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 Development

Reply
 
Thread Tools Search this Thread Display Modes
Old 24th April 2007, 02:38   #1001  |  Link
ChiDragon
Registered User
 
ChiDragon's Avatar
 
Join Date: Sep 2005
Location: Vancouver
Posts: 610
Well I use TDecimate's display mode with TFM's hints to see what both filters are doing, but I guess I understand if keeping them in would interfere with the second decimation.
ChiDragon is offline   Reply With Quote
Old 24th April 2007, 23:03   #1002  |  Link
foxyshadis
ангел смерти
 
foxyshadis's Avatar
 
Join Date: Nov 2004
Location: Lost
Posts: 9,419
Tritical, would you mind updating your page with the smartdecimate compile? Just to have someplace to reference besides a lonely post when telling people where to get the latest.
__________________
There are four boxes to be used in defense of liberty: soap, ballot, jury, and ammo. Please use in that order.
foxyshadis is offline   Reply With Quote
Old 27th April 2007, 17:56   #1003  |  Link
Chainmax
Huh?
 
Chainmax's Avatar
 
Join Date: Sep 2003
Location: Uruguay
Posts: 3,103
I am going to use TDeint+TMMv1 (this one, right?) as Clip2 for an IVTC, would this:

Code:
AssumeTFF()
Deinted=TDeint(order=1,field=1,type=1,emask=TMM(order=1,field=1))
TFM(mode=6,order=1,PP=7,slow=2,mChroma=true,Clip2=Deinted)
TDecimate(mode=1)
be the correct way to do so?
__________________
Read Decomb's readmes and tutorials, the IVTC tutorial and the capture guide in order to learn about combing and how to deal with it.

Last edited by Chainmax; 27th April 2007 at 18:04. Reason: eliminated "type=1" from TMM's arguments
Chainmax is offline   Reply With Quote
Old 28th April 2007, 20:09   #1004  |  Link
Wilbert
Moderator
 
Join Date: Nov 2001
Location: Netherlands
Posts: 6,335
Quote:
Tritical, would you mind updating your page with the smartdecimate compile? Just to have someplace to reference besides a lonely post when telling people where to get the latest.
Tritical? I can't find that version in this thread.
Wilbert is offline   Reply With Quote
Old 28th April 2007, 20:41   #1005  |  Link
Leak
ffdshow/AviSynth wrangler
 
Leak's Avatar
 
Join Date: Feb 2003
Location: Austria
Posts: 2,441
Quote:
Originally Posted by Wilbert View Post
Tritical? I can't find that version in this thread.
Try this post...

np: Gui Boratto - Beautiful Life (Chromophobia)
__________________
now playing: [artist] - [track] ([album])
Leak is offline   Reply With Quote
Old 4th May 2007, 01:10   #1006  |  Link
Zarxrax
Registered User
 
Join Date: Dec 2001
Posts: 1,177
I'm having some trouble fully understanding the d2v parameter that tfm takes.

If I set the d2v parameter, then tfm will check the d2v for illegal field order, and create a new d2v if necessary. Also, tfm will use hints from the d2v to help its field matching process? Will this theoretically give more accurate results, or will be be faster?

Next, is it really necessary to specify the d2v parameter manually? Can't dgdecode pass the data to tfm as a hint or something? And also, is there anyway that tfm could simply fix the d2v file on the fly, rather than making a new one and requiring you to go back and change your script?
Zarxrax is offline   Reply With Quote
Old 4th May 2007, 05:28   #1007  |  Link
tritical
Registered User
 
Join Date: Dec 2003
Location: MO, US
Posts: 999
@Chainmax
The script looks right, but you'll need to use TMMv1 instead of mmask.dll for that syntax to work.

@foxyshadis
I'll do that.

@Zarxrax
Quote:
Next, is it really necessary to specify the d2v parameter manually? Can't dgdecode pass the data to tfm as a hint or something? And also, is there anyway that tfm could simply fix the d2v file on the fly, rather than making a new one and requiring you to go back and change your script?
It would be possible for dgdecode to pass the information to tfm, but dgdecode doesn't currently have this ability. Also, it is slightly easier reading from the d2v and it's faster. TFM could fix the d2v on the fly (meaning it could simply overwrite the existing file instead of creating a new one), but the dgdecode object would still need to be recreated (script closed/reopened) for the changes to effect decoding. My main reason for not overwritting the existing file is I sometimes like to examine the two side by side.

Quote:
If I set the d2v parameter, then tfm will check the d2v for illegal field order, and create a new d2v if necessary. Also, tfm will use hints from the d2v to help its field matching process? Will this theoretically give more accurate results, or will be be faster?
What exactly happens depends on the 'flags' parameter. With the default setting of 4, tfm will check the d2v for illegal transitions and create a new d2v if necessary. It will also pass rff information to tdecimate. The rff information is only used to help aid hybrid detection (detection of 30p parts), so unless you are using hybrid > 0 it won't effect anything (and if the d2v is just all 2's or 0's it won't effect anything either). tfm will also use the trf flags for field matching in film sections (areas of the d2v with the 012301230123 pattern). That means it will determine what match to use based on the d2v flags. It will then check each of these d2v based match decisions to see if the resulting match is combed or not. If it is, then tfm will go back and run its usual match algorithm. Otherwise, it will use the d2v match. This can give a significant speed up if the video has a lot of sections that follow the 01230123 trf pattern, since it means tfm wont have to run its match algorithm. It can also improve things since tfm isn't always correct in its match decisions.

I will say that if you work mostly with r1 anime, the d2v parameter is usually not that helpful since most r1 anime never has 01230123 flagging (the d2v will just have 0's and 2's, 100% NTSC reported by dgindex). In that case, all that setting the d2v parameter accomplishes is checking for illegal transitions.
tritical is offline   Reply With Quote
Old 4th May 2007, 18:13   #1008  |  Link
Zarxrax
Registered User
 
Join Date: Dec 2001
Posts: 1,177
Alright, thanks for the clarification.
Zarxrax is offline   Reply With Quote
Old 5th May 2007, 23:16   #1009  |  Link
ChiDragon
Registered User
 
ChiDragon's Avatar
 
Join Date: Sep 2005
Location: Vancouver
Posts: 610
tritical, can you check this out?

Metrics file ("cap9_TDec.txt")
Code:
BlankClip(length=80061,pixel_type="YV12",fps=30000,fps_denominator=1001)
TDecimate(mode=2,rate=23.483,input="cap9_TDec.txt",batch=true,display=true)
This gives me a "System exception - Access Violation." Other rate values also cause it. 23.482 works fine, as do these rates with mode=7 instead of 2.

I've also got a request... would it be possible to add standalone "+" to the override options to force a decimated frame to be kept? From what I've gathered through use and TDecimate's doc, "+" is currently only used to separate kept and dropped frames for override cycles (which are hard to write out for hundreds of frames at a time) and doesn't actually override anything itself. Currently I usually just edit my TDecimate metrics files and give frames I want to keep huge values, but it's not really practical, especially when I mess up and want to undo something.
ChiDragon is offline   Reply With Quote
Old 6th May 2007, 19:18   #1010  |  Link
chipzoller
Mr. Woof
 
chipzoller's Avatar
 
Join Date: Jan 2002
Location: USA
Posts: 784
Would someone mind taking a look at this sample. I'm trying to figure out how to setup tfm and tdecimate to restore progressive frames from this apparent 2p4i pattern clip but after reading the help files for both and trying samples/experimenting around with decimation cycles am not sure how to restore this. It actually looks like blended fields, but I know this series (Sherlock Holmes with Jeremy Brett) was shot in film, and all the DVDs in the series except the one from which this episode was taken were just soft pull-downs. Or is this not a candidate for tfm+tdecimate but rather some function like restore24 or the likes? Thanks much for the advice.

Last edited by chipzoller; 7th May 2007 at 04:18.
chipzoller is offline   Reply With Quote
Old 7th May 2007, 08:43   #1011  |  Link
Boulder
Pig on the wing
 
Boulder's Avatar
 
Join Date: Mar 2002
Location: Hollola, Finland
Posts: 4,630
The series were originally shot as progressive 25fps I think. Does RePAL do any good?
__________________
And if the band you're in starts playing different tunes
I'll see you on the dark side of the Moon...
Boulder is offline   Reply With Quote
Old 7th May 2007, 17:01   #1012  |  Link
Livesms
Registered User
 
Livesms's Avatar
 
Join Date: Mar 2006
Posts: 184
@All
Can anybody suggest good Bob deinterlace
I tried Tdeint as a bob deinterlace for my TFF PAL video
Code:
topf = separatefields().selecteven().createMM()
botf = separatefields().selectodd().createMM()
tdeint(mode=1,order=1,emask=buildMM(topf,botf,mode=1,order=1,dis=2))
and got this

Yadif(mode=1) give me better result with straight lines but has lots of artifacts
Livesms is offline   Reply With Quote
Old 8th May 2007, 01:55   #1013  |  Link
chipzoller
Mr. Woof
 
chipzoller's Avatar
 
Join Date: Jan 2002
Location: USA
Posts: 784
@Boulder,
Yes I think you may be right. It's just that with the majority of the releases in the "Adventures of..." series they did a proper telecine job, but with this one it seems they went the lazy route and just did a blend-conversion. I'm really not "up" on the methods and processes to get PAL to NTSC so I can't speak from a knowledgeable position...In any case, RePal seemed to do a good job, although I had to use convertfps(25) at the end as it was outputting a 24.75 frame rate otherwise. I really wish there were some tutorials out there that gave examples of various field blendings and avisynth scripts/plugins to try and tackle them.
chipzoller is offline   Reply With Quote
Old 8th May 2007, 04:27   #1014  |  Link
Boulder
Pig on the wing
 
Boulder's Avatar
 
Join Date: Mar 2002
Location: Hollola, Finland
Posts: 4,630
Instead of ConvertFPS, use AssumeFPS.. then again, Restore24 has a RePAL equivalent which outputs 25fps IIRC. You might want to post in the R24 thread or create a new one in the Avisynth usage forum so it will get noticed by the pros
__________________
And if the band you're in starts playing different tunes
I'll see you on the dark side of the Moon...
Boulder is offline   Reply With Quote
Old 8th May 2007, 06:13   #1015  |  Link
ChiDragon
Registered User
 
ChiDragon's Avatar
 
Join Date: Sep 2005
Location: Vancouver
Posts: 610
You should probably use AssumeFPS(25,1,true) followed by an audio resample instead of ConvertFPS. Bump up the video framerate and alter the audio pitch slightly instead of adding blends to your unblended video...

EDIT: Sorry, I see Boulder already said that. I had a cached version open.
ChiDragon is offline   Reply With Quote
Old 8th May 2007, 14:52   #1016  |  Link
chipzoller
Mr. Woof
 
chipzoller's Avatar
 
Join Date: Jan 2002
Location: USA
Posts: 784
Thanks ChiDragon. I'll try AssumeFPS with those params as AssumeFPS(25) produces out of sync audio. However since I'm dealing with a demuxed DVD source, how would, once you demux the audio track, you get AssumeFPS and avisynth to consider and alter the AC3 file?
chipzoller is offline   Reply With Quote
Old 8th May 2007, 16:58   #1017  |  Link
manono
Moderator
 
Join Date: Oct 2001
Location: Hawaii
Posts: 7,265
Quote:
Originally Posted by chipzoller View Post
@Boulder,
In any case, RePal seemed to do a good job, although I had to use convertfps(25) at the end as it was outputting a 24.75 frame rate otherwise.
What's the matter with 24.975fps? Is it any more peculiar than 23.976? If for AVI, keep it there. If for DVD, run the MPV through DGPulldown with the Custom box ticked and 24.975->29.97 filled in. That way you can keep the audio unchanged. ConvertFPS just reblends what RePAL unblended, and AssumeFPS, as you discovered, requires stretching the audio.

If you are going to use AssumeFPS and stretch the audio, you may as well make it for 23.976fps if, as you say, it was shot on film.
manono is offline   Reply With Quote
Old 8th May 2007, 19:14   #1018  |  Link
chipzoller
Mr. Woof
 
chipzoller's Avatar
 
Join Date: Jan 2002
Location: USA
Posts: 784
Manono, these are destined for my iPod and it needs a frame rate of PAL, film, or ntsc to comply with spec.
chipzoller is offline   Reply With Quote
Old 8th May 2007, 20:00   #1019  |  Link
manono
Moderator
 
Join Date: Oct 2001
Location: Hawaii
Posts: 7,265
OK, but if I'm not mistaken, I believe that's the first time you've mentioned requiring specific framerates.

Me, I'd make it 23.976fps (that's what you mean by film, isn't it?), and stretch the audio to do away with the PAL speedup and higher pitched audio. If you want to keep the same audio, then you can do a ChangeFPS(25) after RePAL, which will add one duplicate frame every 1000 frames, or 1 every 40 seconds, to produce a slight stutter which may or may not be noticeable.
manono is offline   Reply With Quote
Old 8th May 2007, 23:50   #1020  |  Link
ChiDragon
Registered User
 
ChiDragon's Avatar
 
Join Date: Sep 2005
Location: Vancouver
Posts: 610
Quote:
Originally Posted by chipzoller View Post
However since I'm dealing with a demuxed DVD source, how would, once you demux the audio track, you get AssumeFPS and avisynth to consider and alter the AC3 file?
I use NicAC3Source, part of NicAudio (check http://avisynth.org/warpenterprises/). You could also use AC3Source or DirectShowSource with an AC3 codec/filter installed. The AC3 wll be decompressed to uncompressed PCM of course.

Last edited by ChiDragon; 8th May 2007 at 23:50. Reason: link fix
ChiDragon is offline   Reply With Quote
Reply

Tags
tdeint, 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 23:03.


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