It's PAL 25i. Multiple sources so I'm not sure what was originally NTSC. I wanted to simply convert to 23.976p with the sound stretched to correct the PAL pitch and tempo to NTSC. If possible, I wanted to remain in MPG2 and keep it progressive (I've used DGPulldown to set flags to make the player do the 2:3 pulldown).
Here's where I'm at:
The original video is at about 3.5 kbits/sec and sound is AC3 192 kbps. I used TMPGenc 2.5 to deinterlace the 25fps to about double the original bitrate, then ran it through again to convert to 23.976fps and back to 3.5 kbits/s. I then used DGPulldown to flag it so that players will see it as 29.97i. For some reason, TMPGenc seemed to have rounded it up to 23.98 so the soundtrack I'd converted on an earlier attempt using VirtualDub (HuffYUV) didn't work anymore. Converting the soundtrack from 25fps to 23.976fps caused synch to drift. It was like it was too long. I had decompressed the AC3 to wav and used GoldWave to stretch it to the exact length of the movie, which caused the same issue. I then tested a demuxed copy of the original PAL version and noticed that the soundtrack didn't quite line up with the video in the first place, even though the synch stayed perfect. I then went back to my NTSC version and stretched the wav to the same length as the video, minus a few seconds, which seemed to help. I have about 4 or 5 test scenes (beginning, pre-middle, middle, post-middle, and end) marked in Womble that I use to check synch, so when I finish a new wav I plug it into the timeline and test it. I got it to within a few frames (I wish Womble had a visual wav line so I could see peaks rather than listen for them) and then would tweak it a few thousandths at a time to get it closer and closer.
The rest of the DVD material, I just used either Procoder or NeroRecode, which both give smooth non-jerky playback while allowing use of the original sound. I didn't want to alter Jambe's work in the extras at all. I only wanted to convert the main film with sound to NTSC speed and pitch. Only real issue is that a couple scenes of commentary were not sped up on the original PAL, so the Empire Strikes Back sound is good, but then a couple times a slowed down Mark Hamill or Carrie Fischer will make a comment. It's an unfortunate compromise. When I get the entire thing encoded, I'll check out both versions and see which is better: Chipmunk movie or normal movie with the occasional s-l-o-w-e-d down comment.
Anyway, that's where I'm at. If there's a more efficient way, I'd genuinely like to know. I couldn't find any guides online that did what I wanted it to do. Most used methods that jerked, or didn't alter the sound.