Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > Capturing and Editing Video > Avisynth Usage

Reply
 
Thread Tools Search this Thread Display Modes
Old 22nd September 2012, 00:35   #1  |  Link
OvejaNegra
ekTOMBE STUDIOS
 
OvejaNegra's Avatar
 
Join Date: Dec 2005
Location: Cuba
Posts: 253
Very weird problem decoding videos using MPEG2Source

Before: I post this here because it seems the more proper place, but the mods may move it to other part of the forum if it fits more.

Im decoding a dvd using MPEG2Source and tivtc after like this:

Code:
MPEG2Source("C:\STRW\01.d2v", cpu=0)
interp = separatefields().eedi2(field=-2)
deint = tdeint(mode=2,edeint=interp)
TFM(mode=2,pp=1,display=true,slow=2,mChroma=true,micmatching=2,cthresh=1,chroma=true,d2v="C:\STRW\01.d2v",micout=2)
(PP is 1 so no post processing to the interlaced frames)


I notice that there are alot of frames detected as interlaced but visually i dont see any combing artifact. If i disable chroma for interlacing detection (chroma=false) most of the false detections go away.

But i decided make a 200% zoom and, in fact, there is a strange problem with most of the frames:

On frames with motion (pans for example) its like parts of the image

start moving before the rest of the frame. Yes its wierd and i dont know how to explain properly (watch the screenshots).

But the thig is, that the parts of the image that move "first" only move on one of the fields of the frame. That creates lots of parts of the picture with combing artifacts.
Dont know if its the source or the MPEG2Source plugin.

Here are two screenshots ( frames 8790 and 8791 taken from avsp). Save both images and alternate between one and another (use a tool with point sampling zoom like the picture viewer of windows, billinear zoom like acdsee hides the problem.) when changing from one screenshot to the other, both may look the same, but look carefully. now imagine that on allmost all the moving frames. (look carefully at her dress, her right arm).

The third screenshot (8792) its actually a moving frame, but you can see more clearly the combing artifacts on the picture, look at her dress again.

the las screenshot shows the frame pure (without tivtc)

Does anyone knows what is this and how can i fix it? Changing different modes of matching on tivtc does not solves the problem

One more thing:
using
bob()
selecteven() (or selectodd)
on the source makes the problem more visible.








Sorry for my english.
Thanks.
__________________
So, it works or not???
OvejaNegra is offline   Reply With Quote
Old 22nd September 2012, 00:55   #2  |  Link
Guest
Guest
 
Join Date: Jan 2002
Posts: 21,923
Post a link to an unprocessed source sample that shows the issue.
Guest is offline   Reply With Quote
Old 24th September 2012, 17:42   #3  |  Link
OvejaNegra
ekTOMBE STUDIOS
 
OvejaNegra's Avatar
 
Join Date: Dec 2005
Location: Cuba
Posts: 253
i cutted a sample using dgindex / demux video. Where should i upload it?
__________________
So, it works or not???
OvejaNegra is offline   Reply With Quote
Old 24th September 2012, 17:59   #4  |  Link
Guest
Guest
 
Join Date: Jan 2002
Posts: 21,923
mediafire.com

Upload it and post the link they give you here.
Guest is offline   Reply With Quote
Old 12th October 2012, 00:34   #5  |  Link
OvejaNegra
ekTOMBE STUDIOS
 
OvejaNegra's Avatar
 
Join Date: Dec 2005
Location: Cuba
Posts: 253
sorry for the very long delay but i'm very busy with my work right now.
here is the sample, i used sendspace because mediafire request an account to upload:

http://www.sendspace.com/file/xsex98

Thanks
__________________
So, it works or not???
OvejaNegra is offline   Reply With Quote
Old 12th October 2012, 02:28   #6  |  Link
Guest
Guest
 
Join Date: Jan 2002
Posts: 21,923
Looks like you are shooting yourself in the foot with that script. With this simple script:

telecide(post=0)
decimate()

I see no such problems.
Guest is offline   Reply With Quote
Old 12th October 2012, 18:24   #7  |  Link
OvejaNegra
ekTOMBE STUDIOS
 
OvejaNegra's Avatar
 
Join Date: Dec 2005
Location: Cuba
Posts: 253
i tested your solution and the problem persist:

MPEG2Source("C:\v send\VTS_01_1.d2v", cpu=0,idct=5)
Telecide(post=0)

If you use avpsmod, zoom the frame (200% or more), seek to frame 988 (maybe you have to turn up a little the brightness of the monitor) and step frame by frame and you'll see the problem.

Look at her black dress, it's like some residual combing or something like that, the funny thing is that the bombed pixels some times move even if the frame is a duplicate. Because of that i get too much false positive on the combed frame detection.

Maybe the authoring house is using a really crappy mpeg2 encoder. As long as you can confirm that is not a bug on MPEG2Source, i'll just leave it as is (or maybe you have some nice trick you can suggest

Thanks
(The DVDs are from Media Blasters)
__________________
So, it works or not???
OvejaNegra is offline   Reply With Quote
Old 12th October 2012, 20:06   #8  |  Link
Guest
Guest
 
Join Date: Jan 2002
Posts: 21,923
Sorry, not seeing anything. Can you post a screenshot showing what you object too?
Guest is offline   Reply With Quote
Reply

Tags
interlace artefacts ntsc, tivtc

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 00:41.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2019, vBulletin Solutions Inc.