Best way to detelecine in Avisynth & a few examples of scripts I've tried.

[LaserBones]

When I detelecine is it a form of deinterlacing or do the interlaced field both form an actual frame and not an interpolated frame? I hope that make sense. I'm still learning.

I've been using the following script to detelecine but does it try to deinterlace every frame?

D = QTGMC(Preset="Slower", FPSDivisor=2, EdiThreads=3)
TFM(slow=2, cthresh=4, mi=0, pp=7, clip2=D)
tdecimate()

I was recently told to use the following script for deinterlacing only frames that needed to be deinterlaced. Should I be using something more similar to this?

#q = QTGMC(preset="Placebo", EdiThreads=4, NoiseProcess=0)#.Subtitle("QTGMC") # for debugging purposes, remove subtitle for production
#d = ChangeFPS(50)
#ConditionalFilter(d, q, d, "IsCombedTIVTC(cthresh=4)") # requires TIVTC package

My main question is how do I detelecine at the highest quality in avisynth?

[johnmeyer]

Deinterlacing and inverse telecine ("detelecine") are two completely different things, although they are often confused. Deinterlacing attempts to create new frames of video where both the odd and even scan lines are from the same moment in time (interlaced video has half of the scan lines from one moment in time, and the other half from the next moment in time).

Inverse telecine (IVTC) removes repeated fields which were added to allow 24 fps film material to play at 30 fps. The smoothest way to create telecine is to repeat fields, not entire frames. It is these repeated fields that must be removed.

Unlike interlaced video where odd and even fields are from different moments in time, telecine exactly repeats, as necessary, a previous field. During IVTC, the excess fields are discarded and you end up with the original 24 frame per second film, without losing any quality whatsoever (other than having to re-encode the result). By contrast, with deinterlacing you always end up with some compromise in the resulting video because you either have to repeat and or re-position fields, or you have to attempt to generate new fields via blending or motion estimation.

Bottom line: if you have 30 fps video of what was original 24 fps film, and you want to get back to that pristine 24 fps (which is a good thing to do), then all you need is TFM followed by TDecimate, usually without any parameters at all.

You do not need QTGMC, which is a deinterlacer, not an inverse telecine tool.

[jagabo]

But when there are orphaned fields (very common when shows are edited after being telecined) using QTGMC for the deinterlacing (as in his first script) is usually better than using TFM's built in deinterlacer (which is more like yadif).

[manono]

My main question is how do I detelecine at the highest quality in avisynth?

In the AviSynth world it's usually called InVerse TeleCine or IVTC.

And you use the TIVTC filter with - as jagabo suggests - QTGMC deinterlacing anything left over after the field matching that might need it. As in your first script. I can't imagine that second script being needed for anything having to do with an IVTC.