I was encoding a movie that was 100 minutes, and when i was done my file size was about 777 MB. I took a look at the kvcd-version and when ever there are very dark backgrounds you can see grayish blocks that are really annoying?

Does anyone know any way to get these off?

The movie is a dvd-rip, so the source is pretty good

Hi muhali3,

What resolution did you encode that movie

-kwag

There are not a lot DVDs that are pretty much perfect.

__________________
j3llyG0053

I encoded it at. 352x480.

I'm seeing something similar. My source is satellite and the dots aren't in the source files.

My guess is that it's a bug in a plugin but I haven't traced it down yet.
It might be due to the temporal smoothing in the script I'm using. Are you using it too?

Im using temporal cleaner.

i get the black block aswell, but i believe this to be the screen res u choose.

I had traced my greyish blocks - not frequent - down to

an avisynth line:
video=coloryuv(video,opt="coring", Autogain=true)

The autogain is what is buggy, you have use:

video=coloryuv(video,opt="coring", Autogain=false)

which works without producing extra grey blocks (mostly during transitions)...

would that edit in avisynth work ok with divx/xvid files?

I cant seem to locate where that line is for autogain?

Heres my scripts in DVD2SVCD;

[AVISYNTH_MA for 4:3 interlaced sources]
0=Edges=^Edges
1=nf=0
2=SeparateFields()
3=UnDot()
4=BicubicResize(^TargetWidth-(Edges+Edges),(^TargetHeight+^BorderTop+^BorderBot tom-Edges-Edges)/2,0,0.6)
5=Import(!MA_Interlaced.avs)
!MA_Interlaced.avs=C:\DVD2SVCD\\D2K\MA_Interlaced. avs
^Edges=16

[AVISYNTH_BilinearResize]
0=BilinearResize(^TargetWidth,^TargetHeight)
1=AddBorders(0,^BorderTop,0,^BorderBottom)

[AVISYNTH_MA script]
0=nf=0
1=GripCrop(^TargetWidth, ^TargetHeight+^BorderTop+^BorderBottom, overscan=^overscan, source_anamorphic=^source_ana, dest_anamorphic=^dest_ana)
2=GripSize(resizer="BiCubicResize")
3=Undot()
4=Asharp(1, 4)
5=STMedianFilter(^S_FilterThreshHold, ^S_EdgeThreshHold, 0, 0 )
6=MergeChroma(blur(^blur_chroma))
7=MergeLuma(blur(^blur_luma))
8=SwitchThreshold = (Width<=352) ? 4 : (Width<=480) ? 3 : 2
9=ScriptClip("nf = round(YDifferenceToNext())"+chr(13)+ "((nf >= SwitchThreshold) && (nf < ^scdtrigger)) ? unfilter(-(fmin(nf*2,100)),-(fmin(nf*2,100))) : TemporalCleaner(6+nf,11+nf) ")
10=GripBorders()
11=function fmin(int f1, int f2) { return (f1<f2) ? f1 : f2 }
12=function ConvertToYUY2(clip c){ return isYV12(c) ? c : convertToYV12(c) }
^overscan=1
^source_ana=true
^dest_ana=false
^S_FilterThreshHold=8
^S_EdgeThreshHold=32
^blur_chroma=1.58
^blur_luma=0.1
^scdtrigger=30

[AVISYNTH_BicubicResize]
0=BicubicResize(^TargetWidth,^TargetHeight,^b_valu e,^c_value)
1=AddBorders(0,^BorderTop,0,^BorderBottom)
^b_value=0.0
^c_value=0.6

[AVISYNTH_Divx Filter]
0=BlindPP(cpu=4)
1=Blockbuster(method="noise",detail_min=1,detail_m ax=3,variance=0.1,seed=1)
2=Convolution3D(1, 6, 12, 6, 8, 2.8, 0)
3=GripCrop(^TargetWidth, ^TargetHeight+^BorderTop+^BorderBottom, overscan=^overscan, source_anamorphic=^source_ana, dest_anamorphic=^dest_ana)
4=GripSize(resizer="^resizer")
5=Undot()
6=TemporalSoften(2,7,7,3,2)
7=DCTFilter(1,1,1,1,1,1,0.5,0)
8=GripBorders()
9=function ConvertToYUY2(clip c){ return isYV12(c) ? c : convertToYV12(c) }
^overscan=1
^source_ana=false
^dest_ana=false
^resizer=LanczosResize

__________________
TheDJ®

Well getting blocks at dark grey scenes is a result of the frequency cutting effect of the Quatisation-matrix. BUT as many times in here diskussed the first try you should do is to add the Blockbuster(x,x,x,x,x,x,x) line BEFORE adding the borders at the end of your script. So look for Dialhots optimal scripts in the avi forum where you can find a typical setting of a blockbuster().
Its NOT the cure (well not in 99% of cases) to do more noisereduction on the source by using temporalsoften because this does exactly what we want to avoid. Plaining the Picture more in already plain parts. Cause thats what happens: The encoder "sees" no details at these plain parts and just "there" acts the frequency cutting done by the filtering of the matrix very well. And so by performing a Blockbuster we add a little "noise" spray to the image which gives the encoder an "information complexitivity fake" so these scenes will be handled more carefully.

@incredible, Thanks for the explanation!

I've added the second blockbuster line as mentioned in Dialhot's latest script, will give that a test.

Cheers

TheDJ®

__________________
TheDJ®

In some cases it helps if the source already is not in a almost unviewable blocky state! In such a case also use the pre-blockbustering in front of your script (also set in Dialhots optimal scripts) this gives also a spry on already existing blocks so they can be handled better by the following filters.
In other cases try to use a tighter GOP or change to the next smaller resoultion your standalone supports.

usually the source is divx/xvid and I always convert to skvcd (because this is the format my DVD standalone prefers).

I know in some cases that divx/xvid files have blocks, so trying to hide them can be hard.

I'm not very experienced with the GOP structure at the moment, could u please enlighten me on how to tighten it?

Thx

__________________
TheDJ®