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. |
21st November 2005, 05:51 | #4 | Link |
Guest
Join Date: Jan 2002
Posts: 21,901
|
For pure interlaced, you might be better off with LeakKernelDeint() or TDeint() rather than FieldDeinterlace(). For field-shifted PAL, any competent field matcher, such as Telecide() or Tfm() will be fine. For 3:2 telecined, either Decomb or Tfm()/Tdecimate() will be good. For field-blended video, slit your wrists.
Last edited by Guest; 21st November 2005 at 05:53. |
21st November 2005, 06:01 | #5 | Link |
AMD fan
Join Date: Jan 2005
Location: Russia, Vladivostok
Posts: 49
|
For PAL i use alparysoft deinterlace, but it's requiring payment.
__________________
Russian GK manual for newbie. |
21st November 2005, 06:56 | #6 | Link | |
interlace this!
Join Date: Jun 2003
Location: i'm in ur transfers, addin noise
Posts: 4,555
|
Quote:
with a stubborn movie that was shot on NTSC DV (lots of high-contrast edges that come out stairsteppy no matter what) i just had a go at this one. it's twice as slow as TDeint: repair( \last.tdeint(1,1,type=2,mtnmode=3), \last.tdeint(1,1,type=3,mtnmode=3),mode=1) the line breaks are just for readibility. it'll work if you paste it into a script this way though. the aim of this is to use edge direction, but repress the stray pixel artefacts by using a more conservative mode. it's quite experimental, but seems to give the best of both worlds. you'll need Tritical's TDeint and Kassandro's Removegrain (search )
__________________
sucking the life out of your videos since 2004 |
|
21st November 2005, 09:19 | #7 | Link | |
Registered User
Join Date: Apr 2002
Location: Germany
Posts: 5,391
|
Quote:
@ Mug Funky: Nice idea! (But what about artefacts actually being about 2-pixel clusters?) edit: Halt, had thought too short. type=2 won't even have those 'holes' , so repair(1) does it. Again: nice
__________________
- We´re at the beginning of the end of mankind´s childhood - My little flickr gallery. (Yes indeed, I do have hobbies other than digital video!) Last edited by Didée; 21st November 2005 at 10:44. |
|
21st November 2005, 10:45 | #8 | Link |
interlace this!
Join Date: Jun 2003
Location: i'm in ur transfers, addin noise
Posts: 4,555
|
actually, i think it needs some tweaking... it worked on the source at work, but not on my sample noisy TV cap at home - it just used the kernel mode the whole time. perhaps i've got type 2 and type 3 mixed up in repair?
__________________
sucking the life out of your videos since 2004 |
21st November 2005, 11:08 | #9 | Link | |
Registered User
Join Date: Jul 2005
Posts: 48
|
Quote:
While we're (loosly) on the subject: Can someone recommend a method for getting rid of blends while still maintaining smoothness during pans/zooms. It might just me doing something wrong though. Only way I see is using 2 types of deinterlace, i.e. some deblend for normal scenes and just straight deinterlacing for pans/zooms. BTW this is regarding pure interlaced field blended PAL footage. |
|
21st November 2005, 12:12 | #10 | Link |
Registered User
Join Date: Apr 2002
Location: Germany
Posts: 5,391
|
If it really is "pure interlaced field blended PAL", you might find something in this thread.
@ Mug Funky: Uh, bad eyes today. Looking closer, you've indeed mixed up the two clips. Should be Repair( [ela] , [kernel] , 1 )
__________________
- We´re at the beginning of the end of mankind´s childhood - My little flickr gallery. (Yes indeed, I do have hobbies other than digital video!) |
28th November 2005, 23:25 | #12 | Link |
plebian
Join Date: Nov 2001
Posts: 775
|
I have tried leakkerneldeint(order=1) and when the avs is viewed in vdubmod, some sort of "ghosting" happens. I was using an anime DVD source and the "ghosting" only happens on the character outlines (the black lines around the drawing). Has anyone encountered this as well? What is the solution?
Thanks |
29th November 2005, 04:41 | #13 | Link | |
Registered User
Join Date: Sep 2004
Location: Near LA, California, USA
Posts: 1,545
|
Quote:
Tritical might find out and start charging us $$.
__________________
Pirate: Now how would you like to die? Would you like to have your head chopped off or be burned at the stake? Curly: Burned at the stake! Moe: Why? Curly: A hot steak is always better than a cold chop. |
|
29th November 2005, 12:11 | #14 | Link | |
ffdshow/AviSynth wrangler
Join Date: Feb 2003
Location: Austria
Posts: 2,441
|
Quote:
np: The Orb - 73.5 ft. nitraM (Orb Sessions Volume 1) |
|
30th November 2005, 00:11 | #17 | Link | |
plebian
Join Date: Nov 2001
Posts: 775
|
Quote:
How do I know if the frame is a blended field? maybe you can point me to a URL? thank you |
|
1st December 2005, 23:41 | #18 | Link |
Registered User
Join Date: Mar 2005
Posts: 65
|
Any thoughts on Tdeint versus LeakKernelDeint/Bob for fully interlaced NTSC DV footage? I just finally threw out DGBob for Leak (nice job Leak!), but I'm not sure if I should try to tackle the big Tdeint...
I'm in kind of a weird situation of having footage involving fast-moving thin string which might be stepped through frame by frame by someone. I've basically been sticking with a threshold of 0 because of that (and not generally caring about the background that much).. I'll have to try Mug's thing tonight, though.. Shawn |
2nd December 2005, 00:35 | #19 | Link | |
Guest
Join Date: Jan 2002
Posts: 21,901
|
Quote:
To check for blended fields, do SeparateFields() and then step through the fields and see if any of them are blends of two pictures. I wouldn't trust an app to tell me the field order. Use the manual method in the Decomb tutorial. Last edited by Guest; 2nd December 2005 at 00:40. |
|
Thread Tools | Search this Thread |
Display Modes | |
|
|