digitalFAQ.com Forums [Archives]

digitalFAQ.com Forums [Archives] (http://www.digitalfaq.com/archives/)
-   Off-topic Lounge (http://www.digitalfaq.com/archives/lounge/)
-   -   Attacking "The Two Towers SEE" (http://www.digitalfaq.com/archives/lounge/6772-attacking-two-towers.html)

Jellygoose 11-18-2003 03:07 PM

Attacking "The Two Towers SEE"
 
Hi everybody...

I just got the Special Extended Edition of LOTR II "The Two Towers"!!
I can't wait to attack these two little discs, trying to squeeze them onto 1 99min. CD-R and see how it will come out... :twisted:
Note the movie is approx. 223 minutes minutes long, and full of action scenes... :roll:
I guess I'll use a resolution of 352x288 and some strong denoise filters to get that movie on 1 CD-R... I know I might not like the quality, but I wanna try it just for the hell of it! :D

Encoder Master 11-18-2003 03:14 PM

I try to put the first part on one 800MB in SVCD size. I hope It's good because there are many dark scenes. I've also done it to put both Matrix parts on one 800 MB in VCD size. There aren't also any block or noise because I've used a CQ of 85 and no Filters. But one disadvantage was the sharpness. :cry:

Jellygoose you will done it. I believe in You. :wink:

Can you say, which settings you want to use and if you have it can you post some screenshots. Then I'm happy. :D

Jellygoose 11-18-2003 04:55 PM

8O 8O

You put "The Fellowship Of The Ring" Extended Edition on one 80min CD-R at 480x576 and CQ 85 ?? 8O
I figure you had a VERY low max. Bitrate then, and the movie must be extremely blurred...

I'll give different settings a go, and sure I'll post some screenshots when I'm satisfied... I'll plan on using a 112kb/sec Audio File and 352x288 resolution, with no sharpening filter applied... I'm also going to test how far I can go with Overscan at that resolution and my TV...

Cheers...

Encoder Master 11-18-2003 05:11 PM

Quote:

You put "The Fellowship Of The Ring" Extended Edition on one 80min CD-R at 480x576 and CQ 85 ??
I figure you had a VERY low max. Bitrate then, and the movie must be extremely blurred...
No you misunderstood. It would be very nice but not realized. I put Matrix one and 2 in 352x288 on one 800 MB CD-R an I want to put First lotr in 480x576 on one 800 MB CD-R. Perhaps I use 900 MB. But it's just a try. :D
Perhaps I set the Max bitrate on 2000 and the minimum on 50 and have to cut the end to code it with a lower CQ. Because the end waste a lot of Bitrate.
Quote:

I'll plan on using a 112kb/sec Audio
I think it's the best choise you can do. With 96 it's very bad but 112 it's quite good.

Quote:

File and 352x288 resolution
But this isn't to bad. It looks like a VHS tape or a 3 VCD Rip but on one CD-R. :lol:

Quote:

with no sharpening filter applied
On this resolution it isn't good to set this. Post you final script please. and interprate it. :D

Choose letterbox with "Letterbox(32, 32, 32, 32) If you have luck you can't see the black borders on TV.

incredible 11-18-2003 06:39 PM

Quote:

Originally Posted by Encoder Master
Quote:

File and 352x288 resolution
But this isn't to bad. It looks like a VHS tape or a 3 VCD Rip but on one CD-R. :lol:

Man ...
he's a moderator in here and so he knows how it will look like refring to sharpness on TV finally when encoding at 352x288!
And on the other hand.. he wants to see how far he can go by pushing da limits and not how to preserve a real good quality ;-)

Quote:


Quote:

with no sharpening filter applied
On this resolution it isn't good to set this. Post you final script please. and interprate it. :D
Exactly the opposite! He's very right not to perform a sharpen on a 352x288! Even if he don't want to push the limits, cause the sharpened edges will be scaled 4 times on the tv (2x width & 2x height)!! That means you will see nice glowing edges by watching it on Tv, not mention the higher risk of mosquito artefacts at the edges after encoding!
Quote:

Choose letterbox with "Letterbox(32, 32, 32, 32) If you have luck you can't see the black borders on TV.
This is a damn lot to choose for letterboxing! Even at 480x576 its too much .. maybe in your case you'll have luck(I don't want to know your TV's age).
Not mention if you set these high values when encoding at 352x288!
Remember my words! These borders also will be scaled 2 times on TV!!
That means finally 64,64,64,64 pixels overscan an your Tv!
Crazy! It can't be really your intention to recommend that???? 8O Think!

Quote:

I put Matrix one and 2 in 352x288 on one 800 MB
Yep! But me and you, we both know which matrix and GOP you choosed by encoding these two movies to one CD-R :!: :wink:

Quote:

I hope you'aren't bad but I would choose to take the EDIT*****-Matrix. It's compress much better than the KVCD Notch. and if it's a test you can try it. And you have at 1800kbps max bitrate an already high CQ. The MDVDex Matrix Compress average 5-10Percent more than KVCD.
Yeeees my friend and exactly that EDIT*****-matrix just applying to a OneCDEncoding still is very very risky AS I ALREADY TOLD YOU!
It still behaves too arbitrary! If it would'nt be that problematic i would had released it earlier, at the same time like the EDIT*****-DVD Templates... THINK!
You can't just recommend this at this state of developing!
I can't belive it! I'm in the developing phase on that ULBR template to use with OneCDEncodings and you are throwing around with things which are definitely not ready to release!
Damn..this sucks!

EDIT:
@ All: Sorry that I'm getting that upset but I already know EncoderMaster from elsewhere.

Encoder Master 11-19-2003 06:07 AM

Quote:

Man ...
he's a moderator in here and so he knows how it will look like refring to sharpness on TV finally when encoding at 352x288!
And on the other hand.. he wants to see how far he can go by pushing da limits and not how to preserve a real good quality
Sorry man but this is a forum and I say what I mean and I don't know whats the problem.

Quote:

Yep! But me and you, we both know which matrix and GOP you choosed by encoding these two movies to one CD-R
With Matrix I mean the Film. :wink: And I haven't use the matrix we know. :wink:

Quote:

Damn..this sucks!
As you sad. Deleate the part with the Matrix and we haven't deleate the problem. :?

incredible 11-20-2003 05:16 AM

@ JellyGoose

As you asked via PN I tried to "attack" this long Movie and it fits very well as 352x288 including 112kbit audio on one CDr80.

- KVCD Template used incl. standard KVCD GOP (1-7-2 therefore GopSeq length = 24) and Notch Matrix.
- Output as 352x288
- min Bitrate 300 (padding on), max Bitrate 1500

The resulted CQ using my Prediction-AVS routine was 67 which gave me a muxed .mpg of 795mb (even using MA)! Which means max Quality is resulted by using these TmpgSettings and those filters in the script at the choosen image Size.
(Edit: I mean "max resultable Quality" refering to max Media space usage, not refering to the choosen AVS filters or Template settings)
(manual Prediction took about 20 mins.)
Therefore the Stream came out in a nice Quality.
(I'll post images this evening)

Here's the script I used including the following filters:
Code:

off=00
gop= 24
gl=2
sa=5
nf =  0
ml =  2
mpeg2source("H:\LOTR_TWO_TOWERS_D1\lotr2.d2v")
BlindPP(CPU=4).ATC().BicubicResize(320,144,0,0.6,5,0,694,416)
ScriptClip("nf = YDifferenceToNext()"+chr(13)+"nf >= 4 ? \
unfilter( -(round(nf)*ml), -(round(nf)*ml) ) : Undot() ")
AddBorders(16,64,16,80).DCTFilter(1,1,1,1,1,1,0.5,0)
FadeOut(200)
Trim(round(framerate()*off), framecount())
#
################### Prediction #####################
#
#pr=sampler(samples=(round(((framecount())/10)/(gop*gl))/sa), Length=(gop*gl))
#Subtitle(text_color=$999999 ,size=12,
#\        "Frames of movie total : "+String(framecount())+" "+
#\        "= ca. "+String(Round(framecount()/Framerate()))+" Sec. "+
#\        "= ca. "+String(ceil((framecount()/Framerate())/60))+" min.",10,13)
#Subtitle(text_color=$999999 ,size=12,
#\        String(round(((framecount())/10)/(gop*gl))/sa)+" Samples * "+
#\        String(gop*gl)+" Frames (GOP="+String(gop)+"*"+String(gl)+")"+
#\        " = ca. "+String(100/10/sa)+"% of Movie total",10,28)
#Subtitle(text_color=$999999 ,size=12,
#\        "Frames of Sampler = "+String(Framecount(pr))+
#\        " = ca. "+String((Framecount(pr)/25))+" sec."+
#\        " = ca. "+String(round((Framecount(pr)/60)/Framerate()))+" min.",10,43)
#Subtitle(text_color=$999999 ,size=12,
#\        "Offset set to : "+String(off)+" sec.  "+
#\        "(Offset intervall should be: 0 sec. / "+
#\        String(Round(((framecount()/Framerate())/60)/(round(((framecount())/10)/(gop*gl))/sa)*60)/2)+
#\        " Sec.)",10,58 )
#sampler(samples=(round(((framecount())/10)/(gop*gl))/sa), Length=(gop*gl))

The only "modding" on the movie I did was cutting some (not all) of the final credits, done in DVD2AVI

Dialhot 11-20-2003 05:59 AM

Did you ever try a lower MIN ?
I use 64 kps and it's fine on my both standalones. Try it. You will raise you CQ by several points !

incredible 11-20-2003 06:22 AM

Yes Dialhot ... I already tried it on other encoding tests a time ago.... but my SAP doesn't support this. But thanks for the hint anyway

This night - for 480x576 output testing - I also did One encoding choosing 480x576, the same filters, but larger different GOP in the Project, and also the lower minBtrate, so bitrates where set to 64 and 1800. But still a CQ of 48 resulted, .... :arrow: Quality failure

But I think .. I'm still going to push the limits a bit more maybe to receive more CQ on that movie at 480x576 ... we'll see cause yesterday it got a bit late in the night and you can assume how do I look now *lol

PS: Maybe cutting a bit more the high frequencies by modding the notch matrix ... shure this will result in a bit unsharper image but let's see all this in order to get the best compromise ... which still will be better looking than 352x288, as it will be 4times up-sized when watching on TV.

Or maybe also a mega-non-standard size at PAL would increase the CQ a lot at higher imagesize, ... like 480x480 @ 25 fps.
Or total conversion to 480x480 @ 23.976 (using an "only-tone-freq-rising-pitch" in Goldwave before resizing the audio length by besweet to avoid deep voices at 23.976 playback)

Jellygoose 11-20-2003 04:08 PM

Ok, for everybody who is interested, here is the script that I'm going to use for this 214 min. Movie...

Code:

nf =  0
ml =  3

Mpeg2Source("E:\DVD1\DVD1full.d2v")

Undot()
BlindPP(cpu=6)
Blockbuster(method="noise",detail_min=1,detail_max=3,variance=0.2,seed=1)
STMedianFilter(3, 3, 1, 1 ) 

BicubicResize(352, 224, 1/3, 1/3, 24, 0, 672, 576)

MergeChroma(blur(1.2))

#
######################################
#
SwitchThreshold = (Width<=352) ? 4 : (Width<=480) ? 3 : 2
ScriptClip("nf = YDifferenceToNext()"+chr(13)+ "nf >= SwitchThreshold ? \
unfilter( -(fmin(round(nf)*ml, 200)), -(fmin(round(nf)*ml, 200)) ) : \
TemporalSoften( fmin( round(2/nf), 6), round(1/nf) , round(3/nf) , 1, 1) ")
#
######################################
#

AddBorders(0, 32, 0, 32)
LetterBox(0, 0, 16, 16) # Depends on situation. Use MovieStacker!

VobSub("E:\DVD1\Neuer Ordner\vts_01_0")

DCTFilter(1,1,1,1,1,1,0.5,0)

## Functions ###
function fmin( int f1, int f2) {
return ( f1<f2 ) ? f1 : f2
}

Use with High Quality Motion Search Precision !! Have fun :wink:

Dialhot 11-20-2003 04:33 PM

Quote:

Originally Posted by Jellygoose
Mpeg2Source("E:\DVD1\DVD1full.d2v")
BlindPP(cpu=6)

If you have a P4, you should use that insteed :
Code:

Mpeg2Source("E:\DVD1\DVD1full.d2v",cpu=6,idct=5)
or even "cpu2='xxxx0x'" insteed of "cpu=2". This way you won't unring the luma (that causes a LOT of blur) but only the chroma.

Quote:

Blockbuster(method="noise",detail_min=1,detail_max =3,variance=0.2,seed=1)
You don't need that here. That is a trick used for breaking DCT blocks in Divx movies. For clean sources as DVD you will just reduce the CQ ! :-(

incredible 11-20-2003 05:13 PM

He got a XP 2xxx , as I know this from his "cooling problem" thread in here. ;-)

Quote:

AddBorders(0, 32, 0, 32)
are you pan-scanning???
(just assuming, not shure)

Mine are
(16,64,16,80)! ok, overscan but more than a double of your top bottom borders! Whats your Source 2.35:1 ???

Dialhot 11-21-2003 03:49 AM

Quote:

Originally Posted by incredible
He got a XP 2xxx

I'm talking about CPU, you answered with the OS version ;-)

Anerboda 11-21-2003 03:58 AM

Quote:

Originally Posted by Dialhot
Quote:

Originally Posted by incredible
He got a XP 2xxx

I'm talking about CPU, you answered with the OS version ;-)

I think he means Athlon XP 2xxx... :D

incredible 11-21-2003 04:51 AM

@ Anerboda : Jackpot! :wink:

Boulder 11-21-2003 05:43 AM

Quote:

Originally Posted by incredible
PS: Maybe cutting a bit more the high frequencies by modding the notch matrix ... shure this will result in a bit unsharper image but let's see all this in order to get the best compromise ... which still will be better looking than 352x288, as it will be 4times up-sized when watching on TV.

I would try this instead of altering the matrix:

DCTFilter(1,1,1,1,1,0.75,0.5,0)

I always place it as the last item in the script, even after the borders. And before you ask, yes, it does produce a lower filesize that way - even if it is minimal but every byte counts in a project like this :wink:

Dialhot 11-21-2003 06:17 AM

Be carreful that you enter in the eye range when you touch to the third (or 6th ;-)) parameter of DCTfilter like this. Do not use too low values.

Boulder 11-21-2003 07:40 AM

I agree that the value may get too low. However, I've never noticed any difference between 1 and 0.75. On the other hand, I've got beads for eyes :bigooh: :lol:

incredible 11-21-2003 07:55 AM

I every time add to all my OneCDencoding-Scripts at the end:

DCTFilter(1,1,1,1,1,1,0.5,0)
(like in Phils scripts and as it is recommended in the readme doc.)

But Ill give your settings a try Boulder ... we'll see! ;-)

Dialhot 11-21-2003 08:50 AM

I already talk about ffdshow that gives the opportunity to test avisynth script "on the fly" to test filters values and effect. Keep that in mind, it's very usefull.

For DCT it's even faster : there is a "DCT" tab in the ffdshow config panel. You can set the values you want and see immediatly how good (or bad) they affect the picture.

incredible 11-21-2003 09:11 AM

NICE! :D

Jellygoose 11-23-2003 04:59 PM

Guys this is fun... Right Now I'm at the state of getting a very good picture quality with CQ_VBR. Still scenes are nice and sharp, I almost killed mosquito noise completely, and DCT Blocks are reduced a lot.
The one problem left for me is blocky picture in action scenes due to bitrate shortage... There's probably not a lot I can do about that since the Average bitrate is at around 510 kb/sec for this movie, but I'm going to try things to better it... Any ideas on that?
TMPGEnc's Soften Block Noise seems to make things worse... :roll:

Dialhot 11-23-2003 05:34 PM

The only way is to lowering your min to save a little more bits for action scenes.

Jellygoose 11-23-2003 06:41 PM

Hi Dialhot!

I know I'm getting on your nerves but... I'm totally in love with PixieDust...
This script is working WONDERS for me on that source at that Ultra-Low-Bitrates... Check it out it's pretty basic!
Code:

Mpeg2Source("E:\DVD1\dvd1.d2v")
Undot()
ConvertToYUY2
PixieDust(limit=5)
Undot()
ConvertToYV12
BicubicResize(352, 224, 0.2, 0.2, 24, 0, 672, 576)
MergeChroma(blur(1.4))
MergeLuma(blur(0.05))
AddBorders(0, 32, 0, 32)
LetterBox(0, 0, 16, 16)
#VobSub("E:\DVD1\Neuer Ordner\vts_01_0")
Levels(0,0.9,255,0,255)
DCTFilter(1,1,1,1,1,1,0.5,0)

It even beats MA by far... Picture looks brilliantly sharp even at that low resolution, PixieDust() takes care of ALL mosquito noise, and the MergeLuma-blur is helping to avoid blocks in action scenes... I know the speed of PixieDust sucks, but to me it's worth every second... :roll:

Boulder 11-24-2003 04:24 AM

Make sure you use ConverttoYUY2() and ConverttoYV12(), otherwise you may get a performance hit. That's one Avisynth strangeness although I don't know if it is fixed by now.

Dialhot 11-24-2003 04:29 AM

Quote:

Originally Posted by Jellygoose
It even beats MA by far... Picture looks brilliantly sharp even at that low resolution, PixieDust() takes care of ALL mosquito noise, and the MergeLuma-blur is helping to avoid blocks in action scenes... I know the speed of PixieDust sucks, but to me it's worth every second... :roll:

Compare it to this scritp please, and tell me what do you think :
Code:

BlindPP(cpu=4)
Deen("a2d",2,8,10)
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)
#Blockbuster(method="noise",detail_min=1,detail_max=10,variance=0.3,seed=5623)
GripBorders()

(uncomment the second blockbuster line only if you have too much DCT blocks in dark areas after encoding).

Note1: I do not know what you call "Noisy" source but we were discussing about PixieDust because you were looking for a good denoiser. The Two Towers SEE... noisy ? ;-)

Note2: All your Chroma is completly screwed up by the 2 "ConvertTo..." but its seems that you don't care about it. Try to compare the color of the source and the ones of the encoding video.
The problem isn't in PixieDust. The problem is that as everyone trying to use these "old" filters, you convert twice. You should perhaps try the internal convert function of mpeg2dec3 (do not remember the name, see the documentation).

incredible 11-24-2003 05:01 AM

In his script as he needs to converttoYUY2() cause of pixiedust.... Is it really needed to convert back to YV12 afterwards???
I only see activated as following ..
Code:

.....
BicubicResize(352, 224, 0.2, 0.2, 24, 0, 672, 576)
MergeChroma(blur(1.4))
MergeLuma(blur(0.05))
AddBorders(0, 32, 0, 32)
LetterBox(0, 0, 16, 16)
#VobSub("E:\DVD1\Neuer Ordner\vts_01_0")
Levels(0,0.9,255,0,255)
DCTFilter(1,1,1,1,1,1,0.5,0)

... does one of these really need YV12 (Im not shure now)??

Dialhot 11-24-2003 06:23 AM

Quote:

Originally Posted by incredible
In his script as he needs to converttoYUY2() cause of pixiedust.... Is it really needed to convert back to YV12 afterwards???

Of course not, and that is what I wanted to point out.

incredible 11-24-2003 06:37 AM

Quote:

Originally Posted by mpeg2dec3 readme doc
MPEG2Source(str "d2v", int "cpu" int "idct" bool "iPP", int "moderate_h", int "moderate_v", bool "showQ", bool "fastMC", str "cpu2")

It doesn't seem that the mpeg2dec3 within the command supports [, string pixel_type]), but I remeber too that I've also read somewhere that by importing a source that could be set. :?:

Only this command is mentioned in the readme....
Quote:

Originally Posted by mpeg2dec3 readme doc
YV12toYUY2(clip, bool "interlaced")

Could it be that we both mean this?:

Quote:

Originally Posted by Avisynth.org
VISource(string "filename" [,...], bool "audio", string "pixel_type")
OpenDMLSource(string "filename" [,...] [, bool "audio"] [, string pixel_type])
AVIFileSource(string "filename" [,...] [, bool "audio"] [, string pixel_type])

There it's supportet as [, string pixel_type])

EDIT: It seems you edited your last Post Dialhot ;-) cause thats why I was typing this

Dialhot 11-24-2003 06:52 AM

Quote:

Originally Posted by incredible
EDIT: It seems you edited your last Post Dialhot ;-) cause thats why I was typing this

You're right ;-). I changed it after having read the documentation myself :-).

However, for those who want to do the convert, think about making a test with "YV12toYUY2" that is the internal convertion function of mpeg2dec3.dll. Perhaps it's better than the avisynth's one.

Jellygoose 11-24-2003 07:57 AM

Thanks for all your inputs mates! Will do more testing on this! :D

@Dialhot: No LOTR II Towers is one of the best authored and cleanest DVDs around. However, when resizing to such a low resolution, I like to get rid of ALL the noise left, because mosquito noise will be terrible at a 352x288 resolution, if the source is not completely denoised before...

incredible 11-24-2003 08:25 AM

Quote:

when resizing to such a low resolution, I like to get rid of ALL the noise left, because mosquito noise will be terrible at a 352x288 resolution, if the source is not completely denoised before...
I know the quality of LOTR II. Well, the difference in my case is that I got the 170mins version but I think both do base on the same master-source.

The best & most effective way to avoid mosquitonoise IMHO is when adding a minimal blur to the video (like mergeLuma/Chroma) but this you're already doing as I saw your last script.

But caution! If you perform to much "denoising" on an already clean source anyway the details will suffer ... and thats what you understand by "more denoising = more compression" this in real means: less details = more compression.
And on the other hand the risk of blocks on big fading clean surfaces gets very high! That's why we often use Blockbuster which adds "spray" as you know to exactly avoid the result of blocks on static surfaces.

Jellygoose 11-24-2003 08:40 AM

Ok, here's one thing I already found out... The blockyness in High Action scenes is NOT due to bitrate shortage. It must be how Automatic VBR (CQ_VBR) allocates the bitrate, because with CQ, this effect does not show up. However Picture seems to be blurrier using CQ, so it's a tough choice for me... :roll:

Boulder 11-24-2003 09:30 AM

Quote:

Originally Posted by Jellygoose
Ok, here's one thing I already found out... The blockyness in High Action scenes is NOT due to bitrate shortage. It must be how Automatic VBR (CQ_VBR) allocates the bitrate, because with CQ, this effect does not show up. However Picture seems to be blurrier using CQ, so it's a tough choice for me... :roll:

That's one problem with TMPGEnc, not being able to do any manual bitrate allocation :cry:

Regarding YV12toYUY2, see this: http://forum.doom9.org/showthread.ph...ght=yv12toyuy2

Jellygoose 11-24-2003 09:50 AM

Thanks For That Link ! :wink:

Jellygoose 11-26-2003 10:17 AM

Ok, with this Project I'll wait until YMPEG gets stable and try it again... :roll: :?

Dialhot 11-26-2003 10:35 AM

Quote:

Originally Posted by Jellygoose
Ok, with this Project I'll wait until YMPEG gets stable and try it again... :roll: :?

??? And what do you expect in this codec that would give you a correct result for Two Towers ?
(I didn't follow all the threads on that tool).

Jellygoose 11-26-2003 11:24 AM

I expect it to allocate the bitrate better than TMPGEnc does. From what kwag said, it has a far better motion search algorithm...
The problem is within TMPGEnc. If CQ_VBR would not produce so many blocks in High Action Scenes (even if they are blurred heavily! ) the movie will look fine...


All times are GMT -5. The time now is 07:21 AM  —  vBulletin © Jelsoft Enterprises Ltd

Site design, images and content © 2002-2024 The Digital FAQ, www.digitalFAQ.com
Forum Software by vBulletin · Copyright © 2024 Jelsoft Enterprises Ltd.