Another IVTC Question: PAL -> NTSC Back to PAL?

Author: Doctor M
Date: 22-Sep-2006, 7:28 PM

Author: Doctor M
Time: 22-Sep-2006 7:28 PM
Post link

I have some material that was originally 25 fps, but they did that god awful conversion to 29.97 using 100%(?) interlacing instead of just adding on some pulldown flags.

Is it possible to restore the original 25 frames from that mess, or once they've gone that far is it garbage?

Dr. M

Author: Karyudo
Time: 22-Sep-2006 9:56 PM
Post link

There are about a half a zillion posts on this very topic over at Doom9.org. The very first step there (or here) is to post a bunch of consecutive frames, so that one can see exactly what's going on.

But in general, if something's been f'ed up well enough, it's virtually unrecoverable.

Author: Doctor M
Time: 22-Sep-2006 11:18 PM
Post link

I tried googling this, but I'm not sure what to look for it as.

You see a lot of Pal to NTSC, NTSC to Pal, but I really don't know what this is called in general.

Dr. M

Author: Doctor M
Time: 23-Sep-2006 12:56 AM
Post link

Hmm, found a couple threads.

What now has me confused is using Bob I end up with (about) 60 fps of unique frames. (No duplicates, stuttering, ghosting, nada).
I wouldn't think that is possible if it was a 25fps source. The runtime seems to be exactly the same between the original PAL and the NTSC. What the heck did they do?

Edit: Heh heh, playing around and used:
Bob()
selectevery(2,0)

I ended up with beautiful looking 29.97fps progressive video. No duplicate or interlaced frames.
Looks better than the original.

Dr. M

Author: Doctor M
Time: 23-Sep-2006 1:36 PM
Post link

I know I'm talking to myself, but if anyone else finds this thread with a similar question.
A great filter chain for this problem is:

bob
repal()

Repal is a nice plugin that is designed exactly for this sort of problem (a PAL source that has been interlaced to 29.97fps and you want to restore the original 25 frames). It takes the 60fps bob generates and calculates/keeps the original 25 out of it.

Dr. M