satanika said:

CapableMetal said:

Interesting point, and certainly worth a try. The stretching is certainly done at 0.1% to match 23.976fps but seems to need very fine tuning as it (thus far) hasn't quite seemed to line up properly.

There could be small frame differences from the theatrical compared to the source you're using.

The quality of the rendered stretch in audition is actually very high, I haven't heard any artifacts on my test mixdown, and it preserves exact pitch too, although thats probably a silly thing to be worrying about as it'll only be 0.2 semitones out.

I'm sure it's fine :) -- 'possible artifacts'
However, less processing is usually better & I believe the standard way of dealing with this is simply slowing the audio down.

I agree with you completely satanika, less processing is better. I used your suggestion and it works very well, syncs almost perfectly with the source I've used and was actually less hassle in the end. Whether it sounds better is subjective and whilst doing this I've come to the conclusion that source audio is less than perfect, but still very good. I'm hopeful that ROTJ will be better (I've yet to try it).

To make the source I layered and compared 5 or 6 different 1997 SE sources to find that all were missing a frame or two (or 15ish in the Reivax!) here and there, but in different places in the film and that together they added up to the exact number of frames in the 2004 release I matched it against, so I'm fairly confident its accurate now. My early attempts were against my NTSC laserdisc capture, but after IVTC a few frames were missing that shouldn't have been so that quickly came to end ;)

The reel changes were perfectly in line with the reel change points Darth Mallwalker posted in the standards thread which means it should should sync to the 2004 NTSC release and the 1997 video source msycamore has uploaded (which I haven't actually seen yet, but numbers correlate so should work).