25fps interlaced to original 24fps?

[zeronegative]

Can anyone tell me if there is a good (or even foolproof) method to get my 25fps interlaced PAL rip back to the original 24fps? They used some kind of trick to add extra frames instead of stretching the movie, because the frame interlacing looks a bit like 3:2 pulldown to me. Can Decomb handle this job?

[hakko504]

Decomb is very well suited for the job:Telecide()
Decimate(cycle=25)

[DJ Bobo]

AFAIK, no way to get progressive 24fps.
Just use

FieldDeinterlace()

to get blended progressive 25fps.
There is no other solution.

[zeronegative]

I did a rip with the FieldDeinterlace() option; the result is 'okay', as in, you can watch it but it is still a bit dirty. It is obvious that the movie was transcoded from 23.978 to 25 fps by adding frames instead of adjusting the length. I'm going to give the other option (telecide/decimate) a shot and see if it solves anything, if not... bad luck for me :)

[Wilbert]

You could do something like this:

Fielddeinterlace # or GreedyHMA(1,0,0,0,0,0,0,0)
Decimate(cycle=25)
AssumeFPS(24) # or AssumeFPS(23.976)

[Dreassica]

Hmm, i hope for your sake u arent alking about an anime DVD, and especcially not madman ones, otherwise u are not lucky. Cause those cannot be properly deinterlaced without getting ghosting and blended fields etc.

[htc10825]

Can anyone tell me if there is a good method to get my 25fps interlaced PAL rip back to the original 24fps?
I suppose you meen the movie rip/avi(PAL, interlaced) from internet. Then it is impossible unless you wait util the official DVD releas.
The most movie rips(PAL, interlaced) of new movies from internet are based on the source material of "Screener", i.e. they were capture at a sliver screen with a PAL DV-Camcorder and then transcoded to (mostly) DivX5 with a very bad quality(very unsharp, brightness/gamma settings totally wrong). The best example is "Spider-Man"(German 1CD-rip DivX5 448x360).
Theoretically you can recover it to original progressive 24fps(25fps with acceleration). But in practice they all did a resize in vertical resolution which make the restore to progressive very very difficult or impossible.

[zeronegative]

Actually, I'm not talking about anime nor internet rips. Tartan released a bunch of my favorite classic Hershell Lewis gore movies on DVD. They're region 2 (PAL 25fps), but the originals were 23.976; taking a look at the frames, not every frame is interlaced; just a bunch. That was why I was wondering if it would help getting the movie back to the original fps.

[Kb_cruncher]

you might also try TomsMoComp(1,9,1),it seems to work better than FieldDeinterlace.

AssumeFrameBased.SeperateField.SelectEven
Mpeg2Source("yourmovie.d2v")
WhateverResize(#,#,#,#).bob #you need to half vertical res and
vertical crop values here.

Also works well but this discards one field so is less sharp but can be made to look close to the original with sharpening.

[zeronegative]

After taking another good look at the source, this is what I found out:

There is an inconsistent pattern of blended & clean frames; I took a look at 75 frames, and completely random most of the frames (75%) are a mix of the current frame + the next frame, while only about 25% seems to be 'clean'. Like I said, the pattern is completely random (no alternating or whatever).

In other words, I don't think there's a lot I can do with it to make it look better. It's just a crappy transfer, right?

[hakko504]

Nope, there isn't much you can do. But take a look at this article (http://www.doom9.org/ivtc-tut.htm), the section about PAL, the Ugly. At least it explains how this kind of thiungs can happen.

[zeronegative]

Thanks; I'll give it a shot with Telecide(post=true,blend=false) followed by SpatialSoftenMMX(2,5,7) and see if the result is watchable :)

[iago]

...I'll give it a shot with Telecide(post=true,blend=false)... (zeronegative)

@zeronegative

If you are using Decomb 4.00, "post=true" is the default, as long as you don't set the "post=false" parameter for Telecide. Therefore, Telecide(blend=false) in your script would be enough imho.

regards,
iago