Dear Sirs,
I want to ask You rather dilettantish question so I apologize in advance for my amateurishness, but I am concerned with this question and ask Your help to make it clear. So here is the question – when dealing with low-quality audio releases (e.g. demo-tapes, bootlegs etc.) should one use the highest available bitrate when coding to lossy or even code to lossless, or it is unworthy and 128-192 kbps is completely enough?
Thanks in advance.
Regards,
Alkslay

Dearest Sir,

ask yourself: Is it a problem to add minimal, hardly audible sound problems to the huge sound desaster or is it not.

If it is, for whatever reason, you may go lossless direction.
If it's not, you may try vorbis aotuvb5/nero aac lossy codecs at maybe ~70 kbps or lower (try!). You'll get impressive results, I'm sure.

Why not LP filter these bad recordings, too, to start with, and perhaps reduce the sampling rate to the signal (as opposed to noise) bandwidth if they are that bad?

Squeller - I wonder if there is an effect of "disaster", as You spoked, being "squared" by audio compression (i.e. big sound problems become even greater).
Woodinville - Thank You very much for Your attention. However I am not an adherent of audio processing.

It doesn't work that way. Chances are your ogg vorbis files at 96 kbps or LAME mp3 files at 128 kbps will be indistinguishable from your original sources TO YOU.

The point of lossy compression is to mimic the quality of the source as much as possible while cutting file sizes down by a lot, that's all.