If some songs in your music collection seem quieter than others, this is either because:
- Their average volume is in fact lower (in dB)
- They are not as heavily compressed as others, so they sound quieter, while in fact their peaks are the same. In my experience, this only happens when you compare e.g. classical records to modern pop, or jazz music to metalcore. So probably not your average listening scenario.
You really don't want to apply dynamic range compression again, since what you'll get is what you mostly experience when listening to radio stations that employ technicians who have no idea what they're doing. The chorus of a song will not seem as loud as the rest, when it's overcompressed.
Plus, guitar equipment is mostly tailored to electric guitars, which have a much smaller frequency band than a typical pop/rock song.
Let mastering engineers do their job, and try to regulate the playback volume only – not the playback dynamics. This is really something that only works well with movie audio.
I'd suggest running a ReplayGain analysis over your music collection, which will add a metadata record to your files that identifies their perceived loudness. Playback devices (and I assume most players) should respect this ReplayGain value and match the volume accordingly so you'll never have to worry about mismatches again.
For Linux, mp3gain
is the de-facto standard for calculating ReplayGain. See here for more: Replay Gain in Linux