Okay, so i got a recorded version of an Episode.
And a DVD version of the Same.
One is PAL (the recorded) the other is NTSC. 29.970 i think.
But the Recorded have a different language(hence why i want to use the DVD picture), is is also a bit faster or slower.
I think itīs cause of the NTSC to PAL convertion.
I thought i would be able to solve it with ease, but i was wrong.
So i wonder if anyone here can help.
If you want, i can upload a clip from the same scenes.
+ Reply to Thread
Results 1 to 30 of 53
Last edited by zerowalker; 12th Feb 2013 at 18:07.
I think i may have solved it. I used indexing on the Video_TS, then TFM and TDecimate to make it 23.976.
After that it seems to be synced, atleast when i played around a bit.
Okay, now i wonder, is there a way to convert NTSC 23.976 to PAL?
I read about many different ways, but wonder if there is a simple, like ChangeFPS(25) (wich didnīt work, it went slower i think;S)
I donīt want to change the speed of the video, just add a Frame per sec.
But better for your purposes would be to encode at 720x576 and progressive 23.976fps. Afterwards run DGPulldown set for 23.976->25.000fps to make it true PAL.
Well, it didnīt seem to work, when i checked with the other file (Stackehorizontal). though the other file was 24.999 for some reason;S
But that seems to be what i want.
Should i deinterlace before or during play?
Edit: Well deinterlacing doesnīt seem to work during play. Is it because it was changed to 23.976 from 29.970?
Last edited by zerowalker; 13th Feb 2013 at 00:26.
...though the other file was 24.999 for some reason
Itīs not really progressive, not all the time, itīs some Interlacing here and there at small areas. Like when someone is standing still talking, his mouth can become interlaced.
Okay, well yeah doubt it as well, must be a bug from when i clipped it or something.
Yepp, Thanks will try it
A good thing to try is lowering the CThresh a bit. From the TFM manual:
This is the area combing threshold used for combed frame detection. It is like
dthresh or dthreshold in telecide() and fielddeinterlace(). This essentially
controls how "strong" or "visible" combing must be to be detected. Larger values
mean combing must be more visible and smaller values mean combing can be less
visible or strong and still be detected. Valid settings are from -1 (every pixel
will be detected as combed) to 255 (no pixel will be detected as combed). This
is basically a pixel difference value. A good range is between 8 to 12.
Default: 9 (int)
So, try 8, or maybe 7 as a last resort before just throwing Vinverse in there. You can check to make sure 'good' frames don't get deinterlaced at the same time by putting on 'Display=True'.
Okay, think i got it to work. I played around a bit, it seems that Chroma=True did alot of help.
But i have a fast question i have wondered about for some time now.
If a video is Interlaced, and some parts are not Full Interlaced while others are.
What is the best way to approach it?
I usually goes with QTGMC() SelectEven/odd().
But that will ruin the parts where itīs fully interlaced, as it will skip every other frame right?
And if i just use QTGMC(), i get 50/60 fps at all time, but that will be a waste on parts that arenīt fully interlaced i think.
You're still talking about anime? I doubt I've seen (and doubt there exists) any anime that's fully interlaced in parts unless it's end credits or unless it's field-blended. If it's a movie, again, no such thing.
Sorry, should have been more specific.
I am talking about, different parts.
Well for example, in a VHS tape, i got,
First a Cartoon, then a non-Cartoon (which seems to be fully interlaced).
So the cartoon is interlaced (doubt itīs fully, see no reason for it), and the non-cartoon is Fully.
From what, i would say itīs shifting from, the cartoon which is Interlaced, but not Pure. And then later on changing to Pure Interlaced.
OK, then I'll repeat:
How to determine field order, interlace, telecine, etc.: http://neuron2.net/faq.html .
The cartoon is likely from film or is progressive+pulldown, so is probably telecined. The non-cartoon is probably straight interlaced. Check the link to see how to view video to analyze it.
Last edited by sanlyn; 16th Feb 2013 at 12:33.Our inventions are wont to be pretty toys, which distract our attention from serious things. They are but improved means to an unimproved end. -- Henry David Thoreau
Okay, can say that i ended up using AnimeIVTC, as the name says, itīs for Animé, and does the jobb very well.
And sanlyn, i still donīt get how i am supposed to work with such videos?
Cause the Cartoon, as you say, is just 25 fps (or 23.976 on NTSC). And the Real World stuff, is 50 fps (60 for NTSC).
There is an intro i will have to capture to show it, where they use Real stuff and Cartoon, and the Cartoon looks good, but the Real stuff become, unsmooth.
Well, i can just upload it here to show it.
Sorry for making it MKV, but didnīt find a way to remove the Audio from an MPG file, so used Mkvmerge to remove it.
This is diagnostic, not prescriptive.
What you have here is a tape that started life as NTSC 29.97 master incorporating live action and animated film elements (probably 24fps film), some 3d elements too -- but it was, as I said, almost certainly mastered at 29.97. Before it came to you, someone converted it to PAL in some (likely) 80's era way. So your tape is a PAL version of an NTSC show with mixed elements.
You are essentially facing the same reconstruction task as Star Trek TNG, but with a VHS PAL generation thrown in to make it interesting.
How far are you willing to go to tease these elements apart, because you can't do it in a single pass?
I personally wouldn't spend my time attempting this, but I hope the diagnosis can be helpful.
VirtualDub to view frame-by-frame, one frame at a time.
IF the source is MPEG, you can extract the video as .m2v using the DGindex utility from DGMPGDec. You've used it before, so I don't know what the problem is to get a short MPEG2 video clip from your source.
Last edited by sanlyn; 7th Mar 2013 at 14:09.Our inventions are wont to be pretty toys, which distract our attention from serious things. They are but improved means to an unimproved end. -- Henry David Thoreau
DGIndex and use [ and ] to trim off a piece. Then File->Save Project and Demux Video. The resulting M2V can be uploaded.
It's field blended from an NTSC source, The animation is a mess and I can't fully unblend it because so many frames don't have any clean fields at all. The live action stuff does better:
Well itīs pretty clear yes.
But i donīt know how to handle it.
Am i supposed to let for example. QTGMC make it fully progressive, even when there are just some parts that are 50fps, or should i let it remain as Interlaced?
Cause normally i use QTMGC() with SelectEven().
But here is the M2V file, thanks for telling me how to do it.
And well, can say that all these cartoons (Captain N, Zelda etc) are totally blended, both the NTSC and PAL, but PAL is probably worse. Or the same, depends if they just speeded it up to PAL speed or added frames ant stuff i guess.
Now I don't know if it's really supposed to be 25fps or not. Sometimes I think that looks better, with fewer frame-jumps. If so, it would be like this:
...even when there are just some parts that are 50fps
That would mean some was shot on video. Is the live action native PAL? But it has blends and duplicate frames too. Where's this video from? So only the animation stuff is converted from NTSC?
...or should i let it remain as Interlaced?
If for DVD, that's one alternative - leave it alone. And seeing how messed up this thing is, I think I might do it that way myself if it were mine and I lived in a PAL country.
Both the cartoon and live are convertet from NTSC to PAL.
So i canīt really use SRestore to get 25, as that would ruin the Live part right?
And itīs "overkill" to make everything 50fps, as the Cartoon is just 25.
The only thing i can make out of it, is to let the Player Deinterlace.
What are the suggestions then if i wanted to work with it in Avisynth?
(No need for filtering and scripts, just want to know if i am supposed to temporarily deinterlace it and then clean it up and stuff like that=
So, i can just use Bob(), then clean it as with any Progressive material. Then reinterlace as if nothing has changed?
Is that the way to go?
Thanks, really helps.
I wonder why these are being converted NTSC to PAL. Don't European players handle NTSC? There are USA players that can handle both.Our inventions are wont to be pretty toys, which distract our attention from serious things. They are but improved means to an unimproved end. -- Henry David Thoreau
Is Yadif mode 1, making it 50p?
cause i always use QTGMC().
And well, hmm, has anyone done this before and can give some clue to what should be working?
As i guess you manono hasnīt done it, as you also are unsure.
If you are talking about DVD, then the PAL speed convertion is just totally worthless really, i have no idéa why they do it, they can just add frames like they could on the VCR but didnīt most of the time, and i think they could have used NTSC as i think DVDs are supposed to be compatible with each other, except for regions.
The only really nice thing about PAL is the extra resolution, itīs quite a bit from NTSC, but sadly all the trouble with framerate conversions doesnīt make up for it.
cause i always use QTGMC().
As i guess you manono hasnīt done it, as you also are unsure.
But, the thing about how to filter a video that is both pure interlaced and not.
As you said you are unsure if the fields are unchanged when using bob().
But i donīt see the reason why Yadif would be different. QTGMC() on the other hand does filtering on itīs own, and isnīt direct approach to make it 50p just for cleaning up.
Or maybe i am just mistaking what you are saying though, if so i am sorry.