DarkT
9th November 2007, 03:31
I was testing this really really great de-rainbower made by MF(mfRainbow v0.31) - and it was pretty much the best, better then bifrost/DFMDerainbow.
Sure, it upped the encoding time from ~5 hours to ~10 hours, but that's no biggy.
HERE however IS somethign which I noticed and IS a biggy:
No MFD:
http://aycu29.webshots.com/image/34388/2002572036593240577_rs.jpg
MFD:
http://aycu29.webshots.com/image/34388/2002513987284361581_rs.jpg
Have you guys encountered it? Anybody knows how to counter it?
The whole script is:
tdeint(mode=1,slow=2,mtnmode=1,ap=55)
Mrestore(numr=2002,denm=5005)
mfRainbow()
If I apply it BEFORE the other two, things get more messed up...
Thing is, this doesn't ocur on every frame...
Anyhow - let me know if you guys have better DeRainbowers, or know how to handle this...
Edit:
After running some mre tests, it appears like MFR only creates a problem on frames which contain blend-ghosts and frames which aren't that well deinterlaced...
Edit2:
It seems that MFD works TWICE faster when preceeded with BiFrost! Weird... By the way, Should MF's Derainbower be used pre or post IVTC? I know you're supposed to remove dot-crawl/rainbows pre IVTC - but I dunno... BiFrost has a special interlaced=true flag...
Sure, it upped the encoding time from ~5 hours to ~10 hours, but that's no biggy.
HERE however IS somethign which I noticed and IS a biggy:
No MFD:
http://aycu29.webshots.com/image/34388/2002572036593240577_rs.jpg
MFD:
http://aycu29.webshots.com/image/34388/2002513987284361581_rs.jpg
Have you guys encountered it? Anybody knows how to counter it?
The whole script is:
tdeint(mode=1,slow=2,mtnmode=1,ap=55)
Mrestore(numr=2002,denm=5005)
mfRainbow()
If I apply it BEFORE the other two, things get more messed up...
Thing is, this doesn't ocur on every frame...
Anyhow - let me know if you guys have better DeRainbowers, or know how to handle this...
Edit:
After running some mre tests, it appears like MFR only creates a problem on frames which contain blend-ghosts and frames which aren't that well deinterlaced...
Edit2:
It seems that MFD works TWICE faster when preceeded with BiFrost! Weird... By the way, Should MF's Derainbower be used pre or post IVTC? I know you're supposed to remove dot-crawl/rainbows pre IVTC - but I dunno... BiFrost has a special interlaced=true flag...