In reference to Peter's post about how to fix nogapped VBR headers, I've been talking with Mark Taylor, one of the lame developers, attempting to get lame's nogap+VBR support fixed.

The base problem is that the VBR headers that lame produced listed the number of frames in the mp3 short by one, on all VBR mp3s, not just the nogapped ones. That's been fixed. However, we're still having some trouble getting it exactly right.

ATTN: Peter!

In your post, you say that you were able to get your test matter to play without gaps once you get mp3tool to fix the VBR headers. I can't seem to reproduce that. There's always a very slight gap in my test cases. Is it possible that you were using a new version of empeg software for your tests that understands VBR headers better than 2.0b13 does? If not, can you give me some pointers on what I might be doing wrong?