The (CPU = 6) line

[TriGGiT]

when ripping a DVD and sending it to Vdubmod you normally use this script line, thats what the guides say you should do

MPEG2Source("E\Disc2\Dolls.d2v", cpu=6)

But my question is what happens if i skip the "cpu = 6" line, also the guides say that with cpu = 6 you are cleaning the dootage of possible blocks etc, but the dvd footage is already in really good quality so is this line of code really necessary or way to important to skip?

discuss!!

tnx

[Kariudo]

not all dvd footage is good quality...take the R1 release of Full Metal Panic! Fumoffu for exapmle. Epic dot crawl...everywhere (I had before/after pics...but they are hopelessly buried somewhere in this forum)

the cpu=6 part takes care of blocking and ringing, to an extent...if it's there.
IIRC, ringing is analogous to rainbowing.

I don't know if there is any advantage to having mpeg2source() take care of some of the blocking/ringing as opposed to other filters...and I suspect that if there is an advantage that it's due to getting extra info from the d2v file.

cpu=6 does some denoising, but I haven't used a source where the difference was noticeable in a long time (I had to use limitedsharpenfaster(edgemode=-1) to see if there was any difference in cpu=6 vs not specifying that parameter)

everything said, I don't think it's a critical parameter in an avisynth script

[TriGGiT]

tnx for teh information
cheers!!!

[Scintilla]

the cpu=6 part takes care of blocking and ringing, to an extent...if it's there.
IIRC, ringing is analogous to rainbowing.

I'm pretty sure ringing is actually analogous to mosquito noise and that they're both manifestations of the Gibbs phenomenon.

I don't know if there is any advantage to having mpeg2source() take care of some of the blocking/ringing as opposed to other filters...and I suspect that if there is an advantage that it's due to getting extra info from the d2v file.

Well, it's either from the D2V file or from the MPEG-2 stream itself, but yes, there is extra info: the quantizers used for each frame. That's why using the postprocessing in MPEG2Source is far superior to using BlindPP later on.

[Zarxrax]

To expand on that, this basically means that if a frame is encoded with a low quantizer (and thus should not have very much compression artifacts) then the postprocessing done is very light. However if a frame is encoded with a high quantizer (and looks really bad), then the postprocessing done is much stronger. So the strength dynamically adjusts to meet the needs of your footage.