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. |
23rd August 2008, 17:42 | #1 | Link |
Registered User
Join Date: Feb 2007
Posts: 45
|
TFM not matching correctly
Hi,
Here is a clip from Avatar: The Last Airbender: http://www.mediafire.com/?zyjsl19ps9w I am doing a two pass intended for VFR encoding avisynth script pass 1: Code:
tfm(movie,mode=0,pp=7,mi=40,output="tfm.txt").tdecimate(mode=4,output="stats.txt") Code:
tfm(movie,mode=0,pp=7).tdecimate(mode=5,hybrid=2,dupThresh=1.0,input="stats.txt",\ tfmin="tfm.txt",mkvout="timecodes.txt") I have used YATTA to manually match the fields, and every one has a perfect match. I am just not insane enough to use YATTA on 61 episodes (yet.) I have played around with many of the options in the TFM manual; micmatch, mode, pp, etc. The only thing that helps is pp set to 2 or 5 - but that is because it uses blend deinterlacing which makes the cartoon kind of 'gooey'. Is there a way to get tfm to match properly? Thanks, Z Last edited by xxxyzzzz; 11th October 2008 at 17:09. Reason: readability |
9th October 2008, 12:23 | #3 | Link |
Sleepy overworked fellow
Join Date: Feb 2008
Location: Maple syrup's homeland
Posts: 933
|
Have you tried animeivtc(mode=5)?
__________________
AnimeIVTC() - v2.00 -http://boinc.berkeley.edu/- Let all geeks use their incredibly powerful comps for the greater good (no, no, it won't slow your filtering/encoding :p) |
9th October 2008, 13:55 | #4 | Link |
Registered User
Join Date: Sep 2003
Posts: 267
|
Sorry, misread the post. Looks like you already tried it.
__________________
"Do everything by hand, even when using the computer." - Hayao Miyazaki Last edited by DigitalDeviant; 9th October 2008 at 14:12. Reason: Misread original post |
11th October 2008, 05:01 | #5 | Link | |
Registered User
Join Date: Feb 2007
Posts: 45
|
Quote:
Z |
|
11th October 2008, 06:23 | #6 | Link |
Sleepy overworked fellow
Join Date: Feb 2008
Location: Maple syrup's homeland
Posts: 933
|
I can't download your clip... could you use mediafire?
__________________
AnimeIVTC() - v2.00 -http://boinc.berkeley.edu/- Let all geeks use their incredibly powerful comps for the greater good (no, no, it won't slow your filtering/encoding :p) |
11th October 2008, 17:02 | #7 | Link |
Registered User
Join Date: Feb 2007
Posts: 45
|
When you click on the link, there is a big brown button that syas 'download'. Click it and 'captcha' request replaces it. Enter the 4 characters it asks and click 'get'.
If it still fails, I can re-uploaded elsewhere. Z Edit: It doesn't work, it only downloads a piece. Here is a new location for it: http://www.mediafire.com/?zyjsl19ps9w Last edited by xxxyzzzz; 11th October 2008 at 17:08. |
11th October 2008, 22:37 | #8 | Link |
Sleepy overworked fellow
Join Date: Feb 2008
Location: Maple syrup's homeland
Posts: 933
|
I don't see any frames that are not matched well... can you either give me the frame # in the stream you uploaded, or upload a different one?
__________________
AnimeIVTC() - v2.00 -http://boinc.berkeley.edu/- Let all geeks use their incredibly powerful comps for the greater good (no, no, it won't slow your filtering/encoding :p) |
12th October 2008, 02:37 | #9 | Link |
Registered User
Join Date: Feb 2007
Posts: 45
|
My apologies,
You are right - using AnimeIVTC(), the frames do match correctly. (I was mistaking the oversharpening of the source during some of the movement as bad fields.) I see what you did - using tdeint(2)'s interlace detecting methods instead of TFM's internal ones allowed TFM to match much better. Pretty smart. That does solve my problems. Thanks. Now if I could just get the de-haloing to my liking for that darn oversharpening.... Z Last edited by xxxyzzzz; 12th October 2008 at 03:28. |
12th October 2008, 04:18 | #10 | Link |
Sleepy overworked fellow
Join Date: Feb 2008
Location: Maple syrup's homeland
Posts: 933
|
I remember seeing a thread not so long ago about Avatar's haloing... ah, here it is
mt_convolution is a very useful tool for halos; see here and below for a practical example + some theory
__________________
AnimeIVTC() - v2.00 -http://boinc.berkeley.edu/- Let all geeks use their incredibly powerful comps for the greater good (no, no, it won't slow your filtering/encoding :p) |
Thread Tools | Search this Thread |
Display Modes | |
|
|