Nasty audio after 12 minutes on OSX10.6.7 using soundflowerbed. restarting soundflowerbed resolves for a repeat crash

[GoogleCodeExporter]

What steps will reproduce the problem?
1. Run any app (I use NI Maschine) and route audio to 16ch SF or 2ch - it dont 
matter
2. Run any target app (I use G-Force v4.0 by Soundspectrum) and select eitehr 
input
3. Run Soundflowerbed and choose my audio device for tapping (I use an Apogee 
Duet)

What is the expected output? What do you see instead?

For 12 minutes (almost precisely) all is great and working as expected - I can 
monitor the visual results in the target app (it's an audio-to-visual app) and 
also hear the signal with acceptable quality by the normal route, no artefacts.

After 12 minutes the app shows no visual evidence of a crash, but audibly the 
sound degrades to digital distortion and to resolve I quit Soundflowerbed and 
restart.

What version of the product are you using? On what operating system?

Soundflowerbed 1.5.2 OSX 10.6.7, Native Instruments Maschine as the source and 
Soundspectrum G-Force 4.0 as the target.

NOTE: it makes no difference to the rsults whicever apps I choose as source and 
target.

Please provide any additional information below.

Original issue reported on code.google.com by RakkMu...@gmail.com on 18 May 2011 at 11:00

[GoogleCodeExporter]

I can see after a review of the issues you have this one across a bunch of 
other issues - please feel free to merge. Thanks

Original comment by RakkMu...@gmail.com on 18 May 2011 at 11:04

[GoogleCodeExporter]

The problem here can be visualized as two "sinks" (destinations) pulling audio 
from the Soundflower at potentially different rates (even the difference is 
only slight).  

For example, the visualization program maybe driven by Core Audio's built-in 
clock whereas the audio monitoring is being driven by the Apogee Duet's clock.  
And at what rate is the buffer being filled by the NI app?

This is the most definitely challenging aspect of Soundflower.  Typically the 
larger you can make the buffer size, the better off you will be because you buy 
yourself more time before the buffer boundary is crossed.

Original comment by t...@electrotap.com on 3 Feb 2012 at 9:41

[GoogleCodeExporter]

nice to get a reply to this one - never got it resolved so looking forward to 
anything in the future

The clock has little effect on the time it takes to get the prob back - usually 
i run at 44k

Original comment by RakkMu...@gmail.com on 8 Feb 2012 at 12:56

[GoogleCodeExporter]

I had this problem and managed to fix it, but i dont know if it will work for 
others, basically i just went into system preferences, clicked on sound and 
where it says 'use audio port for' on both output and input i changed it to 
sound output ^^

Hope it works for you guys too

Original comment by DeadPand...@gmail.com on 8 Sep 2012 at 5:07