Captured a couple of Disney VHS direct to video releases last week...The Return of Jafar and Aladdin and the King of Thieves (I was too cheap to re-purchase these marginal titles on digital). There is quite a bit of hash in some areas, particularly the action segments, which looks like it's due to the interlacing. I plan to keep these videos in a personal digital archive; they will not be re-recorded to disc. What would our experts recommend for cleaning them up (to the extent reasonably practical, of course)?
Edit To Add: Workflow for Return of Jafar was JVC SR-VS30 >> TBC-4000 >> AIW X1800 PCIe. Workflow for King of Thieves was Mitsubishi HS-HD2000U >> TBC-1000 >> AIW 8500DV AGP.
Last edited by ehbowen; 08-04-2018 at 10:20 PM.
Reason: Forgot to add the clips!
Unfortunately they've all been converted to RGB32, so clipped brights and crushed darks due to unsafe luma and chroma levels in the original YUV color can't be repaired. If you captured originally in RGB32, consider not doing it any more. If you captured in the preferred YUY2 for VHS, avoid RGB conversion in Virtualdub by saving your edits in VirtualDub using "direct stream copy" instead of "full processing mode". This will also save storage space and uploading time -- a huffyuv RGB file is twice the physical size of the same file in YUY2.
Your videos are telecined. For any resizing or effective cleanup, you should remove telecine and restore the videos to their original 23.976 fps progressive frame rate. Telecine can later be restored during encoding for 29.97fps. Telecine is removed with Avisynth's TIVTC plugin. The usual code for cleaning up NTSC 3:2 pulldown with the functions in TIVTC is:
Unfortunately they've all been converted to RGB32, so clipped brights and crushed darks due to unsafe luma and chroma levels in the original YUV color can't be repaired. If you captured originally in RGB32, consider not doing it any more. If you captured in the preferred YUY2 for VHS, avoid RGB conversion in Virtualdub by saving your edits in VirtualDub using "direct stream copy" instead of "full processing mode". This will also save storage space and uploading time -- a huffyuv RGB file is twice the physical size of the same file in YUY2.
Your videos are telecined. For any resizing or effective cleanup, you should remove telecine and restore the videos to their original 23.976 fps progressive frame rate. Telecine can later be restored during encoding for 29.97fps. Telecine is removed with Avisynth's TIVTC plugin. The usual code for cleaning up NTSC 3:2 pulldown with the functions in TIVTC is:
For cartoons that's TDecimate(mode=1) (from the official tivtc doc,default mode is 0)
The default for TFM is not 0. The default is -1. The default applies to all videos, not just cartoons.
AssumeTFF() sets the default internal parity in the script to TFF.
From the file TFM - READ ME.txt:
Code:
B.) BASIC PARAMETERS:
These are the only three parameters most users should be concerned with or worry about
setting. They control the basic operation of the filter.
order -
Sets the field order of the clip. The field order must be correctly set for
accurate field matching to occur. Possible values:
-1 = auto (use avisynth's internal parity value)
0 = bottom field first (bff)
1 = top field first (tff)
Default: -1 (int)
From the file tivtc_tfm.htm:
I never heard of a document named " tivtc doc". Please post a copy for us. There is a TIVTC - READ ME.TXT and a TIVTC.htm, but they say nothing specific about default TFM parameters.
Here are two more clips from the original capture. I had VirtualDub on my editing computer set to save in RGB previously...an unchanged default which I just didn't catch. I believe the original was captured in YUY2, but I don't know how to check it for sure...MediaInfo says that the original capture is 4:2:2 YUV, which I'm not familiar with. So what settings would you recommend to detelecine the original capture these clips came from?
Take it easy it's more an observation than a critique i did, everybody can miss things .
So from the tdecimate.txt:
Quote:
Mode 1 = Exactly like mode 0, except instead of decimating the M most similar frames, frames are decimated from the longest remaining strings of duplicates. The duplicate detection uses the dupThresh parameter. This mode is the correct type of decimation for anime and other sources where frames are repeated 2, 3 or 4 times in a row, and also supports blend decimation of video (30p) with hybrid = 1, as well as blending of 24p into 30p (hybrid = 3). Default: 0 (int)
"mode" and "order" are two different parameters.
TDecimate "mode" has nothing to do with TFM field order.
Quote:
Originally Posted by ehbowen
Here are two more clips from the original capture. I had VirtualDub on my editing computer set to save in RGB previously...an unchanged default which I just didn't catch. I believe the original was captured in YUY2, but I don't know how to check it for sure...MediaInfo says that the original capture is 4:2:2 YUV, which I'm not familiar with. So what settings would you recommend to detelecine the original capture these clips came from?
Your two capture samples in the post quoted above are YUY2. If you want more detail about the exact name of the colorspace use Avisynth's info() function:
Code:
AviSource("path:\to\video.avi")
info()
Don't use VirtualDub's file information menu for this. VirtualDub will always say it's RGB.
The telecine pattern in the samples is plain vanilla 3:2 pulldown. To remove telecine fields:
Code:
AssumeTFF()
TFM().TDecimate()
or:
Code:
TFM(order=1).TDecimate()
Both ways of coding the above mean the same thing for inverse telecine.
