Avisynth: Latest Script Discussion - Page 15

girv · #**281** 05-04-2003, 09:49 AM

@bman: I did some tests of resize-before-filter vs. filter-before-resize, encoding an xvid source to 480x480 resolution MPEG1 300-2500kbps CQ ~80.

Peering very closely at the TV I noticed the filter-before-resize video was slightly sharper. A marginal difference but it was there. As a bonus the filesize was slightly smaller for filter-before-resize, about 100Kb over 5 minutes. Again a marginal difference and very possibly dependent on the source material, but there :)

Good suggestion bman, a better picture with more compression!

For reference here are the scripts I was using:

resize-filter (normal):

Code:

LegalClip()

LanczosResize(432,416)
AddBorders(0,32,0,32)
Blockbuster(method="dither",detail_min=1,detail_max=10,variance=0.3,seed=0,cache=4096)
MSoften(3,3,4,3,pre=1,post=1,speed=4,showmask=false)
Crop(0,32,Width(),Height()-64)
FluxSmooth(MergeLuma(last,1))

LegalClip()

filter-resize (bman):

Code:

LegalClip()

AddBorders(0,32,0,32)
Blockbuster(method="dither",detail_min=1,detail_max=10,variance=0.3,seed=0,cache=4096)
MSoften(3,3,4,3,pre=1,post=1,speed=4,showmask=false)
Crop(0,32,Width(),Height()-64)
FluxSmooth(MergeLuma(last,1))
LanczosResize(432,416)

LegalClip()

KingTuk · #**282** 05-04-2003, 10:55 AM

girv,

What decoder are you using to decode the DivX/XviD files?

If you use FFDshow latest alpha http://sourceforge.net/projects/ffdshow/

with all the postprocessing settings ( I use custom and check all the blocks and also Nic's processing)

Then in your script use this for your source...

DirectShowSource("C:\movie.avi", fps=25)

of course change the fps to whatever the movie really is like 23.976, 25 etc...

You should be able to get rid of more blocks and flickering this way...

bman · #**283** 05-04-2003, 12:40 PM

There was no blockiness or flickering there , just Gibbs !
@ girv
I'm glad u find my observations to be correct .

Effect is very noticable when source is LQ DivX or Captured material .
bman

girv · #**284** 05-04-2003, 12:56 PM

@KingTuk:

I tried using ffdshow and DirectShowSource when I started all this but all it did for me was to make the picture a bit softer. I'll look at it again though now that I've got a decent post-post-processing script to see if I can improve the picture further. Thanks for reminding me about it :)

[/b]

girv · #**285** 05-04-2003, 12:58 PM

Quote:

Originally Posted by bman

There was no blockiness or flickering there , just Gibbs !

FWIW, I found FluxSmooth was good at removing a lot of Gibbs. How about you? What do you use?

Jellygoose · #**286** 05-04-2003, 02:21 PM

If you think that FluxSmooth is good at removing Gibbs, give FaeryDust a try

You'll see a big difference. However since FaeryDust is so slow, I usually use Flux too to remove Gibbs!

Wolfi · #**287** 05-04-2003, 03:49 PM

Quote:

Originally Posted by Jellygoose

If you think that FluxSmooth is good at removing Gibbs, give FaeryDust a try

You'll see a big difference. However since FaeryDust is so slow, I usually use Flux too to remove Gibbs!

ehum..ehum... gibbs is a new word for me, what is it

//Wolfi

Bigswaffo · #**288** 05-04-2003, 04:03 PM

@Wolfi
Me too. What is a gibb?

Back on pages 13-14 on this thread, there was a lot of discussion about DivX/XviD encoding scripts. Is there any way that these scripts can be tested and than an optimal script placed in Kwag's optimal script thread?

Kane · #**289** 05-04-2003, 04:08 PM

Quote:

Originally Posted by Bigswaffo

@Wolfi
Me too. What is a gibb?

Back on pages 13-14 on this thread, there was a lot of discussion about DivX/XviD encoding scripts. Is there any way that these scripts can be tested and than an optimal script placed in Kwag's optimal script thread?

if the divx/xvid is clean, i treat it like a clean dvd, if it´s worse, i use the vhs script, works great

Bigswaffo · #**290** 05-04-2003, 04:16 PM

Thanks.

jorel · #**291** 05-04-2003, 06:46 PM

Quote:

Originally Posted by girv

Quote:

Originally Posted by kwag

STMedianFilter(10, 30, 0, 0, 10, 30), I can't see any visual difference with or without the filter

To me the output does look slightly softer with the above filter in effect, but you might not notice the effect on a TV. And interestingly I think it made the output look better but that might just be my noisy source material.

I think this one needs more investigation

[/b]

hey friends,
i am trying to find the best for us.

after hard tests my conclusion comparing samples
with and without STM filer.

with the filter the size is really short and look slighty softer.
without the filter is sharpen,faster to encode and with little more final size.

i see the images only in pc and have little differences.
really different is the size and velocity to encode.

someone make tests too?