Looking for Pulldown code that doesn't cause detail loss.

[Jeremy Duncan]

LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\DGDecode.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\mt_masktools.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\RemoveGrainSSE2.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\RepairSSE2.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\Fluxsmooth.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\MT.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\decomb.dll")

mpeg2source("C:\Users\JD\Documents\Unleashed\VTS_01_3.d2v")

SetMemoryMax(1024)
AssumeTFF()
Telecide(chroma=false)
decimate()
aWarpSharp(depth=20, blurlevel=2, thresh=0.21, cm=0)
Crop(0,56,-0,-62)
FluxSmoothT(3)
MT("
a= last
b=a.Spresso(limit=1, limitC=1, bias=27, biasc=27, rgmode=17, rgmodec=17)
SeeSaw(a, b, NRlimit=0, NRlimit2=5, Sstr=0.90, Szp=1, Slimit=9, Sdamplo=16, Spower=1, sootheT=95, SdampHi=39)
",3)

So, here is my code I use.
I use virtualdub, and I encode using this code.
I also sometimes play movies using ffdshow in realtime.
I prefer to encode as I like to crop this way.

I need to use pulldown on my ntsc dvd's using my preferred encoding method.
I tried ignore pulldown using dgindex and it made the sound all messed up with the lipsync.
So I need to use pulldown using my method.

Here is a sample: Link (http://rapidshare.com/files/102714765/Sample.zip.html)

The sample shows a man's face on the first screen.
Using this code and lanczos resize. When I use pulldown, I see the facial details get dumbed down, when I remode the decomb pulldown the details reappear, viola!

I am interested in suggestions on how to get pulldown that doesn't lose detail.
Thank you.

Edit. Here's some sample comparison pictures comparing pulldown and no pulldown.
Link (http://rapidshare.com/files/102716777/Using_pulldown.zip.html)
For these pictures, I remove the crop code and used ffdshow and mpc.

[FlimsyFeet]

Have you tried using Forced Film in DGIndex?

[Mug Funky]

have you tried comparing without all those denoisers destroying your picture?

my bandwidth is very low, so i haven't downloaded the samples, but my guess is simply turning off telecide/decimate in that chain will severely limit the success of the denoisers, letting through more of the original details.

a spatial denoiser will do much less on an interlaced image because it is inherently more "detailed", and a temporal denoiser will be less effective on 30fps than 24fps for the same temporal radius (in frames).

[Guest]

First, you are talking about IVTC, which is the reverse of pulldown. Saying you need pulldown shows that you don't fully understand what you are doing.

Your clip is soft telecined, so you can use Force Film in DGIndex. If you had read the manual you would know that Ignore Pulldown is intended for diagnostic purposes and not encoding, and that it can cause audio sync problems.

[Jeremy Duncan]

neuron2,

Using this code in ffdshow avisynth:

AssumeTFF()
Telecide(chroma=false)
decimate()


I have to check use pulldown.
It's not called ivtc, so I always assumed it was pulldown I was using with this code.
Thank you for showing it's ivtc.

I will now try your method of using force film and see if it works with lip sync or not.

Thank you both for your kind replies. :)

[Jeremy Duncan]

The Force Film worked on the movie Unleashed.
But on the movie The Mist, 2007, it did not.

How it failed is lip sync.

I will post a link to a clip with sound.
Link (http://rapidshare.com/files/102966625/Sample.zip.html)
I added a fixed d2v that was in the d2v that had the bad lip sync, it is not from the sample I uploaded, two different d2v's.

This is the code I used in virtualdub:

LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\DGDecode.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\mt_masktools.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\RemoveGrainSSE2.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\Decomb.dll")
LoadPlugin("C:\Program Files\AviSynth 2.5\plugins\MT.dll")

mpeg2source("C:\Users\JD\Documents\The Mist\VTS_01_1.d2v")

SetMemoryMax(1024)
MT("LimitedSharpenFaster(ss_x=1.4, ss_y=1.4, Smode=3, strength=175, overshoot=0, soft=60)",3)

So, it looks like I need to use honor pulldown in d2v if I want to get lip sync without guessing it forced film will work or not with lip sync.

So, I am hoping for some suggestions on how to use honor pulldown in d2v, and ivtc without losing detail. :)

[zee944]

So, I am hoping for some suggestions on how to use honor pulldown in d2v, and ivtc without losing detail. :)

This is the code I use regularly to convert telecined 29.97i back to progressive 23.976. Before opening the script I index the m2v file with the default settings (not Force Film mode).

LoadPlugin("DGDecode[1.4.3].dll")
LoadPlugin("Decomb[5.2.2].dll")
MPEG2Source("City_Hunter_Fox.d2v")
AssumeTFF()
Telecide(guide=1)
Decimate(cycle=5)

It's just a simple inverse telecine method, can't see any detail loss or sync problems, it always works on proper sources. IMO Unleashed was shot on film. If this code fails on this movie, then the DVD is probably a screwed up conversion.

[Guest]

Force Film does not change the audio sync.

[Jeremy Duncan]

I tried to encode it again, and I got bad lip sync again, so i then tried to encode it using honor pulldown and the ivtc code I posted and I still had bad lip sync.
So I just watched in in ffdshow and mpc without encoding it. :)

It's a very very scarry show.