I believe that it's this second method that is the reason you may be getting only 90% effectiveness with your gapless playback instead of 100%. The problem is that these "silent" frames can actually contain information necessary to fully decode previous or following frames

Right, I'm aware of the bit reservior issue, and GapKiller does the best with what you've got. It currently does the only two possible things it can: A) trim existing frames via trial and error, B) split a single large MP3 cleanly at frame boundaries.

I don't think the playback of the Empeg Car 2.0 software attempts to preserve the bit reservoir across tracks, so even method B won't sound perfect. It's the correct behavior for a player to empty the bit reservior between tracks when it has no clue whether the material was supposed to be gapless or not.

All I'm saying is that the 2.0 software now sounds (to my ear) exactly like the WinAmp gapless playback plugin that I was using to begin with. In other words, it no longer has a tiny pause between two tracks as it sorts out the next song. It plays one track, then gets right on with the next one.

I'll play with some deliberately-split tracks and see how they sound on the 2.0 software..

___________
Tony Fabris