VirtualDub Saving Takes Ages

User avatar
l33tmeatwad
Joined: Wed Feb 16, 2005 3:22 pm
Location: Christiansburg, VA
Contact:
Org Profile

Re: VirtualDub Saving Takes Ages

Post by l33tmeatwad » Sat Dec 28, 2013 8:36 pm

I would use Ut Codec instead of MJPEG, using trim to isolate the parts you want is what a lot of us do to avoid conversion time being super long, and you should try on every source to find out what combination works (as a lot of users find what works for the worst sources, then use that for EVERYTHING, which is a bad idea).
Software & Guides: AMVpack | AMV 101 | AviSynth 101 | VapourSynth 101
PixelBlended Studios: Website | Twitter | YouTube

User avatar
Qyot27
Surreptitious fluffy bunny
Joined: Fri Aug 30, 2002 12:08 pm
Status: Creepin' between the bullfrogs
Location: St. Pete, FL
Contact:
Org Profile

Re: VirtualDub Saving Takes Ages

Post by Qyot27 » Sat Dec 28, 2013 9:15 pm

The basic point is that the MJPEG copy is used as a quick reference (compare the speed at which VDub will seek with Ctrl+Left/Right Arrow in a slow script - or even one with just the frame stuff - and the MJPEG copy), while the HQ scripts are there in order to generate the proper clips you'd be working with.

I guess there's a chance that, since AviSynth treats filters linearly (unless you load into a variable), that applying processing filters on a small frame range that's already been selected out might shave some time off, but the real problem with that is that if there's a lot of clips to be taken from the source file, you'll have to change the frame range for Trim after every clip gets converted (which is a lot of messing around in Notepad or whatever editor is used), rather than only having to do it once and keeping a clear separation between the script that opens the video and the individual clips. Or it concatenates all the clips into a single clip, but that's almost as unwieldy as using entire episodes.
My profile on MyAnimeList | Quasistatic Regret: yeah, yeah, I finally got a blog

Bollocks
Deranged Baboon
Joined: Wed Nov 13, 2013 12:01 pm
Org Profile

Re: VirtualDub Saving Takes Ages

Post by Bollocks » Sun Dec 29, 2013 7:35 am

This is a different issue, but whilst making the changes to the script that you suggested, I noticed that tfm doesn't actually seem to have done anything in terms of IVTC. There are still blended frames, if that's even the right name, at nearly all scenes of movement. I've changed the settings within tfm based on the avisynth guide (The settings match the example you posted previously), but it still doesn't seem to have had much of the desired affect.
"Never fall for a tennis player. Love means nothing to them"

User avatar
Qyot27
Surreptitious fluffy bunny
Joined: Fri Aug 30, 2002 12:08 pm
Status: Creepin' between the bullfrogs
Location: St. Pete, FL
Contact:
Org Profile

Re: VirtualDub Saving Takes Ages

Post by Qyot27 » Sun Dec 29, 2013 5:33 pm

Post a screenshot of one of the trouble frames. That'll probably help in diagnosing it.

Or better yet, slice out a sample of the original .m2v and upload it to MediaFire or someplace so that it can be more easily seen in action.
My profile on MyAnimeList | Quasistatic Regret: yeah, yeah, I finally got a blog

Bollocks
Deranged Baboon
Joined: Wed Nov 13, 2013 12:01 pm
Org Profile

Re: VirtualDub Saving Takes Ages

Post by Bollocks » Sun Dec 29, 2013 6:55 pm

Example 1 (Barely noticeable)
Image

Example 2 (Much more noticeable)
Image

Example 3 (Hideously Noticeable)
Image
"Never fall for a tennis player. Love means nothing to them"

Bollocks
Deranged Baboon
Joined: Wed Nov 13, 2013 12:01 pm
Org Profile

Re: VirtualDub Saving Takes Ages

Post by Bollocks » Sun Dec 29, 2013 6:59 pm

Failure on my part. Due to the size of the image, you can't notice it at all in the first one (It is there, it's just very faint)
"Never fall for a tennis player. Love means nothing to them"

User avatar
Qyot27
Surreptitious fluffy bunny
Joined: Fri Aug 30, 2002 12:08 pm
Status: Creepin' between the bullfrogs
Location: St. Pete, FL
Contact:
Org Profile

Re: VirtualDub Saving Takes Ages

Post by Qyot27 » Sun Dec 29, 2013 9:55 pm

Don't take a full screenshot of VirtualDub. Use the in and out markers to mark a frame range (or use Trim() in the script itself), and then use File->Export->Image sequence. Save the sequence as PNG.
My profile on MyAnimeList | Quasistatic Regret: yeah, yeah, I finally got a blog

User avatar
l33tmeatwad
Joined: Wed Feb 16, 2005 3:22 pm
Location: Christiansburg, VA
Contact:
Org Profile

Re: VirtualDub Saving Takes Ages

Post by l33tmeatwad » Sun Dec 29, 2013 10:57 pm

I would like to point out, make sure you aren't indexing force film, as it could cause stay interlaced frames which would be blended by daa (which you have in your script).
Software & Guides: AMVpack | AMV 101 | AviSynth 101 | VapourSynth 101
PixelBlended Studios: Website | Twitter | YouTube

User avatar
Qyot27
Surreptitious fluffy bunny
Joined: Fri Aug 30, 2002 12:08 pm
Status: Creepin' between the bullfrogs
Location: St. Pete, FL
Contact:
Org Profile

Re: VirtualDub Saving Takes Ages

Post by Qyot27 » Mon Dec 30, 2013 12:25 am

l33tmeatwad wrote:I would like to point out, make sure you aren't indexing force film, as it could cause stay interlaced frames which would be blended by daa (which you have in your script).
Good point, especially that Decimating a Force Filmed d2v file will muck up the framerate.
My profile on MyAnimeList | Quasistatic Regret: yeah, yeah, I finally got a blog

Bollocks
Deranged Baboon
Joined: Wed Nov 13, 2013 12:01 pm
Org Profile

Re: VirtualDub Saving Takes Ages

Post by Bollocks » Mon Dec 30, 2013 7:13 am

I selected the honor pulldown flags option in DGIndex, so I don't think that's the issue.

http://www.mediafire.com/download/36ah7 ... ending.zip

All of the pngs are in the GG FrameBlending zip. Sorry if there's a better way of uploading them (If you'd rather me save them as a video and upload them that way, just say)
"Never fall for a tennis player. Love means nothing to them"

Locked

Return to “Video & Audio Help”