I used this script to deinterlace my DVD but there are still some remaining interlaced frames?
SetMtMode(5,10)
LoadPlugin("C:\Users\Ryan\Downloads\video editing\MeGUI_2418_x86\tools\dgindex\DGDecode.dll")
LoadPlugin("C:\Program Files (x86)\AviSynth 2.5\plugins\TIVTC.dll")
LoadPlugin("C:\Program Files (x86)\AviSynth 2.5\plugins\MSharpen.dll")
DGDecode_mpeg2source("C:\Users\Ryan\Desktop\VTS_03 _2.d2v")
SetMtMode(2)
AssumeTFF()
TFM()
TDecimate(mode=2, rate=23.976)
Crop(2, 0, -2, -0)
Spline64Resize(720, 544)
Try StreamFab Downloader and download from Netflix, Amazon, Youtube! Or Try DVDFab and copy Blu-rays! or rip iTunes movies!
+ Reply to Thread
Results 1 to 26 of 26
Thread
-
-
I don't see any interlacing in that picture but I'll take your word for it that it's there. What happens when you do this instead:
TFM()
TDecimate(Mode=1,Cycle=25,CycleR=1) -
Read the TFM settings and raise the comb detection settings. MI and cthresh are the options you want, lower them until the combing goes away. (you could try slow=2 as well, PCs aren't exactly abacus' these days...)
Crop(2, 0, -2, -0)
Spline64Resize(720, 544)
And yes, getting the frame-rate right is important. -
Manono you fixed the frame thanks! And mdjamena I'm a little bit confused on aspect ratios and re-sizing and what not. I read that you're suppose to crop out black bars and resize the DVD footage to square pixels for computer viewing.
-
Oh, OK, you're shrinking it to fit the aspect ratio. I normally upsize to 768x576 or 1024x576 but ok. Actually, you don't have to resize to square pixels, most (all) modern containers can handle anamorphic video streams, and most players can play them fine, but you can if you like (although you're losing resolution doing it the way you do).
-
I think what you have is 23.976 fps with pulldown to 25 fps. Use TFM().TDecimate(cycle=25, cycleR=1). That will give you 24 fps -- probably with a duplicate frame every 1000 frames or so. Or use TFM().TDecimate(mode=2, rate=23.976) or some variation.
As ndjamena pointed out, you may need to tune the interlace detection sensitivity in TFM. -
It alternates between in and out of phase because 2:2:2:2:2:2:2:2:2:2:2:3 pulldown was used. That turns 12 frames into 25 fields, 24 frames into 25 frames. If you only use TFM() you will have a duplicate frame in every 25 frames.
-
Anonymous344Guest
Oh, I see. It has a euro pulldown, but it's not been applied by simply doubling every twenty-fourth frame but by inserting fields (one every half second?). Is the second option in post #6 meant to give a 24fps (as opposed to 23.976fps) result?
-
Yes.
The second script should give 23.976 fps. It's hard to tell whether the restored rate should be 23.976 (NTSC film) or 23 (film) -- a difference of only 1 frame in every 1001. The sample is too short. And with animations there are so many duplicate frames it's hard to tell even with longer clips. -
Anonymous344Guest
Yeah. The fact that it was animation and had duplicates anyway threw me: I should have thought to check for a euro pulldown. But I can see it now, most noticeably in the part at the end with the cat.
It's strange, but TFM().TDecimate(mode=2, rate=23.976) gives me a 24fps result i.e. 24/1, not 24000/1001. -
-
I just downloaded the sample to make sure it was actually pal and I wasn't being an idiot about the resize. According to MediaInfo the sample is encoded progressive at 24 fps, and if I load it into virtualdub through an AVS script with MPEG2SOURCE dishonouring pulldown flags I can't see a single combed frame. So I'm a little confused at the moment. MediaInfo says the sample is 720x576 and since the OP has resized his encode to 720x544 I'm assuming it must be the source. What is this thread about exactly???
-
Anonymous344Guest
Are you sure it's 100% soft pulldown? DGIndex says it's 83.29% VIDEO. Even in "ignore pulldown flags" mode, there are some frames that seem to be composed of one field from each adjacent frame e.g. 3, 394. 394 looks messed up even if you separate fields though.
Last edited by Anonymous344; 22nd Apr 2014 at 07:23.
-
83.29% VIDEO might explain the 25fps and there is a lot of shifting of the picture that might be considered an artefact of interlacing, except for the fact that some of that shifting is horizontal, which I don't believe can be caused by any type of interlacing.
-
Frame 394 suffers from encoding or decoding errors, not interlace combing. I saw no comb artifacts in the clip when ignoring pulldown flags.
-
I was thinking last night that I should check to see if the sample used soft pulldown. But I had never seen that used in PAL video so I didn't bother.
-
Code:
MPEG2SOURCE("sample.d2v") FFT3DFilter(sigma=1.5, bt=5, bw=32, bh=32, ow=16, oh=16, sharpen=0.4, plane=4) NNEDI3_Resize16(786, 576, 3, 2, -2, -2) FFT3DFilter(sigma=0.5, bt=5, bw=32, bh=32, ow=16, oh=16, sharpen=0.2, plane=4) Dup() AssumeFPS(24.000)
-
Grrr, .sup was the extension I needed, however the output file isn't compatible with MKVMerge, SubExtractor wouldn't read it so I used SubtitleEdit and downloaded a French dictionary for it. I converted it to srt and from there to a proper sup, but the durations are missing and I think I've already taken this far enough...
-
French? Looks Spanish to me. But he could probably download srt subs from various sub sites. I'm sure this has been translated into many languages