Apple

2) Currently I'm using AudioQueue and doing my processing in the
player callback. My current processing is very simple -- I look for
peaks, and shift as necessary to amplify the sound. I realize that
audiounits is the preferred method for this, but since I don't mind
some delay in the start of the playback, I wonder if I should simply
continue like this, since it's simpler.

sure - there's no reason not to be doing what you are doing, where you
are doing it

But doesn't this mean it will only work with PCM files?
At this point
(I assume this is just before callingAudioQueueEnqueueBuffer() )
mp3 files and other compressed formats are not yet decoded into PCM format.

I'm asking this since I'm struggling with getting the
AudioQueueOfflineRender() function to work
to be able to perform some DSP processing to the audio before sending it out
to the audio out device.

The way I use AudioQueueOfflineRender() it produces silence only.

Would appreciate if somebody could explain how to use
AudioQueueOfflineRender() in a correct way.
The available documentation is very sparse and no sample code

Right -- I presume this implies pcm. I will need compressed format eventually, and I've assumed that I'll have to go to AudioUnits to do my processing in that event... But a confirmation would be appreciated...