I capture with an ATI Tv-Wonder card to MPEG-1. The big mpeg file I'm left with is generally mistake-free, with good audio sync.
Then I use TMPGEnc's mpeg splitter to extract the non-commercial sections, and use a MPEG join program to join the pieces and make the final MPEG file.

However, I usually get audio lag once I've split the file into sections with TMPG. I don't see why this would happen, since the file is in good sync before the split, and it's not like it is going through any encoding.

Is there anything I can do to stop this? Is it a problem with my initial MPEG-1 file, or with TMPG?

Thanks,
Dan