I’m doing a lot of encoding/converting and I’m still trying to understand the intricacies of the process. Recently, I did a conversion that actually worked, and I just want to express a theory of why it worked and if someone could just confirm my thoughts? Hopefully, this will help others out there doing the same thing.

I had an widescreen PAL AVI in which Quicktime didn’t play back the audio (silent playback). VLC, however, did play back audio fine. I was unable to demux it using Mac tools. So I took it over to a PC and had VirtualDub extract the WAV audio. I used Compressor to encode the video portion from PAL 25fps to NTSC 29.97fps. Separately, I converted the WAV file to AC3 with A.Pack. When I put the video/audio back together again in DVDSP 2.0, I was surprised to see that audio/video was perfectly fine in sync! I had attempted to take the audio file into Compressor as well for conversion but it didn’t recognize any of the formats I put into: WAV, AIFF, mp2...and I don’t understand why...it didn’t accept the AVI’s soundtrack...

I was worried that having treated the files separately, that the video’s conversion from 25fps PAL to 29.97 NTSC would have created a file that would have been “longer” or “rendered” in a way that would result in a mis-match with the audio because of the change in frame rate.

But marrying the two files back together worked fine, so my theory is this: In spite of an audio/video split, and in spite of the video frame rate change (and I’m sure the bit rate change as well) the running time is always the same. The video portion may have a different data rate, or a faster frame rate sequence, but its running time is always the same. So, in spite of these changes in video, the audio portion doesn’t get processed in the same way. Synchronization is achieved because of exact same running time? In short, it’s okay to re-encode the video portions separately from the audio and marry them back together again in the end of the process without worrying about going off-synch? (Leading to the question, what creates off-synch?)

And so, now with 2 other files, that are in NTSC Film (23.97fps) I had converted to 29.97NTSC using the 90minute 5Mbps Preset in Compressor but then I realized, “Oh, this is too big!” and I now must re-encode it to a smaller size, the 120min 3.5Mbps Preset. But I’ve already encoded the audio files to AC3 in the 90minute Preset. To save encoding time, I will hold back these files and not encode them with the videos.

I am gambling, once again, that in spite of the frame rate change AND the bitrate change, that if I marry the audio/video later at the end of this process, (90min Preset audio + 120min Preset video) that they will indeed be in synch?

Thanks.