Avisynth+

[LouieChuckyMerry]

Back for more. After receiving enough help here to upgrade from 32 bit AviSynth+ to 64 bit--thanks again for all the help --for my usual Blu-ray sources, I'm now trying to learn how to use 64 bit AviSynth+ for my DVD sources. I've found all the required 64 bit .dll's for my upscale scripts except ColorMatrix (to convert from Rec.601 to Rec.709). After much searching and reading, it seems that I can simply replace the 32 bit line:

ColorMatrix(Mode="Rec.601->Rec.709")

with:

ConvertTo(Proper Symbols Here)

in 64 bit AviSynth+, but I can't figure out just what the proper symbols are. In case it matters, my script is something like:

DGSource Here
### Deinterlace ###
TFM(Mode=7,UBSCO=False)
### Color Conversion ###
ColorMatrix(Mode="Rec.601->Rec.709")
### Adjust Color ###
MergeChroma(aWarpSharp2(Depth=16))
### Crop ###
Crop(8,1,-8,0)
### Gibbs Noise Block ###
Edge=MT_Edge("prewitt",ThY1=20,ThY2=40).RemoveGrain(17)
Mask=MT_Logic(Edge.MT_Expand().MT_Expand().MT_Expand().MT_Expand(),Edge.MT_Inflate().MT_Inpand(),"xor").Blur(1.0)
MT_Merge(Minblur(),Mask,Luma=True)
### Overall Temporal Denoise ###
SMDegrain(TR=3,ThSAD=200,ContraSharp=True,RefineMotion=True,Plane=0,PreFilter=2,Chroma=False,n16=True,n16_Out=True)
### Resize ###
EDI_RPow2(CShift="Spline64",FWidth=960,FHeight=720)
aWarpSharp4xx(Depth=5)
### Darken-Thin Lines ###
FastLineDarkenMod4(Strength=24,Prot=6)
aWarpSharp4xx(Blur=4,Type=1,Depth=8,Chroma=2)
### Deband ###
F3KDB(Y=100,Cb=100,Cr=100,GrainY=0,GrainC=0)

As always, thanks for any help.

Edit: Happy Saturday! After banging my (obviously thick) head against the wall for a while longer I realized that MeGUI has a copy of ColorMatrix.dll in its AviSynth_Plugins folder, so I borrowed a copy from the 64 bit version and all is well. I'd still be curious if "ConvertTo(Something)" would do the same job...

[real.finder]

bug in expr? in float(32)

mt_lut( "x range_half - 0.1 * range_half +", scale_inputs="floatf", U=3,V=3 )

not same as

expr( "x range_half - 0.1 * range_half +" ,scale_inputs="floatf")

without scale_inputs="floatf" they are same

[pinterf]

Originally Posted by real.finder

bug in expr? in float(32)

mt_lut( "x range_half - 0.1 * range_half +", scale_inputs="floatf", U=3,V=3 )

not same as

expr( "x range_half - 0.1 * range_half +" ,scale_inputs="floatf")

without scale_inputs="floatf" they are same

Do you know which one is correct?

[real.finder]

Originally Posted by pinterf

Do you know which one is correct?

the mt_lut one is correct

[pinterf]

Originally Posted by real.finder

the mt_lut one is correct

O.k. thanks, I check them both and found no problem for a quick five-minute investigation, so I need more time.

[pinterf]

Originally Posted by real.finder

the mt_lut one is correct

This issue was fixed, it's on git already, I'll make a build next week.

[real.finder]

Originally Posted by pinterf

This issue was fixed, it's on git already, I'll make a build next week.

thank you

[djonline]

Trim not work after Import, only after DirectShowSource.

DirectShowSource("avisynth-stage1-with-MTS.avs") still not work with LAV splitter 0.71

Originally Posted by LigH

Then write the Trim in a separate line, not after a dot.

If the imported script returns a clip, then it should be passed to the usual internal variable "last", which is then implicitly assumed where no explicit clip variable was written.

If it does not return any clip, then it can't be used as video source anyway.

Still not work with Deshaker, video is shaked when using Import

"00425.MTS-pass2.avs"
vid="00425.MTS"
o=DirectShowSource(vid)
o+o.Trim(0,29)
ConvertToRGB32
LoadVirtualDubPlugin ("c:\Program Files\Vdub\vdub64\plugins64\Deshaker_64.vdf", "deshaker",0) deshaker ("19|2|30|4|1|0|1|0|640|480|1|2|1000|1000|2000|2000|4|0|6|2|8|30|300|3|f:\\00425.MTS.1.1000.1000.2000.2000.log|0|0|0|0|0|0|0|0|0|0|0|0|0|1|70|70|10|30|1|1|30|30|0|0|0|0|1|0|1|10|1000|1|88|1|1|20|30000|100|20|1")

00425.MTS-loader.avs
#DirectShowSource("00425.MTS-pass2.avs").Trim(30,0)
Import("00425.MTS-pass2.avs")
Trim(30,0) #video is shaked with this

[StainlessS]

# "00425.MTS-pass2.avs"
vid="00425.MTS"
o=DirectShowSource(vid)
# o+o.Trim(0,29)  # Join o.trim(0,29) to END OF o
o.Trim(0,29)  # presume this is required
#o.Trim(29,0) # EDIT: Or, is it this
ConvertToRGB32
LoadVirtualDubPlugin ("c:\Program Files\Vdub\vdub64\plugins64\Deshaker_64.vdf", "deshaker",0)
deshaker ("19|2|30|4|1|0|1|0|640|480|1|2|1000|1000|2000|2000|4|0|6|2|8|30|300|3|f:\\00425.MTS.1.1000.1000.2000.2000.log"
 \ + "|0|0|0|0|0|0|0|0|0|0|0|0|0|1|70|70|10|30|1|1|30|30|0|0|0|0|1|0|1|10|1000|1|88|1|1|20|30000|100|20|1")

[Stereodude]

Is there an avisynth+ version of DitherTools by any chance (I searched but came up empty)? If not, is there a way to get dithertool's rgb48y output into RGBP16 which AVIsynth+ handles? I think rgb48y and RGBP16 are basically the same thing just stored differently.

[real.finder]

Originally Posted by Stereodude

Is there an avisynth+ version of DitherTools by any chance (I searched but came up empty)? If not, is there a way to get dithertool's rgb48y output into RGBP16 which AVIsynth+ handles? I think rgb48y and RGBP16 are basically the same thing just stored differently.

avisynth+ version of DitherTools? no, it's still avs25/avs26

some DitherTools functions for avs+? yes, like the one I post here maybe there are others by another persons

[qyot27]

Originally Posted by Stereodude

Is there an avisynth+ version of DitherTools by any chance (I searched but came up empty)? If not, is there a way to get dithertool's rgb48y output into RGBP16 which AVIsynth+ handles? I think rgb48y and RGBP16 are basically the same thing just stored differently.

From the Dither tools wiki page:

"rgb48y" 48-bit RGB. The components R, G and B are conveyed on three YV12 or Y8 (if supported) stack16 clips interleaved on a frame basis.

I don't truly know how to interpret that description. Because from what it sounds like, you'd have to treat a Y8 frame (assuming it would default to Y8 on 2.6 and Plus) as an individual RGB plane, but you *also* have to apply ConvertFromStacked() and Select every third frame into individual variables to try then stitching them back together into a real RGBP16 stream with CombinePlanes. That's just my best guess, assuming the 'interleaved on a frame basis' means something like 6 frames being R1-B1-G1-R2-B2-G2, and not 'interleaved' in the sense that Stack16 format's counterpart Interleaved16 was for doing high bit depth.

IMO, I think it'd be easier to see if other Plus-compatible plugins (or its native HBD and dither functionality) can do what you want without bringing DitherTools into it at all.

[stax76]

I would like to request scaling for Info() because with 4K it's barely readable.

A player such as mpv shows always the same OSD font size regardless of the video resolution.

[qyot27]

Originally Posted by stax76

I would like to request scaling for Info() because with 4K it's barely readable.

A player such as mpv shows always the same OSD font size regardless of the video resolution.

Just set the font size to 72. Or higher. Whatever you want; it's been customizable for three years now (AviSynth Wiki).

[wonkey_monkey]

May I recommend Info2, which is much easier to read and much much faster (it's missing a couple of not-that-useful things from Info).

[VoodooFX]

What AVS+ vs AVS 2.6 does differently at initial script load? I noticed that one of my script loads ~hundred times slower in AVS+.
Most slowness I localized to this pseudo code below, AVS loads it in milliseconds or a second when AVS+ takes ~few minutes. (I don't want to publish script atm).

Trim1 = clip.Trim(22,-30)+clip.Trim(22222,-30)
a = Trim1.Loop(200)

a.PseudoParse()
Trim(0,-1)
setA = PseudoSetting(last)
aClip = clip.PseudoJob(setA) 

Trim2 = aClip.Trim(22,-30)+aClip.Trim(22222,-30)
b = Trim2.Loop(200)

b.PseudoParse()
Trim(0,-1)
setB = PseudoSetting(last)
bClip = aClip.PseudoJob(setB) 

Trim3 = bClip.Trim(22,-30)+bClip.Trim(22222,-30)
c = Trim3.Loop(200)

c.PseudoParse()
Trim(0,-1)
setC = PseudoSetting(last)
cClip = bClip.PseudoJob(setC) 

return cClip

[wonkey_monkey]

Hard to say if we don't know what all the Pseudo functions do. Is there still a speed difference if you take those blocks out (replacing them aClip = a, etc), leaving just all the splices?

[VoodooFX]

It goes slower in some geometric progression with every xClip block, AVS 2.6 loads it in a second, when AVS+ runs script faster than AVS (I think). If it helps: PseudoParse() is AnalyzeLogo(), PseudoSetting() is ImageWriter(), PseudoJob() is DeblendLogo() from AVSInpaint.

Actual script will be more confusing without manual(yet), and I want to solve few other issues before publishing it.

PS:
Replacing to: aClip = Clip, bClip = Clip - just makes script 3 times faster to load. Remove all those "inception" blocks and AVS+ still is significantly slower at loading.

[LigH]

Originally Posted by wonkey_monkey

May I recommend Info2

Pretty pretty!

[magiblot]

What is the range of expected pixel values for floating-point formats? [0..1], [0..255/256], or the former for RGB and the latter for YUV?