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. |
24th March 2024, 19:16 | #1 | Link |
Registered User
Join Date: Aug 2021
Posts: 6
|
Avisynth QTGMC please help me!
This is an interlaced video that is not a DVD. I get this result with Avisynth QTGMC, what is the reason? and is there a solution? Thanks in advance.
Sample: https://mega.nz/file/lUdQUJYS#Eg0gY8...B8LFvidKgjvw8Y Last edited by puppeteer; 24th March 2024 at 20:50. |
24th March 2024, 19:53 | #2 | Link |
Registered User
Join Date: Aug 2017
Location: Italy
Posts: 113
|
Post the source
__________________
A channel on S-VHS / VHS capture and AviSynth restoration https://www.youtube.com/channel/UCMs...h1MmNAs7I8nu4g |
24th March 2024, 19:54 | #3 | Link |
Broadcast Encoder
Join Date: Nov 2013
Location: Royal Borough of Kensington & Chelsea, UK
Posts: 2,883
|
I can see that the video is 640x480, this means that whoever encoded it took the original 720x480 anamorphic file and brutally resized it before deinterlacing it!
That's... bad. Your best bet is to resize to 720x480 first and then try to deinterlace to see if the fields magically match. If they don't, we will have to take more drastic approaches. +1 As lollo said, please upload a sample and we're gonna try to help you. |
25th March 2024, 06:23 | #8 | Link |
Broadcast Encoder
Join Date: Nov 2013
Location: Royal Borough of Kensington & Chelsea, UK
Posts: 2,883
|
Oh crap, I completely missed the framerate, 25i!
Whoever encoded it the first time severely screwed up then as he downscaled a 720x576 25i to 640x480 while still interlaced... No wonder we're left with a field mess... Hahahaha I was gonna say it when I read "Die unschuld von lande" (aka the innocence of the countryside) but I wasn't sure, thanks for pointing it out. By the way, I've seen far kinkier titles at work when I had to encode those kinds of stuff from dorcel and creaways and I'm not gonna lie I still laugh today when I read some of the titles lmao. Why people still watch those contents on linear is very much a mystery to me, though. Older generations I guess? XD |
25th March 2024, 20:02 | #10 | Link |
Registered User
Join Date: Oct 2001
Location: Germany
Posts: 7,259
|
resizing to 640x352 and upscaling to original resolution should remove some of the artifacts
adding some deflickering and spotless will help to remove some of the residual artifacts, but some will still be there |
25th March 2024, 20:53 | #11 | Link |
Registered User
Join Date: Aug 2017
Location: Italy
Posts: 113
|
The problem is in the source, not in QTGMC nor in your script (which is not optimal anyhow); in addition, the source is NOT interlaced, there is no de-interlacing to apply anyhow!
The solution proposed by Selur (Y downsize, resize) may mitigate some of the jagged edges introduced by the wrong processing in the source. Many threads on videohelp forum have attempt to fix this (common) problem, have a look at them.
__________________
A channel on S-VHS / VHS capture and AviSynth restoration https://www.youtube.com/channel/UCMs...h1MmNAs7I8nu4g |
25th March 2024, 22:20 | #12 | Link | |
Broadcast Encoder
Join Date: Nov 2013
Location: Royal Borough of Kensington & Chelsea, UK
Posts: 2,883
|
Unfortunately the source is severely compromised by being resized while still interlaced.
What you can do, however, is a linear blending deinterlacing and blur vertically the lines to get rid of the interlacing artifacts. This is gonna work: Quote:
This also takes care of the levels given that the original source is overshooting significantly way over 0.7V (i.e 235) and it's reaching almost 255, waaaaaay out of range. Before (first picture) after (second picture): This way you can then encode in Limited TV Range as always. Cheers, Frank |
|
26th March 2024, 11:02 | #13 | Link | |
Registered User
Join Date: Aug 2021
Posts: 6
|
Quote:
|
|
26th March 2024, 12:58 | #14 | Link |
Registered User
Join Date: Oct 2001
Location: Germany
Posts: 7,259
|
side note: using a combing mask (i.e. MTCombMask) with the convolution could save some additional details
using "santiag(strh=3)" also works for removing the combing artifacts. Last edited by Selur; 26th March 2024 at 13:38. |
Thread Tools | Search this Thread |
Display Modes | |
|
|