What are you listening on when you bounce to disk? If you're listening to a WAV
file through your onboard soundcard then to headphones, it could be your
soundcard.
Try bouncing the track to disk, and then re-import it to pro tools and have a
listen. If it still sounds slightly different, then phase-reverse the bounce against
everything else (make sure there's no inserts on your master fader, because
the bounce will go through these inserts again). If there's total cancellation
(except for reverbs etc.), then it's your soundcard.
If it's not your soundcard, try setting up an audio track and set the input to
a free stereo bus (eg. bus 31-32). Then set anything with "Output 1-2" to
"bus 31-32", and record everything onto this audio track. (If it's clipping while
recording, set up a master fader for bus 31-32 and pull it down a few dB).
Then, once the track has been recorded onto the new audio track, select
the whole file (it must be ONE continuous audio region), find the region in
the regions list (once you've selected it in the edit window, it should become
highlighted in the regions list). Right-click it in the regions list, and click
"Export regions as files". The window that comes up is very similar to the
"Bounce to Disk" one. The main difference being that it doesn't do it in real
time, and only exports what you've selected.
Also, a thought just came to me. What are your session parameters? Whats
the session sample rate? Bit depth? Are you sure you're not bouncing to disk
as "multiple mono" or "mono (summed)"?
Some people have said that bouncing to disk is not a strong feature, that
there tends to be automation drift and other problems. I haven't found one.
However, I normally record my mixes back on to an audio track and export
them (as I outlined above). I do it because even though I haven't noticed
any problems with bouncing to disk, I'm still paranoid of it. And if I notice any
problems with the mix half way through the bounce, I need to cancel it, fix
the problem, and start again. But if I'm recording my mix back in, I use
destructive record, so that if I notice a problem, I can fix it, then start
recording again a few seconds before where I stopped, and it'll overwrite any
underlying audio unnoticeably (watch out for cutting off reverb tails though).
Also, you can use this if you're doing alternate mixes (without vocals, drums
only etc.) by setting up multiple audio tracks with different bus inputs, and
then setting different multiple outputs on my tracks, and record all mixes down
at once. Then just export them all. It'll take MUCH less time at the end of the
day than bouncing out every single one.
Sorry for babbling on, just thought they were some interesting tips
