progrssive vidoe gets interlace like artifact

[lexor]

http://img56.imageshack.us/img56/1680/lotrcap7pk.th.png (http://img56.imageshack.us/my.php?image=lotrcap7pk.png)

as you can see DGIndex reports it as progressive (so does the box say) it's NTSC source, plays perfectly fine with no artifacts from DVD, but when I ripped it with DVD Decripter, I get those weired lines, it's like some sort of interlace artifact. but the source is for sure progressive. I just ripped it, no editing no nothing, and opened in DGIndex (though any other player would give same problem)

again there is not a slightest hint of this if I play the dvd (even if I go manually frame by frame)

[communist]

DGIndex only reports what flag is set in the MPEG file - this however doesnt have to be true. Also DVD players (software / hardware) will either display it interlaced (on TV) or deinterlace it during playback - so you never really see the interlacing.

I *guess* its telecined and IVTC'ing it will restore the progressive frames - but thats just a guess. Anyone who has has worked with the NTSC version of LOTR can probably tell you what kind of material you have - alternatively you could post a small unprocessed sample of the vob file for people to look at and analyze.

[lexor]

hmm you are right, Telecide(order=0) fixed it right up. why would they mark it as progressive though, I mean if I can't trust movie studios, who can I trust?

[tritical]

The progressive/interlaced frame type only tells how the frame is encoded in the stream. If there are rff flags present (which I am 100% sure there are in this movie) then the result you see from dgindex using "honor pulldown flags" (i.e. the result after pulldown) will not be the individual frames as they are stored. Thus, even if every frame is encoded progressive, if 3:2 pulldown is used you will get interlaced frames after decoding.

If you want to see the individual frames as they are on the dvd before pulldown (to check if the they are actually progressive), then set the field operation to "ignore pulldown flags" in dgindex and then open the result in avisynth with mpeg2source(d2v="",info=1).

[lexor]

just did that tritical, if I'm supposed to check value of "Progr Frame" then it's always true, but "Field Order" is "bottom field first" and I thought only interlaced movies use field orders.

[cw_uk]

that interlacing looks like its just from not setting force film or also a preview has a habit of stopping on a frame like that clicking next or previous frame clears it.

[lexor]

that interlacing looks like its just from not setting force film or also a preview has a habit of stopping on a frame like that clicking next or previous frame clears it.
force film? as for preview getting stuck, I doubt it becouse that happens throughout the whole clip (I ripped chapter 2 only) with any player I have.

[tritical]

just did that tritical, if I'm supposed to check value of "Progr Frame" then it's always true, but "Field Order" is "bottom field first" and I thought only interlaced movies use field orders.
Yes, the method I said was only to give a way to see the individual frames on the dvd and see what the value of "progressive frame" is for each. This also provides a means to see if the progressive frame flag is usually correctly or incorrectly set for the movie.

There is always a field order... an interlaced display (a regular tv for example) has to know which field to display first (it doesn't matter whether the frames are interlaced or not). When dgindex or dgdecode reports the field order it is simply reporting the value of the tff flag for the first frame in the stream (0 or 1). The top and bottom fields are required to alternate (always a top after a bottom and a bottom after a top) throughout the rest of the clip.

In an ideal world, all film material (23.976fps) would be encoded progressively on the dvd and then pulldown via rff flags would be used to make it play back at 29.970 fps. Thus, if you were to view 4 individual frames before pulldown on a progressive display (computer monitor) they would appear as...

0 1 2 3 <= display frame number
0 1 2 3 <= top field (number indicates coded frame #)
0 1 2 3 <= bottom field (number indicates coded frame #)

but after pulldown (0123 trf sequence) you would get:

0 1 2 3 4 <= display frame number
0 1 2 3 3 <= top field (number indicates coded frame #)
0 1 1 2 3 <= bottom field (number indicates coded frame #)

Now you have 5 frames total (23.976->29.970), but display frames 2 and 3 would appear interlaced even though all of the actual frames on the dvd (too which the progressive frame flag applies) are progressive. This scenario, which seems very likely to be the case here, is one way to end up with dgindex reporting "progressive" and seeing interlaced frames during preview.

Under ideal circumstances, in which a dvd has trf flags that follow the 01230123 pattern the entire time and never change order, "ignore pulldown flags" and "force film" will do exactly the same thing. The difference comes when you have non-ideal situations (runs of 2's or 0's somewhere in the movie, etc...), in that case force film will remove frames to maintain correct fps but "ignore pulldown flags" will not (giving an out of sync result).

[Mug Funky]

umm... DGindex is also reporting "FILM". there's no percentage with it, so that means the entire feature is a perfect 3:2. you should be so lucky as to have more films like that.

usually "inverse telecine" is just ticked on the encoder and it's up to it to find 3:2 patterns and trust them enough to replace an entire field with a 1-bit flag.

encodes like that will usually come up as 98% FILM, but if the master has been dubbed a few times, or it's come off analog then you could get down as low as 60% film.