Improved synch when multiplexing...

[ntscuser]

I'm assuming here you already have the sound and video files within about a second of each other after stretching or resampling.

Rather than have to re-stretch or re-sample which is unpredictable anyway and assuming there is nothing of any great importance during the last second of the programme, simply cut the end of the .wav file using CoolEdit or similar. Don't overwrite the original sound file as you may need again it later.

I cut the .wav file in increments of half-second (500ms), quarter-second (250ms), eigth-second (125ms), etc. then I re-encode the programme using TMPGEnc, the shortened .wav file and the original .avi or D2V project file and check the result. TMPGEnc will stretch the sound file to match the picture. I expect it's also possible to cut the video file if the synch error is in the other direction? (I haven't tried that yet).

So far I've been able to achieve near-perfect synch until almost the very end of a programme and even then the error is only noticable because I'm deliberately looking for it.

[DJRumpy]

If you know the delay, then a couple of the multiplexers will simply allow you to specify an audio delay during multiplexing. I wan't to say bbMPEG does this, but I haven't verified. It's either bbMPEG, or ReMPEG. You can also use a more precise audio editor. I think goldwave will let you specify the audio length exactly to match your video's length.