Dogway's Filters Packs

[kedautinh12]

Ryzen 7000 series support AVX 512 now. I think you can buy it
Any chance for speed-up RIFE when RIFE updated with more new models??

[tormento]

Originally Posted by Dogway

IMO installing Win11 on a not supported CPU is risky and bothersome

Enough you can get TPM in some way.

I managed to find a TPM v1.2 HW key for my motherboard and I am using Win 11 since day one on a i7-2600k.

[DTL]

RIFE can be also tested as pre-degrain for SMDegrain (or may be for MAnalyse only) - https://forum.doom9.org/showthread.p...63#post1976763 . But the default 'model' may give some artifacts at repeating structures and also the scenechange may need tweaking.

Originally Posted by Dogway

By the way, it seems Windows 10 22H2 stable was released yesterday. I will be looking into it after I update ResizersPack. IMO installing Win11 on a not supported CPU is risky and bothersome, you have to depend on hacks that might or not work, and later pray for an update to don't ruin it all. Besides it's still too green, many BSODs as seen recently for Nvidia cards, taskbar issues (might be fixed with yesterday's update), and generally no drivers for a PC 10 years old like mine. I will install it when I get a new PC in a few years time when AVX-512 is again a thing, prices go down and Win11 is more mature.

I doubt that you would have ANY issues, I have been using W11 on "unsupported" CPU's, etc, for quite a while now, and have had many successful updates along the way.

Like I mentioned before, you can download a genuine .iso from MS, and when using Rufus to create a bootable USB, it's "kill's" ALL the annoying problems that a "normal" W11 install has, easy peasy.

You can also modify/customise the installation setup using NTLite.

And tbh, I have had more problems with my AMD GPU, than nVidia.

You can also customise the nVidia drivers, so they ONLY install the driver, and not all the other crap that can cause problems.

https://www.techpowerup.com/nvcleanstall/

And there are so many other "addon's" that make W11 a pleasure to use, and easy on the eye , and yesterday's update added even more good stuff.

If you wait too long, it will be W12, and you'll have the "mature" problem all over again....no time like the present.

[tormento]

I have tried the following script on a 1080p (badly upsampled) anime:

PHP Code:

SetMemoryMax()
SetFilterMTMode("DEFAULT_MT_MODE", 2)
LoadPlugin("D:\Eseguibili\Media\DGDecNV\DGDecodeNV.dll")
#LoadPlugin("D:\Eseguibili\Media\DGHDRtoSDR\DGHDRtoSDR.dll")
DGSource("F:\In\Saishuu heiki kanojo\01.dgi", fieldop=1)
DeBicubicResizeMT(target_width=960,target_height=540,threads=1,b=0.3,c=0.3)
nnedi3resize(1920,qual=2,sharpness=30)
fmtc_bitdepth (bits=8,dmode=7)
Prefetch(4) 


It jumps:



Tried with fmtc_resample (960, 540, kernel="bicubic", a1=0.3, a2=0.3, invks=True) instead of DeBicubicResizeMT(target_width=960,target_height=540,threads=1,b=0.3,c=0.3)

too but I get the very same error. If I delete the downsize line, it works fine but with no effect (of course).

I can't find any reference about dh parameter.

[tormento]

Another error:

PHP Code:

SetMemoryMax()
SetFilterMTMode("DEFAULT_MT_MODE", 2)
LoadPlugin("D:\Eseguibili\Media\DGDecNV\DGDecodeNV.dll")
#LoadPlugin("D:\Eseguibili\Media\DGHDRtoSDR\DGHDRtoSDR.dll")
DGSource("F:\In\Tatami\01.dgi")
ConvertBits(16)
CCD(15) # Remove big chroma "blocks"

# Remove blocking and smooth out jagged edges (only applies where motion match fails)
mb=Deblock_QED(quant1=30, quant2=40,UV=1)
mb=mb.ex_smooth(1, mode="SG", limit=true, sharp=true) 

# Prefiltering with radius 3 for more coherence. I found preset "normal" to clean edges better.
pre=ex_BM3D(sigma=10,preset="normal",radius=3,UV=1,gpuid=0,tv_range=true) 

# Large radius for more coherence. LFR to recover some smearing. limits=false to don't sharpen motion areas
SMDegrain(6, 400, prefilter=pre, mfilter=mb, ContraSharp=true, RefineMotion=true, plane=4, LFR=300, limits=false, DCTFlicker=false)
# Debanding to reduce posterization in walls/defocused backgrounds
GradFun3plus(thr=0.25, radius=10, mask=2, smode=0,UV=1) 
Prefetch(4) 


P.S: I updated to your last versions before trying these 2 scripts.

[kedautinh12]

Originally Posted by tormento

I have tried the following script on a 1080p (badly upsampled) anime:

PHP Code:

SetMemoryMax()
SetFilterMTMode("DEFAULT_MT_MODE", 2)
LoadPlugin("D:\Eseguibili\Media\DGDecNV\DGDecodeNV.dll")
#LoadPlugin("D:\Eseguibili\Media\DGHDRtoSDR\DGHDRtoSDR.dll")
DGSource("F:\In\Saishuu heiki kanojo\01.dgi", fieldop=1)
DeBicubicResizeMT(target_width=960,target_height=540,threads=1,b=0.3,c=0.3)
nnedi3resize(1920,qual=2,sharpness=30)
fmtc_bitdepth (bits=8,dmode=7)
Prefetch(4) 


It jumps:



Tried with fmtc_resample (960, 540, kernel="bicubic", a1=0.3, a2=0.3, invks=True) instead of DeBicubicResizeMT(target_width=960,target_height=540,threads=1,b=0.3,c=0.3)

too but I get the very same error. If I delete the downsize line, it works fine but with no effect (of course).

I can't find any reference about dh parameter.

Add propSet("_FieldBased", 0) before nnedi3resize

[kedautinh12]

Originally Posted by tormento

Another error:

PHP Code:

SetMemoryMax()
SetFilterMTMode("DEFAULT_MT_MODE", 2)
LoadPlugin("D:\Eseguibili\Media\DGDecNV\DGDecodeNV.dll")
#LoadPlugin("D:\Eseguibili\Media\DGHDRtoSDR\DGHDRtoSDR.dll")
DGSource("F:\In\Tatami\01.dgi")
ConvertBits(16)
CCD(15) # Remove big chroma "blocks"

# Remove blocking and smooth out jagged edges (only applies where motion match fails)
mb=Deblock_QED(quant1=30, quant2=40,UV=1)
mb=mb.ex_smooth(1, mode="SG", limit=true, sharp=true) 

# Prefiltering with radius 3 for more coherence. I found preset "normal" to clean edges better.
pre=ex_BM3D(sigma=10,preset="normal",radius=3,UV=1,gpuid=0,tv_range=true) 

# Large radius for more coherence. LFR to recover some smearing. limits=false to don't sharpen motion areas
SMDegrain(6, 400, prefilter=pre, mfilter=mb, ContraSharp=true, RefineMotion=true, plane=4, LFR=300, limits=false, DCTFlicker=false)
# Debanding to reduce posterization in walls/defocused backgrounds
GradFun3plus(thr=0.25, radius=10, mask=2, smode=0,UV=1) 
Prefetch(4) 


P.S: I updated to your last versions before trying these 2 scripts.

Maybe you put deblock.dll and DGDecode.dll in same folder got error cause DGDecode has Deblock() function too

[anton_foy]

Originally Posted by kedautinh12

Maybe you put deblock.dll and DGDecode.dll in same folder got error cause DGDecode has Deblock() function too

I get the same error and I even removed DGdecode to load from my script and moved the dll to another folder.

[Reel.Deel]

Originally Posted by tormento

"Deblock_deblock" means that the plugin must be named "deblock.dll" otherwise you will get an error.

Docs:

AviSynth's DLLName_function() feature automatically creates prefix aliases; so for example, foo.dll's bar() function may be called as foo_bar().

Originally Posted by kedautinh12

Maybe you put deblock.dll and DGDecode.dll in same folder got error cause DGDecode has Deblock() function too

Official x64 DGDecode does not have the Deblock function, and even if it does have the unofficial x64 DGDecode loaded it would not cause the error he has.

[kedautinh12]

I think need change Deblock_deblock to deblock won't meet same error in future anymore

[tormento]

Let's wait for the final word of Dogway

[Dogway]

You need to download Deblock plugin.

The plugin prefix Deblock_ is to not be confused with other Deblock called functions, it was a change in 2014 by GMJCZP.
It shouldn't trigger an error, it doesn't for me. As Reel.Deel stated the plugin should be called Deblock.dll (case insensitive).

As for the nnedi3resize bug I will try to do a commit later today with the recent changes.

[kedautinh12]

nnedi3resize bug had same bug with deep_resize when use with AnimeITVC or TDecimate().TFM()

[Reel.Deel]

Originally Posted by Dogway

The plugin prefix Deblock_ is to not be confused with other Deblock called functions, it was a change in 2014 by GMJCZP.
It shouldn't trigger an error, it doesn't for me. As Reel.Deel stated the plugin should be called Deblock.dll (case insensitive).

The problem is that the x64 binary from MysteryX is named Deblock-x64.dll. I doubt many are changing of the name of the binaries. I just download them and stick in the auto-load folder.

I also use 299792458m version so the binary is already named deblock.dll.

[kedautinh12]

But if ver from real.finder, i think Deblock_deblock change to deblock won't have any error when use different others name for Deblock.dll
https://github.com/realfinder/AVS-St...k_QED_MT2.avsi

[tormento]

Originally Posted by Dogway

You need to download Deblock plugin.

Where can I find a build? There are sources only there.

[Reel.Deel]

Originally Posted by tormento

Where can I find a build? There are sources only there.

In the Release and x64/Release folders: https://github.com/299792458m/Avisyn...ree/master/Src

[tormento]

Originally Posted by Reel.Deel

/Avisynth-Deblock/tree/master/Src

Looking for a build inside a src directory, unfortunately, goes beyond the capabilities of my OCD mind.

[DTL]

Originally Posted by kedautinh12

Ryzen 7000 series support AVX 512 now. I think you can buy it
Any chance for speed-up RIFE when RIFE updated with more new models??

As I see AVS RIFE is based on rife-ncnn-vulkan and it is based on ncnn https://github.com/Tencent/ncnn that can support both CPU and GPU on windows. So may be if CPU is underloaded or much faster it may help to ncnn. Need to ask RIFE and rife-ncnn-vulkan developers if they can support both CPU and GPU execution of ncnn.