppleUSBAudio.kext/Contents/_CodeSignature/CodeResources) = 38e08edc2e9788851d0a61c4e459f5ba

MD5 (AppleUSBAudio.kext/Contents/_CodeSignature/CodeSignature) = 44205dbb1047b46de5234c6ff3ec2b36

MD5 (AppleUSBAudio.kext/Contents/Info.plist) = 662b76af792c5e441e7d25f4bfd9b957

MD5 (AppleUSBAudio.kext/Contents/MacOS/AppleUSBAudio) = ec9646206d1f06964313a956720e75b0

MD5 (AppleUSBAudio.kext/Contents/Resources/English.lproj/InfoPlist.strings) = ad6d8e110a0ece6940972d730af09934

MD5 (AppleUSBAudio.kext/Contents/version.plist) = 207b0b449f4bf47e63e921bdf8767beb