Optimal Script v3: Here it is... - Page 3

Dialhot · #41 10-04-2003, 03:14 PM

Quote:

Originally Posted by incredible

Where's Phil? Still in Holydays?
Well I think sometime you really need Holydays even from CD Encodings

I just returned back from Quebec this morning. I don't have the time to read all the 469 messages that were posted here during my vacation

.

But I can explain the mysterious usage of blockbuster done in the optimal script : that was an idea of girv (I think) to use it to "spray" noise on the already existing blocks that are in the MPEG4 sources in order to have the denoiser that just follow to remove most of them.

That is not the same usage of blockbuster we used to have with clean sources (there it is used to avoid creation of new blocks, not to remove existing ones. See the difference ?).

I think that the script will be changed quite soon : before my departure I did a lot of test using vaguedenoiser and it seems that other peoples have had the same idea here. The results were very promising. And as 4 new realases of vaguedenoiser were done in the last three weeks, I have to return to my test labs

)

See ya...

TCB · #42 10-04-2003, 04:59 PM

Welcome "home"

Since your vacation I've enjoyed your v3 script with great pleasure. Last year I've bpught a capturecard and converted lots of my videotapes to divx. Burning to DVD was too expensive, while working with (S)VCD wasn't giving me good results. With KVCD and your optimized script I finally can put them on CD and watch them on TV again...

Thanks!

Zyphon · #43 10-10-2003, 06:40 AM

I Love this new V3 Script Dialhot it is so sweet and has produced some great results on my DivX conversions thanx 4 that.

Just some newb questions:

I notice in all the scripts you have the line:

Code:

AviSource("movie.avi",false)

1. Does this mean its better to save to wav in VDub and then use Headache to convert it to MP2 or can i ommit the ,false if my DivX has audio?

2. Does it affect the predictions if you dont use external audio encoding?

Thanx.

Edlund · #44 10-10-2003, 09:53 AM

Quote:

Originally Posted by Zyphon

I notice in all the scripts you have the line:

Code:

AviSource("movie.avi",false)

1. Does this mean its better to save to wav in VDub and then use Headache to convert it to MP2 or can i ommit the ,false if my DivX has audio?

Yes, it's better to use VDub + headac3he, the sound quality is much better.

Quote:

Originally Posted by Zyphon

2. Does it affect the predictions if you dont use external audio encoding?

No.

Zyphon · #45 10-10-2003, 10:02 AM

Quote:

Originally Posted by Edlund

Yes, it's better to use VDub + headac3he, the sound quality is much better.

Thanx 4 the reply Edlund i appreciate the useful info you have given me here, now i know the best way to encode the audio.

TOK uses BeSweet and ive heard from a lot of sources say that headac3he is better for the audio conversion.

zes · #46 10-10-2003, 03:14 PM

Quote:

TOK uses BeSweet and ive heard from a lot of sources say that headac3he is better for the audio conversion.

that i don't get. how can the quality differ when they both use the same engine 'mpgenc' ? with besweet u can even use toolame which, ppl used to say, gives even better quality?

Edlund · #47 10-10-2003, 05:02 PM

I haven't compared besweet and headac3he. I use headac3he and am satisfied with the results. I have used besweet only when working with ToK and was satisfied with the results too

Zyphon · #48 10-11-2003, 05:58 AM

Quote:

Originally Posted by zes

Quote:

TOK uses BeSweet and ive heard from a lot of sources say that headac3he is better for the audio conversion.

that i don't get. how can the quality differ when they both use the same engine 'mpgenc' ? with besweet u can even use toolame which, ppl used to say, gives even better quality?

Yeah i like Toolame i used to use to that with the Toolame gui whichi got from Doom9's site and i used to use that for my wav2mp2 conversions with great results.

Im not saying that Headac3he is better than BeSweet its just that most ppl seem to reccomend Headac3he for conversions.

Also another question about the V2 & V3 scripts:

Version 2:

Code:

AviSource("PATH\NAME.avi",false) 
BlindPP(cpu=4) 
Blockbuster(method="noise",detail_min=1,detail_max=3,variance=0.1,seed=1) 
ATC(2,3,5,0.5,false) 
GripCrop(HEIGHT, WIDTH, overscan=1, source_anamorphic=false) 
GripSize(resizer="LanczosResize") 
Undot() 
TemporalSoften(2,7,7,3,2) 
DCTFilter(1,1,1,1,1,1,0.5,0) 
GripBorders()

Version 3:

Code:

AviSource("movie.avi",false) 
BlindPP(cpu=4) 
Blockbuster(method="noise",detail_min=1,detail_max=3,variance=0.1,seed=1) 
ATC(2,3,5,0.5,false) 
GripCrop(HEIGHT, WIDTH, overscan=1, source_anamorphic=false) 
GripSize(resizer="LanczosResize") 
Undot() 
TemporalSoften(2,7,7,3,2) 
DCTFilter(1,1,1,1,1,1,0.5,0) 
GripBorders()

Is it me or are these two scripts identical?

Quote:

Originally Posted by Dialhot

Please note that I've just modified some values in the script. I you took it before this present post, check the new script.
Thanks !

(the new values result in less mosquitoes and a smaller file. In fact I deed a mistake in the previous post on the variance of the noise - 1 insteed of 0.1 !)

Does this mean that that i should change the variance to -1 in the V3 Script or leave it as 0.1?

Thanx 4 any help given.

bigggt · #49 10-11-2003, 08:29 AM

Hi Zyphon where did you get those versions becuase in the optimal versions script it is Vesion 1 and version 2 and they are different
Version 1

Code:

AviSource("PATH\NAME.avi",false) 
Blockbuster(method="noise",detail_min=1,detail_max=8,variance=0.3,seed=5823) 
Deen() 
GripCrop(WIDTH, HEIGHT, overscan=1, source_anamorphic=false) 
GripSize(resizer="LanczosResize") 
Undot() 
TemporalSoften(2,7,7,3,2) 
DCTFilter(1,1,1,1,1,1,0.5,0) 
Blockbuster(method="noise",detail_min=1,detail_max=10,variance=0.5,seed=5623) 
GripBorders()

Code:

Version 2 
AviSource("PATH\NAME.avi",false) 
BlindPP(cpu=4) 
Blockbuster(method="noise",detail_min=1,detail_max=3,variance=0.1,seed=1) 
ATC(2,3,5,0.5,false) 
GripCrop(HEIGHT, WIDTH, overscan=1, source_anamorphic=false) 
GripSize(resizer="LanczosResize") 
Undot() 
TemporalSoften(2,7,7,3,2) 
DCTFilter(1,1,1,1,1,1,0.5,0) 
GripBorders()

Zyphon · #50 10-11-2003, 08:43 AM

Hi bigggt yeah i know that V1 & V2 are different but in this thread on Page 1 at the beginning Dialhot has posted his V3 script.

Now to me the V3 scripts looks identical to the V2 script and im confused by this and was hoping some one could clear this up for me i know im a newb.

Thanx.

bigggt · #51 10-11-2003, 09:01 AM

OK i get you,i remember a while ago Dialhot had two different scripts in the thread i was talking about but after he made this thread(version3) he switched the one in the other thread to get peopls's opinions in the pole..

So in short the optimal script used to be different until this thread was made

Hope that makes sense

Zyphon · #52 10-11-2003, 09:51 AM

Quote:

Originally Posted by bigggt

OK i get you,i remember a while ago Dialhot had two different scripts in the thread i was talking about but after he made this thread(version3) he switched the one in the other thread to get peopls's opinions in the pole..

So in short the optimal script used to be different until this thread was made

Hope that makes sense

Thanx 4 clearing that up for me bigggt i was racking my brain last night studying the two scripts trying to see what the difference was lol.

Yes this make perfect sense now thank you very much.

KunGFu-L33t · #53 10-11-2003, 08:09 PM

Hey

Thanks for the script, i finnally got it working

(as i am a total n00b)

Ok now here are a few problems im getting.

1/ when i played the avs in Windows Media player it came out Upside down wtf is with this.

2/ and also i was reading this thread and it said that if you had a high quality movie you shouldn't use many filters like ATC etc etc, As my video i have is a DVD Quality XViD should i get rid of a few filters and what ones if i need to...

3/ Theres this green line on the right hand side of the video when i play it.

Here is my code..

Code:

AviSource("E:\fahisligad\video1.avi",false) 
BlindPP(cpu=4) 
Blockbuster(method="noise",detail_min=1,detail_max=3,variance=0.1,seed=1) 
ATC(2,3,5,0.5,false) 
LanczosResize(496, 288, 0, 0, 640, 256)
AddBorders(16, 144, 16, 144)
Undot() 
TemporalSoften(2,7,7,3,2) 
DCTFilter(1,1,1,1,1,1,0.5,0)

Dialhot · #54 10-12-2003, 05:38 AM

2/ You can drop at first Undot, then reduce values of temporalsoften to (2,5,5,3,2) or even (1,5,5,3,2)

1/ and 3/ both problems are related to the codec used on your PC for playing Mpeg. Which is it ?

incredible · #55 10-12-2003, 07:18 AM

@ KungFu

The green Border (no matter where it appears) results from a wrong size calculation in Moviestacker. Be shure to set everythng right.
Or use GripFit as included in the optimal script. GripFit includes an automatic detection and calculation routine to avoid problems like in your case.

Try this on very clean and good Dvix/Xvid Sources (like 2x700 MB Dvix/Xvid Movies).

#############################
AviSource("PATH\NAME.avi",false)
BlindPP(cpu=4)
GripCrop(528, 576, overscan=1, source_anamorphic=false)
GripSize(resizer="LanczosResize")
TemporalSoften(1,5,5,3,2)
Asharp(1,2) # 1= sharpen factor, 2= threshold for only sharpen edges and no cleaned surfaces
#DCTFilter(1,1,1,1,1,1,0.5,0)
#Blockbuster(method="noise",detail_min=1,detail_ma x=10,variance=0.5,seed=5623)
GripBorders()
#############################

You don't need to use the Blockbuster at the beginning if there do not appear any artefcts or blocks in your Dvix source. TemporalSoften just cleans a little the surfaces.
And only if needed:
DCTFilter performs the DCT routine in combination with the Blockbuster at the end to avoid new Blocks in your encoding.
First try to make a test encoding using sampler() at the end of your script WITHOUT DCT & Blockbuster. When there will appear blocks etc. remove the "#"s in front of DCT & Blockbuster an do another test with sampler() included at the end of the script.
Then you decide

Only use Asharp if the Value of the predictionned CQ allows this. If the result of the CQ will be to low .. remove the Asharp line.

Dialhot · #56 10-12-2003, 07:57 AM

1/ an asharp after Lanczos ? That is the first time I see someone no telling me that Lanczos is too sharpy

2/ DCTfilter has nothing (or very little) to do with DCT blocks and you should use it everytime as it lower the file size wihtout any visual effect.

incredible · #57 10-12-2003, 08:07 AM

Your totally right Dialhot, .. I meant the Combination Between them (DCT and Blockb.) = less filesize = more CQ = less blocks
I see the things in the result

And I testet it on many movies.

And still to use Lanczos in the script I posted above is very funny, I agree.
Thats because why I write too fast and copied the lines of the script before.

In this case KungFu, change the line above

GripSize(resizer="LanczosResize")

to

GripSize(resizer="BicubicResize")

Cause Lanczos seems to me that it sharpens to much the wohle picture and I think it doesn't work adaptive, so Temp.Soften with this should work harder on the sharpened Surfaces by Lanczos when used before.
Thats why I use bicubicresize and a little a(daptive)sharp after the cleaning.

But still I'm not at the end of my Dvix/Xvid testings ...

Markymoo · #58 10-21-2003, 02:43 AM

If i'm using a CQ of 90 with pretty good quality source which filters in general are best left in and the settings?

AviSource("123.avi",false)
BlindPP(cpu=4)
GripCrop(704, 576, overscan=1, source_anamorphic=false)
GripSize(resizer="LanczosResize")
#TemporalSoften(1,5,5,3,2)
#Asharp(1,2) # 1= sharpen factor, 2= threshold for only sharpen edges and no cleaned surfaces
#DCTFilter(1,1,1,1,1,1,0.5,0)
#Blockbuster(method="noise",detail_min=1,detail_ma x=10,variance=0.5,seed=5623)
GripBorders()

Edlund · #59 10-21-2003, 02:48 AM

You can use dctfilter.