Re: Audio distortion after several hours?
Re: Audio distortion after several hours?
- Subject: Re: Audio distortion after several hours?
- From: Peter Lang <email@hidden>
- Date: Tue, 07 Jul 2009 23:52:26 +0200
I've got a similar problem with my MacBook Pro, C2D 2.16, 3GB. I use an RME
Multiface as my main sound device, connected via express card.
After starting the system, everything is fine for a while. Then, from time to
time, audio starts to sound like short buffers played with a very high volume
(therefore sounding distorted).
Reconnecting the audio device does not help.
This behavior occurred the first time with 10.5.7. It never happened with 10.5.6
or earlier.
What I've also seen is kernel_task consuming a lot of CPU at the moment of the
audio failure - together with a short mouse cursor hang.
At the moment, I see no possibility to give a reproducible scenario to get these
results, but maybe this information helps for further investigation.
Jeff Moore wrote:
Probably the best thing you can do is to nail down for certain how to
cause the problem to happen and file a bug. Be sure to also figure out
as much as you can about which Macs have the issue and which ones don't.
Chances are that given the circumstances (only certain Macs, only your
USB Audio device), the problem has to do with the interaction between
your device and the USB controller. Filing a bug is a fine way to get
the ball rolling on figuring out a solution.
On Jul 6, 2009, at 5:44 PM, Daniel Mack wrote:
On Mon, Jul 06, 2009 at 03:25:51PM -0700, Jeff Moore wrote:
Without a better description of the problem, particularly what the
distortion sounds like, how often it happens, precisely when it started,
etc. There really isn't much we can do to help you out.
Well, I know. Guess what I tell users to come up with ;) Unfortunately,
this vague reports (but many of them) is all we currently have. I've
only ever got to hear one piece of recorded audio material as evidence
for it all - hope to be able to provide that by tomorrow.
I'm totally aware that such an error description doesn't help much, I
just wanted to share our state of awareness, hoping someone with similar
problems chimes so we could look for commonalities.
FWIW, the usual cause of problems like this is that somebody somewhere
has rolled over a 32 bit number. Basically, somebody is using a 32 bit
number where they should be using a 64 bit number.
The recorded audio material shows audio artefacts every 128 samples.
While this is in general something the driver could be made resposible
for, there is still the fact that the issue persists after unplugging
and replugging the device. As the device is bus-powered, everything is
reset electrically, and the IOAudioEngine is deregistered. My verdict is
that if that doesn't help anything, the bug is supposed to be in some
other software layer. And again - this is something that only newer
machines with 4GB of RAM show. After many hours of usage.
Anyway, I'll provide more input on this as soon as I can.
_______________________________________________
Do not post admin requests to the list. They will be ignored.
Coreaudio-api mailing list (email@hidden)
Help/Unsubscribe/Update your Subscription:
This email sent to email@hidden