How to make different tracks all the same volume?

This isn't quite as newbie as it may look. I use MP3 tracks as backing for my live gigs, played from an old Toshiba laptop. The trouble is, lots of tracks have been made over many years and many different computers and the difference in perceived volumes can be a problem. I want to try and process them all to make them all at least roughly the same volume.
Normalise won't work, because any decent peak will keep the rest of the track quiet. Hard Limiting will kill the dynamics, right?
I'm keen to experiment and figure it out for myself, but for the sake of a quick posting I thought maybe someone else had already solved a similar problem.

Cheers, Bag.

[loudness]

You can adjust the loudness of your files with this free tool:

http://www.maazl.de/project/mp3/mp3gain.html

It uses a relatively complex ear response model to adjust perceived loudness.  There is an equivalent function "wavGain" for wav files.  The nice thing is that mp3gain does the loudness adjustment using the mp3 file header  "gain" factor so it does not need to decode and then re-encode the mp3.

Audition also has a Group Waveform Normalize function that will apply to wav files (and seems to use a simpler ear model), but you will have to decode and re-encode (involving further loss of information).
.

Looks interesting! I'll check it out.

Cheers, Bag.