logo Sign In

2006 SE/GOUT hybrid script project - UPDATE: 05-24-11 see 1st & 2nd post... — Page 2

Author
Time

Thanks for the info Darth Mallwalker. 

RU.08, I appreciate anybody that wants to help.  I hope this info from Darth Mallwalker will give you what you need.  I know Pittrek had done a PAL version of the script and he said that he had a problem with the move along sequence in mos eisley.  I've double-checked it on the NTSC side of things and it is correct.

As far as the plugins for avisynth.  I believe  most of them are core filters.  The rest should be available for download on the avisynth wiki page under external filters.  I believe the latest versions should work just fine.

Right now I'm going through the script and trying to figure out which scenes should be weighted average and which should use the chroma copy.  I prefer the weighted average method as long as it works because i don't have to worry about noise or bad alignment.  When this is finished i'm gonna try to replace more of the wipes.  It should be pretty close to finished after this.  I'm always open to ideas to make it even better though.

Luke threw twice…maybe.

Author
Time
 (Edited)

It would be cool for archival purposes if someone could do something so nice with the NTSC & PAL S.E. (I love Lee's work on the PAL version)  ....just as a preservation....a timeline of changes for historical reference if nothing else.

Awesome work being done here, thanks for your efforts.

Author
Time

Happy Moth3r 's day, rofl!

Thanks Mallwalker, I'll check it out. I'm going to have to rip all 6 discs again, haven't got a single iso stored (rofl). It looks like the new gforce script is made to work on ntsc discs? Ahh it doesn't matter I can easily adapt it. All I really need now is that list of the required plugins for this script. Too bad it'll take a whole week to g-force all my discs again, lol.

[ Scanning stuff since 2015 ]

Author
Time

^^Jetrell Fo It'd be easy enough to drop-in replace the gout version that you're using (NTSC-GOUT/PAL-SE), in fact all you'd have to do is set it to resize the gout to the pal resolution and check the frame nubering between the pal/ntsc SE. It wouldn't give you much improvement though since most of the scenes would be from the SE anyway, maybe some very slight detail here and there since its unresized from the LD master tape. Anyway, I have ripped my 6x 2006 SW DVD's, I'll have a play with the g-force script (output non-resized, etc) to simply "clean" the source.

BTW did I mention how bad the gout SW looks!

[ Scanning stuff since 2015 ]

Author
Time

RU.08 said:

All I really need now is that list of the required plugins for this script.

RU.08, If you load the script into AVSP it will tell you which ones are missing.

The ones I see that aren't internal filters are:

Ylevels

Vaguedenoiser

Depan

HDRAGC

Luke threw twice…maybe.

Author
Time

Thanks for the info guys, to Darth Mall and althor1138 especially. I've been real busy lately :( but I finally have some time to try and re-apply the gforce script, which I'll post here as soon as I've completed tinkering with it.

[ Scanning stuff since 2015 ]

Author
Time

Okay, here we are (at last). Now, hopefully these screencaps line up correctly to your NTSC frame numbering (I've cliped 12 frames off the start of the video in an attempt to sync it)...

These frame numbers are from VDubMod, the first frame in VDub is 0 (not 1).

Frame 690, first frame of "Star Wars" logo:
http://img820.imageshack.us/img820/9879/690c.jpg

Frame 81507, first frame of Alderaan in the shot where it is destroyed:
http://img535.imageshack.us/img535/4291/81507.jpg

Frame 168773 disolving to credits...
http://img844.imageshack.us/img844/5161/168773.jpg

And finally, I've completed playing with the g-force script, you'll notice that for now subtitles have been completly removed, and the AA is no longer performed (as it can be done in the next stage when combining the gout+se), the reason I removed it for now is because if I resized it to anamorphic after the AA filter now I will be increasing the "uncompressed" file size by like if anyone notices any glaring errors please point it out to me before I spend 48hrs encoding it!

#######################################################################################
#################### GOUT Filter By G-force V.15.00 USE ONLY ON ANH ###################
#################### ______________________________________________ ###################
#################### Modified for PAL by RU.08 ###################
#######################################################################################

# video source
Mpeg2Source("SW_GOUT.d2v") # <-----Set path

# Trim the first 12 frames to match the NTSC frames
# (don't worry it's easy enough to add back 11 frames of black later if needed)
Trim(11,0)

# cut off all black bars (no longer Mod16!)
Crop(0,124,0,-126,align=true)

# I've removed the levels for now, they can be done later.
# Plus there was really no need to do it first in the script.

#black level, gamma, hue
#Levels(10,1.08,255,0,255).Tweak(hue=-3)

# image stabilization and etc...
# de-ring
pm = last
h4 = pm.RemoveGrain(4,-1)
Median1(pm, MT_Convolution(horizontal="1",vertical=" 0 0 0 1 0 .1 -.1" ,chroma="copy first"),h4,Chroma="copy first")
Median1(last,MT_Convolution(horizontal="1",vertical="0 0 0 0 1 0 0 .1 -.1",chroma="copy first"),h4,Chroma="copy first")
Contra(last,pm)

# global stabalization
DeShaker()

# local motion stabilization/degrain stage
source = last
sigma = 5
fft = source.FFT3dFilter(sigma=sigma,sigma2=sigma*.75,sigma3=sigma*.5,sigma4=sigma*.25)
fft.Clense(grey=true)
filt = Contra(last.Repair(fft,2,-1),last).MergeChroma(source)
idx1 = filt.MSuper(pel=2,sharp=1)
bw_vec1 = MAnalyse(idx1,isb=true, delta=1,overlap=4,plevel=0,dct=1)
fw_vec1 = MAnalyse(idx1,isb=false,delta=1,overlap=4,plevel=0,dct=1)
idx2 = source.MSuper(pel=2,sharp=1,levels=1,chroma=false)
pcorr = source.MDegrain1(idx2,bw_vec1,fw_vec1,thSAD=800,thSCD1=150,thSCD2=75,plane=0)
corr1 = Median1(source,pcorr,filt,chroma="copy first")
idx3 = corr1.MSuper(pel=2,sharp=1,levels=1,chroma=false)
corr1.MDegrain1(idx3,bw_vec1,fw_vec1,thSAD=400,thSCD1=150,thSCD2=75,plane=0)
Median1(corr1,last,filt,chroma="copy first")
Contra(last,source)

########## restore stars
prest = last
prest.MT_Binarize(threshold=20,upper=true,chroma="copy first").MT_Expand(chroma="copy first")
MT_Merge(prest,source,last,chroma="copy first")

########## additional "Lucasfilm" and "Along time ago..." stabilization
prest.Trim(0,265)++prest.Trim(266,687).TemporalSoften(3,255,255,25,2)++last.Trim(688,0)

# I've removed the anti-aliasing part for now (and hence resizing), they can be done later.

########## anti-alias
#NNEDI2(dh=true, field=0)
#NNEDI2(dh=false,field=1)

# Resize (to anamorphic PAL).
#Spline16Resize(720,432)

################# End of script, functions follow... ####################

Function Contra(clip denoised, clip orig)
{# re-write of Didée's contra sharpening routine from TemporalDegrain.avs

rg12d = MT_MakeDiff(denoised,denoised.RemoveGrain(12,-1))
rg12d.Repair(MT_MakeDiff(orig,denoised),9,-1).MT_LUTxy(rg12d,"x 128 - abs y 128 - abs < x y ?")

denoised.MT_AddDiff(last,chroma="copy first")

Return(last)
}

Function Median1(clip input_1, clip input_2, clip input_3, string "chroma")
{# median of 3 clips from Helpers.avs by G-force

chroma = Default(chroma,"process") #default is "process". Alternates: "copy first" or "copy second"

Interleave(input_1,input_2,input_3)
chroma == "process" ? Clense(reduceflicker=false) : Clense(reduceflicker=false,grey=true)
SelectEvery(3,1)

chroma == "copy first" ? last.MergeChroma(input_1) : chroma == "copy second" ? last.MergeChroma(input_2) : last

Return(last)
}

Function DeShaker(clip orig)
{# Image stabalization stage, rewritten as a function by RU.08
temp = Repair(orig.TemporalSoften(3,255,255,25,2),orig)
source = Interleave(temp,orig)
smotion = DePanEstimate(source,trust=0,dxmax=1,dymax=0)
Return DePan(source,data=smotion,offset=-1,mirror=15).SelectEvery(2,0)
}

[ Scanning stuff since 2015 ]

Author
Time

I'm going to need to start my own thread soon. :) I was playing with the 4-eyed storm trooper scene, and I've written an avisynth function I may well use again to take care of it. There is a slight side-effect cause the frames didn't line up absolutly perfectly (visible in the far-left), see what you think. I can probably fix it by spliting the frame into parts and stretching the far-left at a different amount. I left the wipes in so you know it's "real" :) as if it's fake (rofl). Oh, and it would have looked better if I'd used a g-forced source rather than the raw GOUT DVD ;)

http://www.sendspace.com/file/xet9ec 5.68MB

[ Scanning stuff since 2015 ]

Author
Time
 (Edited)

RU.08 said:

I'm going to need to start my own thread soon.

I know I wish you would, some of us are trying to follow what althor is doing here, if you are doing your own thing, than it shouldn't be here in his thread anyways.