Groucho's Avisynth Stuff

[Groucho2004]

All files listed below can be found here: Download page (http://www.mediafire.com/folder/x6f7yqjufdg7c/Groucho's_Avisynth_Stuff)


[Tools]

AVSMeter (forum thread (https://forum.doom9.org/showthread.php?t=174797))
AVSMeter is a CLI (command line interface) tool which measures the speed/efficiency of Avisynth scripts and checks if Avisynth and its plugin sub-system are set up and working correctly.

Avisynth Info Tool (forum thread (https://forum.doom9.org/showthread.php?t=176079))
This program performs a thorough test of your Avisynth installation and associated plugins, checking if external dependencies are installed, etc.

Universal Avisynth Installer (forum thread) (http://forum.doom9.org/showthread.php?t=172124)
Installer including 9 Avisynth versions and the necessary batch file to switch between them.

avsr 0.2.2
CLI (command line interface) tool which simply executes an Avisynth script. This is useful when, for example, you want to write an image sequence to the disk by using the "ImageWriter" function.

Screen shot:
https://i.postimg.cc/FHZ5d92G/avsr.png





[Plugins & Support DLLs]

Most of the plugins listed were built based on requests for 64 bit versions of existing 32 bit plugins. Many of them are much faster than the originals due to Intel compiler optimisations, some just have slight speed improvements.

ApplyEvery 0.3.1 (x86 & x64)
Stickboy's ApplyEvery (32 & 64 bit) with updated AVS2.6 headers. Documentation included.

AudioGraph 0.0.1 (x86 & x64)
Richard Ling's AudioGraph (32 & 64 bit) with updated AVS+ headers. Documentation (http://avisynth.nl/index.php/AudioGraph).

AvsRecursion 0.2 (x86 & x64)
Kassandro's AvsRecursion (32 & 64 bit) with updated AVS2.6 headers.

ColorMatrix 2.6.0.0 (x86 & x64)
tritical's ColorMatrix with updated AVS2.6 headers and SoraYuki's REC.2020 addition. Documentation included.
Note:
These ColorMatrix builds do not have the 'opt' parameter. The binaries produced with the Intel C compiler using C/C++ code are at least as fast as the ones with the inline asm.

ColorYUV2 1.05 (x64)

DeScratch 1.1.0 (x86 & x64) (Documentation) (http://avisynth.org.ru/descratch/descratch.html)

DeSpot 3.6.3.1 (x86 & x64) (Up-to-date documentation included)

DGMPEGDec 1.5.8 (x86) (Documentation) (http://avisynth.nl/index.php/DGDecode/MPEG2Source)
This DGMPEGDec package contains a highly optimised version of DGDecode.dll.

EEDI2 0.9.2 (x86 & x64) (Documentation) (http://avisynth.nl/index.php/EEDI2)

Ghostbuster 0.2 (x86 & x64)
Sansgrip's Ghostbuster (32 & 64 bit) with updated AVS2.6 headers. Documentation included.

GScript 1.1 (x86 & x64)
Gavino's GScript (32 & 64 bit) with updated AVS2.6 headers. Documentation included.

LoadDLL 1.0 (x86 & x64)
tsp's LoadDLL (32 & 64 bit) with updated AVS2.6 headers. Documentation included.

RemapFrames 0.4.1 (x86 & x64) (Documentation) (http://avisynth.nl/index.php/RemapFrames)

RequestLinear 1.0.5 (x86 & x64)
This plugin contains Kevin Stone's RequestLinear function which is part of TIVTC. For those who just need that function. Documentation included.

SSIM 0.25.1.0 (x86 & x64)
A filter that computes an objective video quality metric between two videos. Documentation included (and also here (http://avisynth.nl/index.php/SSIM)).

SysInfo (x86 & x64)
This filter returns info about OS/CPU/System. Documentation included. (forum thread (https://forum.doom9.org/showthread.php?t=176131))

TBilateral 0.9.11 (x86 & x64) (Documentation) (http://avisynth.nl/index.php/TBilateral)

TMM 1.0.0 (x86) (Documentation) (http://avisynth.nl/index.php/TMM)
There are two builds, one that needs a processor with at least SSE support and one with at least SSE4.x support. I made a separate SSE4.x build because it's quite a bit faster than the SSE build.

TNLMeans 1.0.3 (x86 & x64) (Documentation) (http://avisynth.nl/index.php/TNLMeans)

TUnsharp 0.9.3 (x86 & x64) (Documentation) (http://avisynth.nl/index.php/TUnsharp)

VagueDenoiser 0.35.1 (x64) (Documentation) (http://avisynth.org.ru/vague/vaguedenoiser.html)

[Sparktank]

Nice index!
Subbed for reference.

[StainlessS]

In my opinion, you can open a thread titled Plugins built with ICL or something for simplification, instead for separate thread for every plugins. Like tp's "modernization effort" thread.
Good solution to the bxyhxyh comment, should keep everybody happy.

Oh WOW, just noticed DGDecode 1.5.8 ICL10, I am a happy bunny. :)
:thanks:

[bxyhxyh]

This build of NNEDI3 can use TP7's fturn.dll which is much faster than Avisynth's internal turnleft/turnright functions. Fturn.dll has to be located in the same directory as NNEDI3.dll. Without fturn.dll it works like the "normal" NNEDI3.dll.

Does this nnedi support Y8, Y16 and Y24?

[Groucho2004]

Does this nnedi support Y8, Y16 and Y24?
No. This is built with tritical's 0.9.4 code base. If you need 2.6 color spaces, use jpsdr's updated version.

[Taurus]

So, all in one place.
Really appreciated.:thanks:

[Groucho2004]

Added 64 bit build of FFT3DFilter.

[Groucho2004]

Updated the "Avisynth version switcher"

[asarian]

Hmm, SmartScreen refused to download nnedi3_0.9.4_fturnmod.7z; and now, after disabling it. Windows Defender says it contains malware, and has been removed as well.

EDIT: Allegedly contains 'severe' 'Trojan/Win32/Spursint.A!cl

[StainlessS]

So disable your Smartscreen and Windows Defender, download and submit to VirusTotal.com, get some real idea from a
good number of (probably better) AV software. Then its make your mind up time.

PS, just done it for you, here results:- https://www.virustotal.com/en/file/31e00c0ed92ce6b58baff7c73652ce405d6b7e07ec73405b73f08a96ab3d3077/analysis/1461861203/

0/56 report as malware, including MicroSoft.

Update your AV database, if continues, submit as false +ve.

SHA256: 31e00c0ed92ce6b58baff7c73652ce405d6b7e07ec73405b73f08a96ab3d3077
File name: nnedi3_0.9.4_fturnmod.7z
Detection ratio: 0 / 56
Analysis date: 2016-04-28 16:33:23 UTC ( 1 minute ago )



EDIT: Had not previously been submitted, ie nobody else had problems with it.
Get a decent anti malware, Comodo is good and free.

EDIT: Comodo on Wikipedia:-
https://en.wikipedia.org/wiki/Comodo_Group
https://en.wikipedia.org/wiki/Comodo_Internet_Security

https://www.comodo.com/home/internet-security/free-internet-security.php
Get the free one, Pro version just has 24/7 support [EDIT: Called GeekBuddy, free for 30 days I think].

EDIT: Pro also has some kind of public network protection (wifi), free for 30 days in free version,
I disable it so I dont get too used to, and reliant upon it. Remembered name, TrustConnect.

EDIT: Here also the Comodo forums:- http://forums.comodo.com/

[asarian]

Wasn't really meant as an accusation towards Groucho2004, of course, but I figured maybe that upload site had been compromised or something.

I shall seek a different means to download these modded stuffz anyway. Thx for the virus test.

[StainlessS]

Wasn't really meant as an accusation towards Groucho2004

Sorry, if I gave impression that I thought you you did, I did not.
But once you get it onto your machine, you are likely to have exact same problem again (if same machine).
Really, suggest Comodo as worth a genuine look. Some little time ago there was some kind of report from
FBI/NSA or similar (linked on Comodo forums), that they had problems breaking through Comodo firewall and additional
security, published as (short) list of security suites that cause authorities problems :)

EDIT: Also the Anti-Virus (although not immune to false +ve's) consistently scores very high in independent
AV tests. (way better than eg Norton, which seems to have lost the plot a bit in recent times).

[asarian]

EDIT: Comodo on Wikipedia:-
https://en.wikipedia.org/wiki/Comodo_Group
https://en.wikipedia.org/wiki/Comodo_Internet_Security

https://www.comodo.com/home/internet-security/free-internet-security.php
Get the free one, Pro version just has 24/7 support.


^^ Good links! :) Thx.

And thanks to Groucho2004 too, naturally! I had very good results with a recompiled VapourSynth dll with ICL10 recently, so this will all be very good and welcome!

[Groucho2004]

I had very good results with a recompiled VapourSynth dll with ICL10 recently
Huh? I thought I was the only one making ICL10 builds.

[asarian]

Huh? I thought I was the only one making ICL10 builds.

And right you are. :) It was VS built with ICC16:

vapoursynth.dll built with ICC16 for Windows (http://forum.doom9.org/showthread.php?t=173406)

[Groucho2004]

And right you are. :) It was VS built with ICC16:

vapoursynth.dll built with ICC16 for Windows (http://forum.doom9.org/showthread.php?t=173406)
I wonder why he calls it "ICC". Last time I checked, "icc.exe" was the (free) Intel compiler for Linux.

[asarian]

I wonder why he calls it "ICC". Last time I checked, "icc.exe" was the (free) Intel compiler for Linux.

This is probably the one he used:

https://software.intel.com/en-us/intel-cplusplus-compiler-16.0-user-and-reference-guide

Seems to be able to do both:

Intel® 64 architecture applications can run on the following:

•Windows* operating systems for Intel® 64 architecture-based systems.
•Linux* operating systems for Intel® 64 architecture-based systems.

[Groucho2004]

Added dfttest 1.9.4 built with ICL10. See first post for details.

[asarian]

Added dfttest 1.9.4 built with ICL10. See first post for details.

Isn't it time you upgraded to ICC16 too?! :) Just kidding, really; if I recall correctly, those Intel compilers are wickedly expensive.

Thanks for this one, btw.

[Groucho2004]

Isn't it time you upgraded to ICC16 too?! :) Just kidding, really; if I recall correctly, those Intel compilers are wickedly expensive.
I chose ICL10/11 because they are faster than newer versions in most cases.

Take for example jpdsr's nnedi3 which apparently has numerous (assembly) optimizations and uses Agner Fog's asmlib memcpy functions (which is weird because Intel's internal memcpy is at least as fast):

Script (source is a short PAL SD clip):
MPEG2Source("F:\Test\test.d2v", idct = 3).trim(0,2999)
fturnleft()
NNEDI3(dh = true, nsize = 0, nns = 4, qual = 2)
fturnright()
NNEDI3(dh = true, nsize = 0, nns = 4, qual = 2)

My nnedi3 0.9.4 ICL10:
[Runtime info]
Frames processed: 3000 (0 - 2999)
FPS (min | max | average): 22.74 | 179.7 | 75.49
Memory usage (phys | virt): 529 | 551 MiB
Thread count: 14
CPU usage (average): 78%
Time (elapsed): 00:00:39.739

jpdsr's NNEDI3_v0_9_4_21_x86
(Release_Intel_XP_Core2_SSE4.2):
[Runtime info]
Frames processed: 3000 (0 - 2999)
FPS (min | max | average): 22.64 | 177.0 | 74.76
Memory usage (phys | virt): 529 | 551 MiB
Thread count: 14
CPU usage (average): 78%
Time (elapsed): 00:00:40.130

[asarian]

^^ That was a very illuminating post! :goodpost:

[Groucho2004]

Moved all my stuff to a different page (http://www.iol.ie/~schubert/avs/) without the annoying crap you get from Sendspace/Mediafire and updated all links accordingly. However, I added a secondary link (https://www.dropbox.com/sh/6kb3723po5oqd4b/AADbP8gIJ3YrHVoLU3joqJDma?dl=0) to a Dropbox directory, just in case.

[StainlessS]

I was curious to know what the AvisynthSuppository was in your downloads and so DL'ed it.
Collection of Avisynth versions.

Thanx for that :)

[Sparktank]

Suppository

don't tell me I've been installing it wrong this whole time... :(

[StainlessS]

fraid so.

If at first you don't suck seed, try try try again http://www.cosgan.de/images/smilie/ekelig/a010.gif EDIT: SparkTank, it really is worth that extra effort. :)

[Groucho2004]

Added ColorMatrix 2.5 (ICL11)

[Motenai Yoda]

can't you mod colormatrix to use dithering and/or 16bit in/out?

also IINW your nnedi3 don't sports some of jpdsr's one patches/features.

[Groucho2004]

can't you mod colormatrix to use dithering and/or 16bit in/out?

also IINW your nnedi3 don't sports some of jpdsr's one patches/features.
I only provide optimized builds for existing plugins. I hate fiddling with other people's code apart from small modifications, I'd rather write something from scratch.

If you need 16 bit you can use DitherTools. There's also nothing wrong with jpdsr's NNEDI3. Lot's of choices.

[Sparktank]

Thanks for the updates.

Currently trying to figure out what happened to Windows 10 and batch scripting.
Upgraded from 7 to 10 and %~dp doesn't seem to work anymore. Several batch scripts are having issues lately.

Looking forward to next Repo update.
I saw on SourceForge AVS got some updates to 261. And I see printf AVS+ got some updates, too.

I have got a lot to catch up on for working with Win10.

[Groucho2004]

Upgraded from 7 to 10
Why???

[Groucho2004]

Updated Avisynth Version Switcher:

2016/06/03
- Updated AVS+ DLLs with the latest versions from pinterf (r1858)
- Removed 2.5.7
- Added 2.6.1 Alpha

[Motenai Yoda]

Upgraded from 7 to 10 and %~dp doesn't seem to work anymore. Several batch scripts are having issues lately
w10 here and all my scripts with %~dp work without issues
maybe you don't run those scripts as administrator and they point out to some protected folders?

[Sparktank]

Thanks for the switcher update!

Why???

Laziness on my part. It became more and more difficult to keep track what Windows Updates was doing to try keep it from upgrading.

as administrator

I think that's it. I notice a lot has changed in regards to Admin privileges. Even UAC is not even the same as Win7.

[Groucho2004]

Added TIVTC 1.0.5 built with VC10.

[Groucho2004]

Added TDeint 1.1.0 built with VC10

[StainlessS]

What is the reason for VC10 rather than ICL10/11, faster ?

Thanx in advance.

[Groucho2004]

What is the reason for VC10 rather than ICL10/11, faster ?
Speed is about the same since most of the critical stuff is inline asm anyway. In order to build it with ICL I would have to change some of the code to make it compatible. Not worth it in these cases (TDeint / TIVTC).
So, this is the same code built with a newer compiler.

[GMJCZP]

Thank you for your effort, Groucho2004.
Only is operative for download TDeint the Dropbox link.

[Groucho2004]

Only is operative for download TDeint the Dropbox link.
Ooops, fixed.

[real.finder]

thank you for updates

is VC10 faster than the old original one?

and there are TMM that if used with TDeint will gave more quality

[Groucho2004]

is VC10 faster than the old original one?Yes, but it's a very small improvement. The point is to have a build compiled with a newer (and better) compiler.

there are TMM that if used with TDeint will gave more quality
Thanks for the tip. I'll add it to the list.

[Groucho2004]

Added tritical's TMM 1.0.0 built with ICL10. See first post for details.

[real.finder]

Added tritical's TMM 1.0.0 build with ICL10. See first post for details.

cool, thanks

[real.finder]

does TDeint and fft3dfilter have -vit- changed for mt?

[Groucho2004]

does TDeint and fft3dfilter have -vit- changed for mt?
Don't think so. Do you have links to these modifications?

[real.finder]

Don't think so. Do you have links to these modifications?

http://rgho.st/74sNg9CG2

and there for nnedi3 too, cretindesalpes did include them in his plug-ins

[Groucho2004]

does TDeint and fft3dfilter have -vit- changed for mt?
Thanks for pointing this out. I'll update the source and binaries for the relevant plugins.

[Groucho2004]

Updated FFT3DFilter (32 Bit), TDeint and NNEDI3 with VIT's changes to improve stability when running multi-threaded on SEt's Avisynth MT.

[real.finder]

Updated FFT3DFilter (32 Bit), TDeint and NNEDI3 with VIT's changes to improve stability when running multi-threaded on SEt's Avisynth MT.

thank you for updates

[GMJCZP]

Groucho2004, :thanks:

I suggest you publish the date of each build in the first post, therefore will be more easy search in the list new updates.

I know it's more work but might consider. Thanks again.

[Groucho2004]

I suggest you publish the date of each build in the first post, therefore will be more easy search in the list new updates.
Good idea, will do.

[GMJCZP]

Thank you very much. It's amazing.

So is much better, I just thought something like this format, for example:

( 12/06/2016)
AvsMeter
.
.
.
(20/01/2014)
ColorMatrix
.
.
.
(15/05/2016)
TMM

Therefore all the builds is associated with your last date of updating (and your explanation, is good your idea of include this). Thanks again for your effort.

[kypec]

Therefore all the builds is associated with your last date of updating (and your explanation, is good your idea of include this). Thanks again for your effort.

Please use standard ISO date format, ever! [YYYY-MM-DD] or [YYYY/MM/DD], it also helps a lot to automate sorting of textual data. Locale US/EU date formats are confusing because one can never be sure which is MONTH and which is DAY in those. :thanks:

[Groucho2004]

Thank you very much. It's amazing.
The update list is amazing? You're easy to please! :)



( 12/06/2016)
AvsMeter
.
.
.
(20/01/2014)
ColorMatrix

What Kypec said - "YYYY-MM-DD" is the proper format.

[GMJCZP]

When I refer to the date it is to place next to the item, not the format itself. My mother language is not English, sorry.

PD: By being polite does not give you the right to talk like that. This is a community and we are here to help each other.

EDIT: PD is nonsense, bad interpretation for me. :)

[Sparktank]

Thanks for the updates! (I'll see myself out)

Another suggestion, can you please add your site to your sig? It would make for faster navigation. :thanks:

[GMJCZP]

I don't have own internet.
Can be today connected, but then not to be for months, so it is difficult for me to track updates.

[Groucho2004]

PD: By being polite does not give you the right to talk like that. This is a community and we are here to help each other.
Talk like what? What are you referring to?

[GMJCZP]

LOL , the translator played a trick on me, sorry Groucho.
The word "easy" translated very badly, my apologies. Stupid system for posting that I have right now.:p

[Groucho2004]

LOL , the translator played a trick on me, sorry Groucho.
The word "easy" translated very badly, my apologies. Stupid system for posting that I have right now.:p
What's your native language?

[GMJCZP]

What's your native language?

Spanish. I wrote a couple of words and is a torture, a loaned Ipod for this is not commendable for anything, lol. I know english but sometimes i need a bit of help. :p

[StainlessS]

HaHa, two Spaniards gettin into bother whilst conversing in English :)

[Groucho2004]

HaHa, two Spaniards gettin into bother whilst conversing in English :)
Huh? I'm not Spanish, I just live (and work) here.

[GMJCZP]

HaHa, two Spaniards gettin into bother whilst conversing in English :)

LOL *1000, tin man.:)

[StainlessS]

Oh my god, what a terrible mistake, what is you, Australian ?, worse ? New Zealand ?, worse, cant think of worse.
I know you are not american)

Grouchy, I was totally convinced that you were of the Espaniola persuasion,
I did think that your English was way too better than mine (for some damn foreigner),
but given that my english is so bad, I just thought, so....so.

I did some years back think that Gavino must be code for 'like 'Gavin' (Welsh) or some such,
but thought that the world is a real strange place filled with sooo many weird people
that I dismissed the thought (why would some Welch fella wanna hide in Spain. [ages since last great train robbery]).

Can I ask, (please dont you feel at all compelled to answer), what is is that you and Gavino did that make you have to lie about
your lives and live in secret in Spain ( Nobody here is gonna tell, really).

EDIT: yeh and cheers GMJCZP :)

[Motenai Yoda]

I thought Gavino was from Berserk the manga, but actually that was Gambino XD

@Groucho did you were used to post on DigitalExpression?
but as I suspect your name is Michael I think you don't.

[Groucho2004]

what is you, Australian ?, worse ? New Zealand ?, worse, cant think of worse.
I know you are not american)
German, born and raised. However, I left Krautland 20+ years ago and lived in Ireland, then Spain.


what is is that you and Gavino did that make you have to lie about
your lives and live in secret in Spain
A dark and sinister past full of hideous crimes. Well, not quite. I was lucky to get good jobs in Ireland and later in Spain.

[Groucho2004]

@Groucho did you were used to post on DigitalExpression?
No.
but as I suspect your name is Michael
How did you figure that out?

[Motenai Yoda]

We are spying you since december 2013 as you show liking brexit and podemos movements...
naa you signed some post with it.

[GMJCZP]

Groucho, would it be worthwhile update TUnsharp?

[Groucho2004]

Groucho, would it be worthwhile update TUnsharp?Looks like a potential candidate. I'll check it out.

[Groucho2004]

Added TUnsharp 0.9.3 (ICL)

[Groucho2004]

Oh my god, what a terrible mistake, what is you, Australian ?, worse ? New Zealand ?, worse, cant think of worse.
I know you are not american)

Grouchy, I was totally convinced that you were of the Espaniola persuasion,
I did think that your English was way too better than mine (for some damn foreigner),
but given that my english is so bad, I just thought, so....so.

I did some years back think that Gavino must be code for 'like 'Gavin' (Welsh) or some such,
but thought that the world is a real strange place filled with sooo many weird people
that I dismissed the thought (why would some Welch fella wanna hide in Spain. [ages since last great train robbery]).

Can I ask, (please dont you feel at all compelled to answer), what is is that you and Gavino did that make you have to lie about
your lives and live in secret in Spain ( Nobody here is gonna tell, really).
Methinks you were baked out of your skull when you wrote that. ;)

[johnmeyer]

Methinks you were baked out of your skull when you wrote that. ;)Hey, look at his tagline; what do you expect?? :)

[GMJCZP]

Added TUnsharp 0.9.3 (ICL)

:thanks: one more time, very easy ;-)

Now ss will be more convenient.

[StainlessS]

Methinks you were baked out of your skull when you wrote that.

Absotively poselutely, baked to a crisp :)

German, born and raised
yeh, that explains the absence of humour

Yeh, well it does :)
(Not Spanish, I have to re-evauate my entire life)

GMJCZP
LOL *1000, tin man.

You be quiet :)

[GMJCZP]

I know it's off topic, but let the disquietude: plugins reverse engineering.
That is, obtain the source code of certain plugins. It's posible?

PD: tin man, I have a cool head right now :)

[Groucho2004]

plugins reverse engineering.
That is, obtain the source code of certain plugins. It's posible?
Possible - Yes. Practical - No.
No matter which de-compiler you use, you will spend more time transforming its output to usable C/C++ code than writing the thing from scratch.

[GMJCZP]

I mention this because there are plugins is lacking source code, such as marcfd. It would be good to obtain this for future optimizations and revitalize such projects.

[StainlessS]

As Grouchy said, would be easier to write from scratch.
I kinda miss 'The Peach' (Lindsay Drew I think), no source. (something about blowing Orange smoke up your kilt),
dont know how I would implement the Orange smoke thing and too much work to figure out what it's doing.

[GMJCZP]

I'm not expert on the subject, but I guess I write something from scratch would be ideal.
But then what to do? I share my impotence to do anything about it.

[Groucho2004]

Added TNLMeans 1.0.3 (x64)

[real.finder]

Added TNLMeans 1.0.3 (x64)

This time you beat Joshy D :cool:, your TNLMeans is faster by about 10%

[Magik Mark]

Groucho2004

How do I incorporate this in staxrip? Is it even possible specially the ICC avisynth?

[Groucho2004]

How do I incorporate this in staxrip?Incorporate what exactly?

Is it even possible specially the ICC avisynth?As far as I'm aware, staxrip is 64 bit only, so that won't work.

[Groucho2004]

I updated TNLMeans x64, now built with ICL13. There are two versions, one for CPUs with at least SSE2 and one with at least SSE4.2. The latter has a slight performance advantage. The speed is about 20% better than the previous version which was built with VC10. Benchmarks:

Script:
colorbars(width = 1024, height = 576, pixel_type = "yv12").killaudio().assumefps(24000, 1001)
trim(0, 199)
tnlmeans(ax = 1, ay = 1, az = 1)


JoshyD's 64 bit version:
Frames processed: 200 (0 - 199)
FPS (min | max | average): 4.492 | 8.238 | 5.482
Memory usage (phys | virt): 22 | 18 MiB
Thread count: 9
CPU usage (average): 25%

Time (elapsed): 00:00:36.486

ICL13, SSE2:
Frames processed: 200 (0 - 199)
FPS (min | max | average): 5.320 | 10.38 | 6.940 (+27%)
Memory usage (phys | virt): 22 | 18 MiB
Thread count: 9
CPU usage (average): 25%

Time (elapsed): 00:00:28.818


ICL13, SSE4.2:
Frames processed: 200 (0 - 199)
FPS (min | max | average): 5.370 | 10.58 | 7.070 (+29%)
Memory usage (phys | virt): 22 | 18 MiB
Thread count: 9
CPU usage (average): 25%

Time (elapsed): 00:00:28.288



Note:
If you are using this with AVS+ MT, try to increase the number of threads you pass to "Prefetch()". In my case, the optimum value was 6 (with a 4 core CPU).
However, in combination with other filters, this may not be a good idea. As always, test it with AVSMeter ;).

[Groucho2004]

Updated TNLMeans x86, now built with ICL13.There are two versions, one for CPUs with at least SSE2 and one with at least SSE4.2. The latter has a slight performance advantage. Benchmarks:

Script:
colorbars(width = 1024, height = 576, pixel_type = "yv12").killaudio().assumefps(24000, 1001)
trim(0, 199)
tnlmeans(ax = 1, ay = 1, az = 1)


Tritical's original DLL:
Frames processed: 200 (0 - 199)
FPS (min | max | average): 2.728 | 4.179 | 2.783
Memory usage (phys | virt): 25 | 27 MiB
Thread count: 9
CPU usage (average): 25%

Time (elapsed): 00:01:11.861

ICL13, SSE2:
Frames processed: 200 (0 - 199)
FPS (min | max | average): 4.879 | 8.869 | 5.946 (+214%)
Memory usage (phys | virt): 25 | 27 MiB
Thread count: 9
CPU usage (average): 25%

Time (elapsed): 00:00:33.638

ICL13, SSE4.2:
Frames processed: 200 (0 - 199)
FPS (min | max | average): 5.010 | 9.320 | 6.249 (+225%)
Memory usage (phys | virt): 25 | 27 MiB
Thread count: 9
CPU usage (average): 25%

Time (elapsed): 00:00:32.006

[Magik Mark]

Hi! Is it worth even compiling something that would work in staxrip? Cpu avx2 perhaps?


Sent from my iPhone using Tapatalk

[Groucho2004]

Hi! Is it worth even compiling something that would work in staxrip?
I can't see why this would not work in staxrip.

Cpu avx2 perhaps?
Tried that, not worth it.

[Groucho2004]

Added EEDI2 x64 built with ICL13. Benchmarks:

Script:
colorbars(width = 1920, height = 540, pixel_type = "yv12").killaudio().assumefps(24000, 1001)
trim(0, 499)
eedi2()
JoshyD's EEDI2 x64:
Frames processed: 500 (0 - 499)
FPS (min | max | average): 8.880 | 22.69 | 22.47
Memory usage (phys | virt): 40 | 37 MiB
Thread count: 9
CPU usage (average): 25%

Time (elapsed): 00:00:22.252

ICL13:
Frames processed: 500 (0 - 499)
FPS (min | max | average): 10.74 | 27.65 | 27.39 (+22%)
Memory usage (phys | virt): 40 | 37 MiB
Thread count: 9
CPU usage (average): 25%

Time (elapsed): 00:00:18.257

[tuanden0]

@Groucho2004: Hir Sir, Do you plan to compile Fft3dfilter 2.23 to ICL?

[Groucho2004]

Do you plan to compile Fft3dfilter 2.23 to ICL?
No, pinterf mentioned that he plans to re-write the inline asm code with simd intrinsics.

[tuanden0]

No, pinterf mentioned that he plans to re-write the inline asm code with simd intrinsics.

:thanks: Groucho2004 for the information

[Groucho2004]

Added 64 bit versions of GamMac, RoboCrop and FredAverage (by StainlessS)

[Groucho2004]

Cleanup. Removed plugins that were superseded by versions supporting new colorspaces (TIVTC, fft3dfilter, NNEDI3), 64 bit versions, etc.

[Groucho2004]

Added RemapFrames 0.4.1 (x86 & x64).

[dlnm]

Yes, link is dead and I couldn't find the file anywhere else. So, I have taken cretindesalpes' last update to the code of RemapFrames from here (https://forum.doom9.org/showthread.php?p=1644971#post1644971), updated it to AVS2.6 interface and built 32 and 64 bit versions. I tested them very briefly, let me know how it goes.

Link is in my signature.

Thank you so much.

[amayra]

just out of curiosity
AVSMeter is gone for good ?

[Groucho2004]

just out of curiosity
AVSMeter is gone for good ?Of course not. The link to the dropbox folder is in the first post.

[amayra]

Of course not. The link to the dropbox folder is in the first post.
sorry i mean to ask about development status

[Groucho2004]

sorry i mean to ask about development statusIt's still being developed although there's not much to add or fix.

[Groucho2004]

Updated AVSMeter:

v2.5.6

Fixed a problem with plugin enumeration
Refactoring and tweaks

[Groucho2004]

Updated AVSMeter:

v2.5.7

Added sections for uncategorized files/DLLs in the plugin directories
Fixed a bug in the detection of installed VC++ runtimes
Improved error handling for plugin check

[Groucho2004]

Updated AVSMeter:

v2.5.8
- Re-organised the auto-load sections and added section for AVSI
- Fixed a problem with log file content (regression)
- C 2.0 and CPP 2.0 plugins no longer have a dedicated section in the list of plugins.
They are still enumerated but now fall into the section "Uncategorized".
The reasons for this are
- They are not supported by AVS+ at all
- They need ancient support DLLs and special loading mechanisms in order to be
used with other Avisynth versions

[StainlessS]

You are the Bee's Bollocks (no that cant be right, how bout the Dogs Knees).

Thank you most graciously kind Sir :)

[cooli]

Hello,

GamMac64 = 404

this link is fine
https://www.dropbox.com/sh/oxx5cm9hkbpj5oz/AAD0QBnTlczv7xW3jEdSjenHa?dl=0&preview=GamMac64_1.09.7z

Is it possible to get the last version?

[Groucho2004]

Hello,

GamMac64 = 404

this link is fine
https://www.dropbox.com/sh/oxx5cm9hkbpj5oz/AAD0QBnTlczv7xW3jEdSjenHa?dl=0&preview=GamMac64_1.09.7z

Is it possible to get the last version?
All links are working in my dropbox directory. Also, GamMac 1.09 is the latest version, at least according to the first post in Stainless' thread.

[StainlessS]

I shall validate latest version v1.10 whan I get any/some indication that the dithering stuff works OK.

[LigH]

I wonder what DropBox messed up this time to make their JavaScript code incompatible with Pale Moon. I need the original Firefox or a Chromium based browser not to see an empty page only. — Or is it the other way round, Pale Moon missing a compatibility tweak?

Not your fault, Groucho. Just annoying to get more and more limited in my free choice of web browsers. I already know I have to abandon Opera 12 as soon as Vivaldi will integrate a mail client, to become again what Opera once used to be.

[Groucho2004]

I wonder what DropBox messed up this time to make their JavaScript code incompatible with Pale Moon. I need the original Firefox or a Chromium based browser not to see an empty page only. — Or is it the other way round, Pale Moon missing a compatibility tweak?
I added an alternative location (MEGA). Let me know how that works.

[LigH]

Yes, MEGA works. And DropBox would work as well with a User Agent override or Content Security Check exception in Pale Moon (https://forum.palemoon.org/viewtopic.php?f=3&t=16805) (overly specific browser check).

I found MediaFire to work quite reliably, during the last years of hosting x265 builds and other stuff. But yes, once I had to clear old cookies after updating PM...

[LigH]

Pale Moon 27.5.1 takes now special care of DropBox... :rolleyes:

[Groucho2004]

Rebuilt EEDI2 (x86 & x64) with AVS+ 2.6 headers.

[Groucho2004]

Added 64 bit build of StainlessS' ColorYUV plugin.

[StainlessS]

Verification, as ColorYUV2, bit different.

Thanks awfully, you Gouchy ol' wastrel :)

[Groucho2004]

Verification, as ColorYUV2, bit different.Elaborate please.

[StainlessS]

ColorYUV2 has also s style power curve (r curve being as Gamma).

EDIT: eg
https://s20.postimg.cc/e05bqbhq5/Coloryuv2_Graf.jpg (https://postimg.cc/image/r4aw309rt/)

EDIT:
Elaborate please.
Dont know, bit pissed :( EDIT: forgot.

EDIT: Thought Ligh asked for elaboration on ColorYUV2, I'm a mess, bit pissed, arh, who cares. :)
(Maybe got threads a bit muxed up)

[LigH]

No, I just read. :cool: Cool tool!

[StainlessS]

As Groucho2004 seems to be Avisynth leading proponent of Ricktendo's All-In-One CPP Runtimes, thought I'de post this here

Visual C++ Runtimes AIO (10-15-2017), [15 Oct 2017 update - CPP runtimes for VS 2005 -> VS 2017 + other stuff, x86 and x64),

Here:- http://repacks.net/forum/viewtopic.php?f=6&t=125

or direct link to MEGA file host:- https://mega.nz/#F!SpQgkAII!07wjhxEh89CUQlTV8oXPyg

VBCRedist_AIO_x86_x64.exe is the file, right click and Download (~27.7MB non-zipped).

Cheers Grouchy.

[Groucho2004]

As Groucho2004 seems to be Avisynth leading proponent of Ricktendo's All-In-One CPP Runtimes, thought I'de post this here
There's another AIO (http://repacks.net/forum/viewtopic.php?f=6&t=237) on repacks.net which works just as well, is more up-to-date, has a selection dialog and uninstalls previous runtimes. Very nice.

[LigH]

And absolutely up-to-date: DIY with wsusoffline.net (http://download.wsusoffline.net/) and download installers of today from Microsoft.

[Groucho2004]

And absolutely up-to-date: DIY with wsusoffline.net (http://download.wsusoffline.net/) and download installers of today from Microsoft.
There are no options for only selecting the runtimes, pack the runtimes into a nice AIO, automatically uninstall old runtimes and avoiding the idiotic installer duplicates.

I don't see how this is an alternative to the AIOs.

[LigH]

I did not even know that the AIO packs have customized installers with additional features. The original MS installers used to be all I needed ever.

Thank you, another reason to feel stupid now.

[StainlessS]

Thanx G2K4, that one might have confused me as it has day-month-year as opposed to RickTendo month-day-year format in the thread.
A bit bigger @ ~41MB. [EDIT: actually ~44MB]

Ligh, is the latest XP version wsusOffline (v923, I think, leastwise thats what I downed a few weeks ago) still working for XP, does it need the POSReady registry fix to work ?
Have not used wsusOffline since 6th April 2014. [EDIT: with v921 wsusOfflline, so two version increments since I last used]

v923 got 2 hits on VirusTotal when I downloaded it, from Cylance ('Unsafe') and Sophos ML ('heuristic'), obviously both false +ve's.

EDIT: wsusOfflline also updates for security fixes, etc (All fully automated, optional updates for office XP and other stuff, [EDIT: eg PowerShell, WMP, Silverlight, DirectX plus more] with auto reboot/logon/continue where necessary, takes a long time though).
EDIT: Think I read somewhere about a recently released security fix for XP, so that at least might update, think I'll give it a try some time over the next week. (maybe without, and then with PosReady fix).
EDIT: My old version of WinRar baulked at decompress of the VS CPP runtime AIO rar file (some years since update), but 7Zip did the job OK.
I dont like to update WinZip or WinRar, in case of recent better compression modes with recipient user confused as to why his old version WinRar/Winzip will not decompress, I assume that all users have recent versions of 7Zip when I use that.

[Groucho2004]

The original MS installers used to be all I needed ever.Most people use the MS installers and are perfectly happy with them.
another reason to feel stupid now.Hardly. Very few even know of the existence of these custom installers.

[LigH]

@Tinman:

Don't fear false alarms. Heuristics are sensitive. AutoIt binaries may contain powerful libraries.

I have no more XP. "ESR" will have its meaning...

[StainlessS]

Yes indeed Ligh, AutoIt was originally produced for OS rollouts, and so can do pretty much anything at almost any level,
AutoHotKey is spurred off (fork) of AutoIt and is equally powerful.
Perhaps you tried the Sawbones thing, wonder if you managed to get it working OK with QWERTZ keyboard.
The user abolibibelot (AZERTY) did not yet come back to say if further problems in most recent version.
I have no more XP
Arh, you poor poor soul, I lament for you :)

[monohouse]

is there no build of NNEDI3 for AMD ? I am using latest version jspdr and my script on core 2 duo give 0.03 fps/core but on my thuban 1090T I have only 0.01 fps/core

the jspdr version has SSE2 but the thuban has newer instructions that nnedi3 doesn't appear to use (SSE3, SSE4A, 3DNow!, 3DNow+!), it has AVX but there is no AVX in thuban :x

[Groucho2004]

@monohouse
You should post this in jpsdr's nnedi3 thread.

[Groucho2004]

Added TUnsharp 0.9.3 (x64)

[magiblot]

Hello Groucho,

Would it be possible to compile the old LGhost plugin for 64-bit AviSynth+?
Source code (http://nullinfo.s21.xrea.com/data/LGhost0301.zip)

Thanks.

[LigH]

I noticed that AVSMeter 2.7.3 prompts for a location to save logs to. Would it be possible to store a default path / filename[x86|x64] in the INI file instead? I did not put it in an UAC protected directory so I don't need to worry about file access rights that much, they are writeable in '.' for me.

[Groucho2004]

I noticed that AVSMeter 2.7.3 prompts for a location to save logs to. Would it be possible to store a default path / filename[x86|x64] in the INI file instead? I did not put it in an UAC protected directory so I don't need to worry about file access rights that much, they are writeable in '.' for me.v2.7.4 re-introduced the possibility to specify a directory for .log/.csv files through the "LogDirectory" ini option.

[LigH]

Yay! :thanks:

[real.finder]

Hello Groucho,

Would it be possible to compile the old LGhost plugin for 64-bit AviSynth+?
Source code (http://nullinfo.s21.xrea.com/data/LGhost0301.zip)

Thanks.

I wish that too

[Groucho2004]

Hello Groucho,

Would it be possible to compile the old LGhost plugin for 64-bit AviSynth+?
Source code (http://nullinfo.s21.xrea.com/data/LGhost0301.zip)

Thanks.
I'll have a look when I have some more time.

[magiblot]

Cool, Groucho. Please, take all the time you need.

[LigH]

TDeint is also available in 64 bit (linked in the QTGMC – Optional Plugins and Scripts (http://avisynth.nl/index.php/QTGMC#Optional_Plugins_and_Scripts)); can you make TMM available too? Not important, just for completeness.
_

P.S.: Ah, wait, chikuzen released TMM2 (https://github.com/chikuzen/TMM2).

LOL: The missing closing parenthesis from the TMM documentation in the second example persists.

[Groucho2004]

Added a new tool "avsr". See first post for details.

[LigH]

So, verbosely, you could call it the "AviSynth Runner" or similar ... nice tool for experimenting, thank you.

[althor1138]

AVSR sounds pretty interesting to me. Is there a link where I can grab it? The one in the first post doesn't seem to work.

[Groucho2004]

avsr v0.1.1

Added clip info and progress indicator

[Groucho2004]

AVSR sounds pretty interesting to me. Is there a link where I can grab it? The one in the first post doesn't seem to work.In the first post, second paragraph, you'll see a link to the download page (http://www.iol.ie/~schubert/gas/gas.html). The file you want is named "avsr_xxx.7z" where "xxx" stands for the current version. I just updated so that may be the reason why you had a problem downloading.

[althor1138]

I will give it another try sometime later today. Thanks!

[Groucho2004]

avsr v0.1.2

Added SEH (Structured Exception Handling) and additional error handling

[Groucho2004]

So, verbosely, you could call it the "AviSynth Runner" or similar ... Let your imagination run free. :)

[Groucho2004]

avsr v0.1.3

Minor fixes
Cosmetics

[LigH]

Would it be hard to add a datetime column to your downloads table to see which tool has recently been updated? And possibly make it sortable by name or date?

[Groucho2004]

Would it be hard to add a datetime column to your downloads table to see which tool has recently been updated? And possibly make it sortable by name or date?
I added a build date column and sorted it by date/descending. I hope that's sufficient.

In order make the sorting customisable I'll have to look for another tool that generates the table.

[LigH]

No pressure, it was a mere suggestion. What has been done is already useful. Thank you.

[Groucho2004]

avsr v0.1.4

Added console colour scheme
Cosmetics/additional info

[zorr]

Groucho, would it be possible in avsr to write the error messages to standard error instead of standard output like now? Then I could determine in AvisynthOptimizer if the script failed and print the error message from avsr.

[Groucho2004]

Groucho, would it be possible in avsr to write the error messages to standard error instead of standard output like now? Then I could determine in AvisynthOptimizer if the script failed and print the error message from avsr.
Yes, this is already on my todo list for the next version.

[Groucho2004]

avsr v0.1.5

Re-write of command line processing
Added switch "-p" which pauses the program at the end and returns to the console prompt only after pressing a key
All error messages go to stderr
Refactoring


https://i.postimg.cc/FHZ5d92G/avsr.png

[zorr]

Yes, this is already on my todo list for the next version.

Thanks, you always deliver. :)

I tried reading the standard error from avsr (v0.1.5) but it wasn't smooth sailing. Looks like every time the script is run successfully there is still "bad allocation" in the standard error.

When the script fails there is nothing in the standard error (or at least I'm not getting anything).

With version 0.1.4 I am able to read the standard output which contains both normal output and the error message and there's nothing in standard error.

[Groucho2004]

I tried reading the standard error from avsr (v0.1.5) but it wasn't smooth sailing. Looks like every time the script is run successfully there is still "bad allocation" in the standard error.

When the script fails there is nothing in the standard error (or at least I'm not getting anything).

With version 0.1.4 I am able to read the standard output which contains both normal output and the error message and there's nothing in standard error.
Without knowing how you capture the stdout/stderr streams I can't really comment...

Edit:
I tested the stderr output with this command line:
avsr script.avs 2> output.err

The error message(s) were properly re-directed.

[zorr]

Without knowing how you capture the stdout/stderr streams I can't really comment...

Well I could show you the Java code but I don't know if that would do any good... anyway it's a standard method for reading these output streams. There are other ways and I'm going to try them too.


Edit:
I tested the stderr output with this command line:
avsr script.avs 2> output.err

The error message(s) were properly re-directed.

I can confirm, that works. However for some reason I wasn't able to read both the stdout and stderr. I tried to get them into separate files:

avsr script.avs > stdout.txt 2> stderr.txt
There's text in stdout.txt but stderr.txt is empty.

I also tried getting both streams into same file:

avsr script.avs > out.txt 2>&1
In this case the file contains only what was in standard out.

[Groucho2004]

However for some reason I wasn't able to read both the stdout and stderr. I tried to get them into separate files:

avsr script.avs > stdout.txt 2> stderr.txt
There's text in stdout.txt but stderr.txt is empty.

I also tried getting both streams into same file:

avsr script.avs > out.txt 2>&1
In this case the file contains only what was in standard out.
I can confirm this behaviour but so far have not been able to figure out why this happens. Investigating...

[Groucho2004]

However for some reason I wasn't able to read both the stdout and stderr. I tried to get them into separate files:

avsr script.avs > stdout.txt 2> stderr.txt
There's text in stdout.txt but stderr.txt is empty.

I also tried getting both streams into same file:

avsr script.avs > out.txt 2>&1
In this case the file contains only what was in standard out.
@zorr
Found the problem, should be fixed now. Please try this test build (https://www.dropbox.com/s/25oec97odhhmo8y/avsrtest.7z?dl=1).

[zorr]

@zorr
Found the problem, should be fixed now. Please try this test build (https://www.dropbox.com/s/25oec97odhhmo8y/avsrtest.7z?dl=1).

That's better, but still something is not quite right. I can now read the stderr sometimes. But quite often the process "hangs" when I try to read the stream and never returns. The same happens even when I don't try to read the stream but just wait for the avsr process to finish. Note that this doesn't happen when I run avsr from the command line, only when I start it with Java.

Version 0.1.4 behaves differently, it never hangs when I wait for it to finish.

[Groucho2004]

That's better, but still something is not quite right. I can now read the stderr sometimes. But quite often the process "hangs" when I try to read the stream and never returns. The same happens even when I don't try to read the stream but just wait for the avsr process to finish. Note that this doesn't happen when I run avsr from the command line, only when I start it with Java.

Version 0.1.4 behaves differently, it never hangs when I wait for it to finish.
I don't know. I do think that the sentence I highlighted in blue might be a clue. Maybe a timing issue?

[zorr]

I don't know. I do think that the sentence I highlighted in blue might be a clue. Maybe a timing issue?

Progress! Turns out I need to read both streams. If I only read stderr the thing is going to hang when the script is run successfully. With this fix I can get the error message out and display it. :D

[Groucho2004]

avsr 0.1.6

Changed console write functions a bit making it (hopefully) easier to capture stdout/stderr
Removed switch "-p". Will be implemented differently at a later stage.
Various fixes

[Groucho2004]

avsr 0.1.7

Minor fixes/cosmetics

[LigH]

I see AVSMeter 2.86 is new as well.

[Groucho2004]

avsr 0.1.8

Added switch "-i" which displays just the script properties
Error handling improvements

[zorr]

I'd like to request a new feature for the avsr: a switch similar to AVSMeter's -range but it would be a list of frames to run (preferably in the order specified).

I have a new feature planned for the AvisynthOptimizer, an automated frame selection. This switch would make implementing the feature much easier. :thanks:

[Groucho2004]

I'd like to request a new feature for the avsr: a switch similar to AVSMeter's -range but it would be a list of frames to run (preferably in the order specified).Need more info - list of frames from a file or supplied on the command line?

[zorr]

Need more info - list of frames from a file or supplied on the command line?

From the command line. Could be called like this:

avsr script.avs -frames 1,3,8,23,4

[Groucho2004]

avsr 0.1.9

Added switch "-frames=n,n,n...". The user can specify the frames to be read from a script on the command line.

[zorr]

avsr 0.1.9

Added switch "-frames=n,n,n...". The user can specify the frames to be read from a script on the command line.


:thanks: Just what the doctor ordered.

[asarian]

There were some requests to group the plugins built with ICL/VC10 into one thread. While I'm at it I'll include a couple of useful tools. Most of the plugins listed are much faster than the originals due to the Intel compiler optimisations, some just have slight speed improvements.


Didn't you also compile an ICL/VC10 version for VapourSynth?! Or is my memory playing tricks on me? :)

[Groucho2004]

Didn't you also compile an ICL/VC10 version for VapourSynth?I did not.

[Groucho2004]

Added RequestLinear 1.0.5. See first post for details.

[LigH]

Thank you. Can be pretty useful, good to have it solo (PAL people hardly need TIVTC).

[Natty]

any anti ghost plugin in 64bit if possible ��

http://avisynth.nl/index.php/Category:Ghost_Removal

[Groucho2004]

Added ApplyEvery 0.3.1. See first post for details.

[Groucho2004]

Added Ghostbuster 0.2. See first post for details.

[Natty]

:thanks::goodpost:

[Groucho2004]

Added LoadDLL 1.0 (x86 & x64). See first post for details.

[wonkey_monkey]

Does Avsmeter still do some kind of pre-scanning even when -o is specified? I've noticed it takes longer to start processing (pauses on "Frame (current | last): 0 | 10789199"), and updates less frequently, the more frames I give it. The pause and update interval are also dependent on the speed of the filters I'm running.

[Groucho2004]

Does Avsmeter still do some kind of pre-scanning even when -o is specified? I've noticed it takes longer to start processing (pauses on "Frame (current | last): 0 | 10789199"), and updates less frequently, the more frames I give it. The pause and update interval are also dependent on the speed of the filters I'm running.Using the "-o" switch, the initial delay you see is entirely down to how fast Avisynth reads the first frame. With certain filters, this can take a while. I guess it has to do with how many frames a filter wants to read in advance and/or the general processing speed of that filter.

If you have a clip with a very large number of frames (such as your example above), AVSMeter changes the frame interval for measurements. The data for each measurement (fps, CPU usage, etc.) is stored in an array. Increasing the frame interval keeps the memory usage of that array reasonable.

[wonkey_monkey]

Using the "-o" switch, the initial delay you see is entirely down to how fast Avisynth reads the first frame.

VirtualDub doesn't exhibit the same delay. For example:

colorbars.loop(X).blur(1)

This loads instantly in VirtualDub with X=10, X=100, or X=1000, but AVSmeter gets progressively slower to start.

Take out the blur(1) and it's a lot faster, which indicates some frames are being fetched somewhere.

If you have a clip with a very large number of frames (such as your example above), AVSMeter changes the frame interval for measurements. The data for each measurement (fps, CPU usage, etc.) is stored in an array. Increasing the frame interval keeps the memory usage of that array low.

The update speed also changes depending on whether blur(1) is included or not, so it doesn't seem to be purely due to the number of frames.

(This is probably purely academic as 100,000,000 frame scripts are unlikely to come up very often)

[Groucho2004]

VirtualDub doesn't exhibit the same delay. For example:

colorbars.loop(X).blur(1)

This loads instantly in VirtualDub with X=10, X=100, or X=1000, but AVSmeter gets progressively slower to start.
You're right, it has to read the first interval of frames before it can display any meaningful numbers. So, in your example, with X = 10, the interval is 10, with X = 1000, the interval is 1000. You can verify this if you use the -log switch and look at the 'performance data' in the log file.


The update speed also changes depending on whether blur(1) is included or not, so it doesn't seem to be purely due to the number of frames.I don't understand the logic here. 1000 frames with blur() is of course slower than 1000 frames without it.

[wonkey_monkey]

Ugh, never mind, I've finally got it through my thick head.

[Groucho2004]

avsr 0.2.0

Fixed wrong AVISYNTH_INTERFACE_VERSION

[Groucho2004]

Added tritical's ColorMatrix (64 bit), see first post.

[hydra3333]

thank you

[StainlessS]

Lovely, thank you G2K4. :)

[Natty]

:goodpost:tyvm

[Groucho2004]

Added DeSpot 3.6.1.1 (x86 & x64)

[StainlessS]

Damn, nuther one for my archive. :(

Thanx Muchly :)

[Groucho2004]

Damn, nuther one for my archive. :(I'm terribly sorry. :sly:

Interesting fact - I removed all the SSE inline asm stuff and it turns out that it is even slightly faster by just using the C code. I guess the Intel compiler helps a lot (note to self - try MS compiler).

Edit - Intel compiler produces a DLL twice as fast as the one built with VC. :cool:

[StainlessS]

Intel pay for if not on Linux, yes ?
[mucho wonga]

[Groucho2004]

Intel pay for if not on Linux, yes ?Correct.

[VoodooFX]

How to read first test? Did it parsed clip in avs 2.6 test?

https://i.imgur.com/aDf6CZs.png

[Groucho2004]

How to read first test? Did it parsed clip in avs 2.6 test?
The error message seems pretty clear to me. You'll have to elaborate a bit as to what is unclear. Also, posting this in the AVSMeter thread would have made more sense.

[VoodooFX]

The error message seems pretty clear to me. You'll have to elaborate a bit as to what is unclear. Also, posting this in the AVSMeter thread would have made more sense.

Sorry I somehow ended in this tread.
I managed to benchmark script with avsr, can limit for very fast speed be removed in avsmeter, it can be still meaningful (https://forum.doom9.org/showthread.php?p=1883114#post1883114).

[ChaosKing]

Is the download page down?

[Groucho2004]

Is the download page down?
Works for me.

[ChaosKing]

Seems like a dns problem on my machine. Or maybe Quad9 DNS 9.9.9.9 is blocking http://www.iol.ie/~schubert/gas/gas.html 🤔

EDIT it is indeed blocked https://www.quad9.net/result/?url=www.iol.ie

[LigH]

Works for me.

:o Oops, missed a new page.

[Groucho2004]

Seems like a dns problem on my machine. Or maybe Quad9 DNS 9.9.9.9 is blocking http://www.iol.ie/~schubert/gas/gas.html ��

EDIT it is indeed blocked https://www.quad9.net/result/?url=www.iol.ieJust report it as a false positive to Quad9.

[Meloware]

I thought zorr ask for this, a year ago. You obliged him with a " -frames" feature, which seems to require each desired frame to be listed. How about a real “-range” feature? I would love a starting point and number of frames (a range) to render!

My video script and 32 core Threadripper is processing out about 20, 1440 x 1080 frames per second. The CPU utilization is only about 16% and there are around 16 threads doing most of the work. Now, I'd love to discuss how to light up the other cores, in a different conversation, but I could really use this “range” feature. If I start an instance of AVSPmod (64), while Avsr was also running, I can specify a later range of frames to process. AVSPmod happily jumps to the selected start frame and begins to process. Then, my CPU load climbs to 30%.
I suppose I am greedy, but I'd love to see what my rendered frame rate would be, with a 85% CPU load! Is there an easy way to specify a start frame/range in the avsr command line?

[Groucho2004]

I thought zorr ask for this, a year ago. You obliged him with a " -frames" feature, which seems to require each desired frame to be listed. How about a real “-range” feature? I would love a starting point and number of frames (a range) to render!

My video script and 32 core Threadripper is processing out about 20, 1440 x 1080 frames per second. The CPU utilization is only about 16% and there are around 16 threads doing most of the work. Now, I'd love to discuss how to light up the other cores, in a different conversation, but I could really use this “range” feature. If I start an instance of AVSPmod (64), while Avsr was also running, I can specify a later range of frames to process. AVSPmod happily jumps to the selected start frame and begins to process. Then, my CPU load climbs to 30%.
I suppose I am greedy, but I'd love to see what my rendered frame rate would be, with a 85% CPU load! Is there an easy way to specify a start frame/range in the avsr command line?
I'm not sure about the usefulness of what you're trying to do but for now you can use AVSMeter (https://forum.doom9.org/showthread.php?t=174797) to achieve this (switch "-range", see documentation). Don't forget to set the parameter "AllowOnlyOneInstance" in avsmeter.ini to 0 if you want to run multiple instances.

[Meloware]

I have used AVSMeter to check scripts, but wasn't aware you could render as well. Command line rendering is what I have been using AVSr for.
All my script rendering is to frame sequences. This allows me to stop and resume big jobs as my needs for the computer change, over the day. I also have had issues where a script may just stop running. If I were rendering to an avi, everything would break, right? With a frame sequence, I can restart where things left off, with no loss of the work already rendered.

If a script is only running at 20% of my machine's capacity, I can run multiple instances of the script, with each instance rendering different portions of the big job. Unlike rendering to an AVI container, individual numbered frames may be rendered in any order, and make it possible to correct portions of the larger project, without needing to re-render everything from scratch. Each instance might grab it's own 20% of my system resources, but 4 instances combined will give me four times the performance.

When it is all done, I can then review the work and then possibly re-render portions, using different script settings. For these reasons, I feel the ability to suspend and resume script jobs is very powerful and needed.

Thanks, and I will explore doing this with AVSMeter.

[Groucho2004]

OK, I understand now. Multiple instances of AVSMeter should do exactly what you want.

[Atak_Snajpera]

I thought zorr ask for this, a year ago. You obliged him with a " -frames" feature, which seems to require each desired frame to be listed. How about a real “-range” feature? I would love a starting point and number of frames (a range) to render!

My video script and 32 core Threadripper is processing out about 20, 1440 x 1080 frames per second. The CPU utilization is only about 16% and there are around 16 threads doing most of the work. Now, I'd love to discuss how to light up the other cores, in a different conversation, but I could really use this “range” feature. If I start an instance of AVSPmod (64), while Avsr was also running, I can specify a later range of frames to process. AVSPmod happily jumps to the selected start frame and begins to process. Then, my CPU load climbs to 30%.
I suppose I am greedy, but I'd love to see what my rendered frame rate would be, with a 85% CPU load! Is there an easy way to specify a start frame/range in the avsr command line?

Why don't you just use ripbot264 in distributed encoding mode? With 8 servers (2 per numa) you will get constant 100% CPU load.

[Meloware]

OK, I understand now. Multiple instances of AVSMeter should do exactly what you want.

I just tried AVSMeter64 2.9.7. The first instance works, but I am not allowed to run multiple instances from multiple terminal boxes.
I get the message "One instance of AVSMeter is already running"

[poisondeathray]

I just tried AVSMeter64 2.9.7. The first instance works, but I am not allowed to run multiple instances from multiple terminal boxes.
I get the message "One instance of AVSMeter is already running"

Edit the .ini

See post 206


Don't forget to set the parameter "AllowOnlyOneInstance" in avsmeter.ini to 0 if you want to run multiple instances.

[stax76]

I decided to use avsr for learning how to use avisynth and wonder why there is no build or project file? It took me a while to make it compile in VS.

[Groucho2004]

I decided to use avsr for learning how to use avisynth and wonder why there is no build or project file?I build my projects with makefiles from the command line. I can post the makefiles if you want.

[stax76]

I build my projects with makefiles from the command line. I can post the makefiles if you want.

I have it working in Visual Studio 2019 but only x86, with x64 (I've only avs+ x64 installed and I don't have any x86 filters) I get many errors unfortunately.

Severity Code Description Project File Line Suppression State
Error (active) E0167 argument of type "char *" is incompatible with parameter of type "LPWSTR" avsr D:\Projekte\CPP\avsr\common.h 79
Error (active) E0167 argument of type "const char *" is incompatible with parameter of type "LPCWSTR" avsr D:\Projekte\CPP\avsr\avsr.cpp 169
Error (active) E0167 argument of type "const char *" is incompatible with parameter of type "LPCWSTR" avsr D:\Projekte\CPP\avsr\common.h 67
Error (active) E0167 argument of type "char *" is incompatible with parameter of type "LPWSTR" avsr D:\Projekte\CPP\avsr\avsr.cpp 125
Error (active) E0167 argument of type "char *" is incompatible with parameter of type "LPTSTR" avsr D:\Projekte\CPP\avsr\common.h 95
Error (active) E0167 argument of type "char *" is incompatible with parameter of type "LPCWSTR" avsr D:\Projekte\CPP\avsr\avsr.cpp 125
Error C4996 '_vsnprintf': This function or variable may be unsafe. Consider using _vsnprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_WARNINGS. See online help for details. avsr D:\Projekte\CPP\avsr\utility.h 174
Error C4996 '_snprintf': This function or variable may be unsafe. Consider using _snprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_WARNINGS. See online help for details. avsr D:\Projekte\CPP\avsr\exception.h 1465
Error C4996 '_snprintf': This function or variable may be unsafe. Consider using _snprintf_s instead. To disable deprecation, use _CRT_SECURE_NO_WARNINGS. See online help for details. avsr D:\Projekte\CPP\avsr\utility.h 253
Error C2664 'UINT GetWindowsDirectoryW(LPWSTR,UINT)': cannot convert argument 1 from 'char [1026]' to 'LPWSTR' avsr D:\Projekte\CPP\avsr\common.h 79
Error C2664 'UINT (LPTSTR,UINT)': cannot convert argument 1 from 'char [1026]' to 'LPTSTR' avsr D:\Projekte\CPP\avsr\common.h 95
Error C2664 'HMODULE LoadLibraryW(LPCWSTR)': cannot convert argument 1 from 'const char [9]' to 'LPCWSTR' avsr D:\Projekte\CPP\avsr\avsr.cpp 169
Error C2664 'HMODULE GetModuleHandleW(LPCWSTR)': cannot convert argument 1 from 'const char [13]' to 'LPCWSTR' avsr D:\Projekte\CPP\avsr\common.h 67
Error C2664 'DWORD GetModuleFileNameW(HMODULE,LPWSTR,DWORD)': cannot convert argument 2 from '_Ty *' to 'LPWSTR'
with
[
_Ty=char
] avsr D:\Projekte\CPP\avsr\utility.h 248
Error C2664 'DWORD GetModuleFileNameW(HMODULE,LPWSTR,DWORD)': cannot convert argument 2 from 'char [2048]' to 'LPWSTR' avsr D:\Projekte\CPP\avsr\exception.h 67
Error C2664 'DWORD GetFullPathNameW(LPCWSTR,DWORD,LPWSTR,LPWSTR *)': cannot convert argument 1 from 'char *' to 'LPCWSTR' avsr D:\Projekte\CPP\avsr\avsr.cpp 125
Error C2664 'DWORD GetFileVersionInfoSizeW(LPCWSTR,LPDWORD)': cannot convert argument 1 from '_Ty *' to 'LPCWSTR'
with
[
_Ty=char
] avsr D:\Projekte\CPP\avsr\utility.h 258
Error C2664 'DWORD GetFileAttributesW(LPCWSTR)': cannot convert argument 1 from 'const _Elem *' to 'LPCWSTR'
with
[
_Elem=char
] avsr D:\Projekte\CPP\avsr\utility.h 137
Error C2664 'DWORD FormatMessageW(DWORD,LPCVOID,DWORD,DWORD,LPWSTR,DWORD,va_list *)': cannot convert argument 5 from 'char [2048]' to 'LPWSTR' avsr D:\Projekte\CPP\avsr\utility.h 231
Error C2664 'BOOL VerQueryValueW(LPCVOID,LPCWSTR,LPVOID *,PUINT)': cannot convert argument 2 from 'const char [2]' to 'LPCWSTR' avsr D:\Projekte\CPP\avsr\utility.h 272
Error C2664 'BOOL GetFileVersionInfoW(LPCWSTR,DWORD,DWORD,LPVOID)': cannot convert argument 1 from '_Ty *' to 'LPCWSTR'
with
[
_Ty=char
] avsr D:\Projekte\CPP\avsr\utility.h 264

[Groucho2004]

No idea (except C4996). Search MSDN for these errors. The code builds fine with the VS2010 compiler, linker and libraries for x86 and x64.

[stax76]

It's running now in x64 after changing few lines, thanks Groucho2004.

[wonkey_monkey]

It's running now in x64 after changing few lines, thanks Groucho2004.

For the benefit of others, what were the few things?

In my experience those kinds of errors about char* conversions are usually solved by changing the project's Character Set property.

[Groucho2004]

For the benefit of others, what were the few things?

In my experience those kinds of errors about char* conversions are usually solved by changing the project's Character Set property.That sounds plausible since newer versions of VS have UNICODE defined by default (as far as I recall).

[stax76]

My changes were not perfect like changing func to funcA or LPWSTR to LPSTR.

[Groucho2004]

My changes were not perfect like changing func to funcA or LPWSTR to LPSTR.Yeah, not really the way to do it. Just change it in the preprocessor definitions (UNICODE -> MBCS).

[stax76]

For my purpose it's fine, I'm much of C++ n00b, only piece of C++ code I still own is this:

https://github.com/stax76/OpenWithPlusPlus/tree/master/OpenWithPPShellExtension

It was written 10-15 years ago and meanwhile I have forgotten few things.

When I've finished this project I'm doing for staxrip (involves avisynth, vapoursynth and hw accelerated rendering) my C++ knowledge will be improved 1-2 levels, I'm still in the learning and experimenting phase and don't try to rush it.

[Groucho2004]

avsr 0.2.1
- Updated some internal string formatting functions
- Added "-range" switch


How to use the "-range" switch:

"-range=first frame, last frame"
Specifies the range of frames to be processed. For example, "-range=1500,1999" processes 500 frames starting at frame 1500. Without "-range" avsr processes all frames of the script.

[kedautinh12]

I see v2.6 of colormatrix add Rec.2020 support.
https://github.com/sorayuki/ColorMatrix/release

Can you compile x64 of it?

[Groucho2004]

ColorMatrix 2.6.0.0 (x86 & x64)
tritical's ColorMatrix with updated AVS2.6 headers and SoraYuki's REC.2020 addition. Documentation included.
Note:
These ColorMatrix builds do not have the 'opt' parameter. The binaries produced with the Intel C compiler using C/C++ code are at least as fast as the ones with the inline asm.

[kedautinh12]

ColorMatrix 2.6.0.0 (x86 & x64)
tritical's ColorMatrix with updated AVS2.6 headers and SoraYuki's REC.2020 addition. Documentation included.
Note:
These ColorMatrix builds do not have the 'opt' parameter. The binaries produced with the Intel C compiler using C/C++ code are at least as fast as the ones with the inline asm.

Thanks, but where your link download??

[Zetti]

You should check first post.

[kedautinh12]

You should check first post.

I was checked but only have colormatrix 2.5.0.1 x64

[Groucho2004]

I was checked but only have colormatrix 2.5.0.1 x64
Try refreshing the download page.

[Groucho2004]

Added SSIM 0.25.1.0 (x86 & x64). See first post.

[Groucho2004]

Added AvsRecursion 0.2 (x86 & x64).

[Selur]

btw, what happend to TDeint?

[Groucho2004]

btw, what happend to TDeint?If you're referring to a 64 bit version I think you'll find the answer here (https://forum.doom9.org/showthread.php?p=1844348#post1844348). You do know that this forum has a search function, right? :sly:

[tormento]

If you're referring to a 64 bit version I think you'll find the answer here (https://forum.doom9.org/showthread.php?p=1844348#post1844348).
Who did that? Tritical, Vit or you?

[Selur]

@Groucho2004: I used the search but must have overlooked the post.
Also the 32bit version points back to this thread.

[Groucho2004]

@Groucho2004: I used the search but must have overlooked the post.
Also the 32bit version points back to this thread.I see. I never made a 64 bit version. However, I'll put up the 32 bit version again.

[ChaosKing]

btw, what happend to TDeint?

I also made a git repo for TDdeint (64bit binary included) https://github.com/avisynth-repository/TDeint

[Groucho2004]

I also made a git repo for TDdeint (64bit binary included) https://github.com/avisynth-repository/TDeintAh, OK. So I don't have to make a build. :)

[StainlessS]

ChaosKing,
Putting a version number on a zip is rarely a bad idea [Or leave it as it is and remain King of Chaos] :)
[Hope Wonkey sees this]

[tormento]

I was checking if some plugins were 2.5 or 2.6 and I got this strange error at the end (latest AVS Meter):

[Plugin errors/warnings]
__________________________________________________________________________________________________________________

"D:\Programmi\Media\AviSynth+\plugins64+\MVTools-2.7.41-pinterf.dll" requires the FFTW library for some functions.

Note: 'libfftw3f-3.dll' can be downloaded here:
http://www.fftw.org/install/windows.html

'libfftw3f-3.dll' should be placed in "System32" and/or "SysWoW64"
__________________________________________________________________________________________________________________

I never knew MVTools need FFTW! :confused:

Any hint?

[Groucho2004]

I never knew MVTools need FFTW! :confused:

Any hint?The DCT general transform uses the fftw library.

[tormento]

The DCT general transform uses the fftw library.
Luckily I never used that function. Thanks!

[StainlessS]

DCT used in some functions with DCT parameter non 0 {where requires FFTW library libfftw3f-3.dll or fftw3.dll}, [Non default can be very slow]

MAnalyse

MAnalyse (
clip super,
int blksize (8),
int blksizeV (blksize),
int levels (0),
int search (4),
int searchparam (2),
int pelsearch (pel),
bool isb (false),
int lambda (0 | 1000*blksize*blksizeV/64),
bool chroma (true),
int delta (1),
bool truemotion (true),
int lsad (400 | 1200),
int plevel (0 | 1),
bool global (truemotion),
int pnew (0 | 50),
int pzero (pnew),
int pglobal (0),
int overlap (0),
int overlapV (overlap),
string outfile (""),
int dct (0), <<<<<<<<<<<<<<<<<<<<
int divide (0),
int sadx264 (0),
int badSAD (10000),
int badrange (24),
bool isse,
bool meander (true),
bool temporal (false),
bool trymany (false),
bool multi (false),
bool mt (true),
int scaleCSAD (0)
)



dct

Using of block DCT (frequency spectrum) for blocks difference (SAD) calculation. In particular it can improve motion vector estimation at luma flicker and fades.
0 Usual spatial blocks, do not use DCT.
1 Use block DCT instead of spatial data (slow for block size 8x8 and very slow for other sizes).
2 Mixed spatial and DCT data; weight is dependent on mean frame luma difference.
3 Adaptive per-block switching from spatial to equal-weighted mixed mode (experimental, a little faster).
4 Adaptive per-block switching from spatial to mixed mode with more weight of DCT (experimental, a little faster).
5 SATD instead of SAD for luma.
6 Same as 2 only use SATD.
7 Same as 3 only use SATD.
8 Same as 4 only use SATD.
9 Similar to 2, use SATD and weight ranges from SAD only to equal SAD & SATD.
10 Similar to 3/4,use SATD weight is on SAD, only on strong luma changes.


+ MRecalculate.

[Groucho2004]

Strictly, it should say "... in some functions with certain parameter values". The current warning could lead people to think that they need fftw in any case.

[kedautinh12]

here have code Tdeint 1.2 32&64bit but he don't releases it: https://github.com/pinterf/TIVTC

[Groucho2004]

here have code Tdeint 1.2 32&64bit but he don't releases it: https://github.com/pinterf/TIVTCInteresting. Have you considered asking pinterf about this?

[kedautinh12]

Interesting. Have you considered asking pinterf about this?

No, i just find out when check tivtc new releases or not

[Groucho2004]

No, i just find out when check tivtc new releases or notDon't know what that means.

Anyway, stay safe, wash your hands and don't touch your face.

[StainlessS]

Pinterf will do it when he feels its ready.

[kedautinh12]

Don't know what that means.

Anyway, stay safe, wash your hands and don't touch your face.
Sorry, my English is bad, i don't asking him yet

[Reel.Deel]

here have code Tdeint 1.2 32&64bit but he don't releases it: https://github.com/pinterf/TIVTC

I am sure he will release it when it's ready. 3 days ago:

It is a big achievement that last week I made TDeint compilable on msvc x64 and clang as well. More code clean will follow, drop yuy2 and add yv16.

[Groucho2004]

Sorry, my English is bad, i don't asking him yetWell, you have your answer now. I don't use TIVTC so I don't really follow its development. I was just a bit stumped when your question popped up in this thread.

[Groucho2004]

Added Gavino's GScript 1.1, 32 & 64 bit.

[Groucho2004]

Updated GScript (32 bit). It wasn't working with Avisynth < 2.6 because of the new AVS+ headers. The 32 bit version is now using the "old" header (AvisynthPluginInit2).

[StainlessS]

Thanks for all your efforts :)

[Groucho2004]

Updated GScript (32 bit). It wasn't working with Avisynth < 2.6 because of the new AVS+ headers. The 32 bit version is now using the "old" header (AvisynthPluginInit2).Of course, since it's using 'AvisynthPluginInit2', the 32 bit version now crashes with AVS+ 3.5.2. :devil:

I guess I'll have to release 3 versions of the plug.

[Groucho2004]

Updated GScript. There are now 3 versions:

GScript_25_32.dll - 32 bit, AVS2.5 interface
GScript_26_32.dll - 32 bit, AVS2.6 interface
GScript_26_64.dll - 64 bit, AVS2.6 interface

[StainlessS]

Thanks GG, those 3 versions are rarely a bad idea.

[Groucho2004]

Interestingly, if I put both 32 bit versions (2.5 and 2.6) into the auto-load directory, every Avisynth version works with GScript. Since I'm not quite sure why, I advise against doing that.

[Groucho2004]

Thanks GG, those 3 versions are rarely a bad idea.I wish 2.5.8 would just die. Maybe we can send it to a farm in the countryside where it can live out the rest of its days. :sly:

[StainlessS]

And what bout the guy thats still using 2.57 [was in pole a couple years back, bet he's just too embarrassed to admit it now].
Was I think 3 in last pole still using 2.58, scaled up for global population, how many is that, [more than 10 ?]

[videoh]

And we thought DLL hell was bad!

[Groucho2004]

And what bout the guy thats still using 2.57I remember that. As far as I know, 2.5.7 is already happily running around in the fields 'upstate'. So, 2.5.8 can join it now (with the appropriate social distancing of course).

[manolito]

Not for me...

From my notes about the ancient but still very useful DVD2SVCD software:
In DVD2SVCD the preview only works under AviSynth 2.56 and 2.57.
If a newer AviSynth version is installed, the file avisynth.dll v 2.57 needs to be copied into the DVD2SVCD folder.

[Groucho2004]

avsr v0.2.2

Added new color space definitions (CS_RGBP8 and CS_RGBAP8), removed redundant CS_RGBP / CS_RGBAP
Fixed a version resource issue
Updated AVS+ headers
Increased range for AVISYNTH_INTERFACE_VERSION detection

[real.finder]

hi, what about this https://forum.doom9.org/showthread.php?p=1908063#post1908063

is it forgotten?

[Groucho2004]

hi, what about this https://forum.doom9.org/showthread.php?p=1908063#post1908063

is it forgotten?Not forgotten. I haven't had the time to do it.

[Reel.Deel]

@Groucho2004

Hi. Any possibility of a DeSpot update? cretindesalpes modified the plugin (http://ldesoras.free.fr/prod.html#src_ditheravsi) some time ago:

v3.6.3.0
A modified version of the DeSpot plug-in, allowing to dump the masks of the detected spots in an Aegisub file for further editing.


Version 3.6.3.0 (October 8, 2013):
* Added the mc, spotmax1 and spotmax2 parameters.
Version 3.6.2.0 (August 19, 2011):
* Added the outfile parameter.

[Groucho2004]

Any possibility of a DeSpot update?Sure, will do.

[Groucho2004]

Updated DeSpot to v3.6.3.0.

[Reel.Deel]

Updated DeSpot to v3.6.3.0.

Thank you!

[Drmsy]

Updated DeSpot to v3.6.3.0.
:goodpost:
Thanks

[tormento]

Updated DeSpot to v3.6.3.0.
What is the difference with undot?

[StainlessS]

What is the difference with undot?
Spots can be a quite big clump of pixels.

EDIT:
From Undot
UnDot is a simple median filter for removing dots, that is stray orphan pixels and mosquito noise. It basicly just clips each pixel value to stay within min and max of its eight surrounding neighbors.
so single pixel noise, purely spacial.

DeSpot is temporal detector, if spot exists on current frame but not on previous or next then can be detected. [rough description]
EDIT: See a 'spot' here:- http://www.avisynth.nl/users/fizick/despot/despot.html

Thanks Grouchy :)

EDIT: Despot on wiki [updated as per Grouchy post below, thanks Reel.Deal]:- http://avisynth.nl/index.php?title=DeSpot&redirect=no

[Groucho2004]

Reel.Deal updated the documentation and links on avisynth.nl.

Thanks for the continued Wiki maintenance and updates, RD!

[vcmohan]

Will avisynth+ 3.6 version be included in your universal installer? Or can I download the 3.6 version and replace the 3.5 in the avsplus dll folder?

[Groucho2004]

Will avisynth+ 3.6 version be included in your universal installer? Or can I download the 3.6 version and replace the 3.5 in the avsplus dll folder?There are 2 versions, one with 3.5.1 AVS+ and one with 3.6.0 AVS+. Just download the one you want.
The only differences between the two are the DLLs in the AVS+ directories and the setavs.cmd batch file.

[vcmohan]

I am updating my plugins to work in v8 as well as in pre v8 versions. So I need to alternate between avsplus 3.5 and 3.6 . Can some thing be done so that by just running your command script I can alternate between 3.5 and 3.6 easily?

[StainlessS]

Groucho will give definitive answer, but below what I do.
In eg AVSPLus_x86 directory
https://i.postimg.cc/0Ns55yGt/Untitled-00.jpg (https://postimages.org/)
I just make directories, each containing Avisynth.dll and DevIL.dll for the folder named version.
When I want to change, I just copy [not move] the dll's into the main AVSPLus_x86 dir,
Cmd file shows wrong version info on menu, but I can live with that, the
info at top of menu for current installed version for x86 and x64 show correct current installed version.

EDIT: If you update to a later version Universal Installer,
then copy over the substitute version directories into the new installer directories, and make new substitute version directory for the new native version, and copy
over/duplicate the Avisynth.dll and DevIL.dll to it.

[Groucho2004]

I am updating my plugins to work in v8 as well as in pre v8 versions. So I need to alternate between avsplus 3.5 and 3.6 . Can some thing be done so that by just running your command script I can alternate between 3.5 and 3.6 easily?
What you could do is download both, 3.5.1 and 3.6.0 Installers and rename the directories to AvisynthRepository_351 and AvisynthRepository_360 respectively. Then you rename the setavs.cmd files the same way and create shortcuts to both of them.

Seems (slightly) less complicated to what Stainless suggested. Just my 2 cents.

[StainlessS]

That just sounds too easy [it'll make you lazy].

Fishy story:- https://metro.co.uk/video/doctors-surgically-remove-dead-fish-mans-rectum-sat-it-2187564/

EDIT:
Seems (slightly) less complicated to what Stainless suggested.
Maintaining two (or more) lots of current plugins directories seems somewhat more complicated to me.

[Groucho2004]

Fishy story:- https://metro.co.uk/video/doctors-surgically-remove-dead-fish-mans-rectum-sat-it-2187564/The all important question is - Was the fish dead before it entered the rectum or did it die inside after the man had his way with it?

[StainlessS]

"Give it to us raw — and wriggling! You keep nasty chips!".
[I hear raw and wriggling Conger eel is just the job].

EDIT: To below.
Did you mean: Gagh?
Nope, Conger for me, although you could try that earther ersatz Gagh, Cha ruoi, [made with ragworm, watch out, they bite],
but me dont like that foreign stuff, so you know where you can stick it.

[LigH]

:confused: Did you mean: Gagh? :sly:

[Groucho2004]

Added AudioGraph 0.0.0 (x86 & x64).

[Groucho2004]

Updated AudioGraph to v0.0.1

- Added version resource
- Provided default values for:
graph_scale (0),
frames_either_side (25),
middle_colour ($00FF00),
side_colour ($7F7F7F)

Please note that the parameters now don't have a underscore prefix.

[Dogway]

Latest DesPot is crashing on AVS+ 3.6.1, using also AvsPmod v2.6.3.1 x64. Does it have any dependency?

Over YV12, 8-bit:
DeSpot(pwidth=12,pheight=12,p1=16, p2=8,mthres=8,ranked=false,merode=33,show=0,mark_v=255,sign = 1)

[Groucho2004]

Latest DesPot is crashing on AVS+ 3.6.1, using also AvsPmod v2.6.3.1 x64. Does it have any dependency?

Over YV12, 8-bit:
DeSpot(pwidth=12,pheight=12,p1=16, p2=8,mthres=8,ranked=false,merode=33,show=0,mark_v=255,sign = 1)
New version uploaded.

[LigH]

The name means de-spot. ;)

[Boulder]

The name means de-spot. ;)

Doesn't Despot, or just despot, describe these times much better? :D

Thanks for the new version, I'll check it as soon as I can update the dll.

[StainlessS]

The name De-Spot is plainly clear,

Despot,
noun. a king or other ruler with absolute, unlimited power; autocrat. any tyrant or oppressor. History/Historical. an honorary title applied to a Byzantine emperor, afterward to members of his family, and later to Byzantine vassal rulers and governors.

I know that G2K4 can seem a bit severe, but he is in no way responibile for everyones perceptions of him,
and quite frankly It upsets me that people would think of him in that way. I'm guessin that he is probably a really nice flellow. (given the chance).

[Groucho2004]

I know that G2K4 can seem a bit severe, but he is in no way responibile for everyones perceptions of him,
and quite frankly It upsets me that people would think of him in that way.Whaddaya talkin' about? Put the bong down (or the pint, whatever applies). :D

[StainlessS]

It seemed a good idea at the time, but then me is bit pissed, and is likely gonna be a bit more pissed too,,
and given that I''m listenening to the pipes, i and probalbly gonna be a bit more, un-nice,
but I'll be lovely when In wake up, jjust have not done it yet.

I'm quite sure that G2K4 is lovely, depite what some others may habitually say bout him, he is my friend, I think, he is a bt strange but wonderful,as we all are.

EDIT: the pipes, well actuall, jeane michell Jarr equinox, always gets me in the mood for discussion.

[Groucho2004]

but then me is bit pissed, and is likely gonna be a bit more pissed tooGo easy on the booze, I had a rude awakening 3 weeks ago at the doctor, my liver took quite a beating over last decades. I've been dry for 3 weeks! Withdrawal symptoms are a bitch but I think I have the worst behind me (I hope).

[Groucho2004]

he is my friend, I think, he is a bt strange but wonderful,as we all are.Yep, strange and wonderful, you nailed it buddy. :D

[FranceBB]

Whaddaya talkin' about? Put the bong down (or the pint, whatever applies). :D

The UK scheme is "eat out to help out" but I think he heard "drink out to help out" xD
But what can I say, even Jacob Rees Mogg said that he was gonna drink a yard of ale, so... xD


I'm quite sure that G2K4 is lovely, depite what some others may habitually say bout him, he is my friend, I think, he is a bt strange but wonderful,as we all are.

Yeah, he definitely is. :)
By the way, I consider many Doom9 users as "friends" as replying to you on this forum is way better than posting memes on any social network. Oh and it's not just funny or entertaining, but also useful ('cause we're also helping other people) and enlightening ('cause I learn new stuff every time despite having been an Avisynth user for years). :)

I had a rude awakening 3 weeks ago at the doctor, my liver took quite a beating over last decades. I've been dry for 3 weeks! Withdrawal symptoms are a bitch but I think I have the worst behind me (I hope).

Oh... sorry to hear that...
Take care, you both.

[Boulder]

The name De-Spot is plainly clear,

Despot,

I know that G2K4 can seem a bit severe, but he is in no way responibile for everyones perceptions of him,
and quite frankly It upsets me that people would think of him in that way. I'm guessin that he is probably a really nice flellow. (given the chance).

My remark was more like pointing out that the world out there seems to be full of despots so why not have a dedicated plugin :)

[Dogway]

Thanks for the update! I got fooled unconsciously by the syllabus break in des-pot.

I noticed you updated recently ColorMatrix but it doesn't support HDR (actually meant HBD), what is the easiest way to change matrix in HDR nowadays? I know HDRTools but it requires ConvertToDoubleWidth().

Take care!

[FranceBB]

ColorMatrix but it doesn't support HDR, what is the easiest way to change matrix in HDR nowadays? I know HDRTools but it requires ConvertToDoubleWidth().

Well actually I think you're referring to HDRCore which does indeed work in 16bit interleaved and doesn't support HDR afaik, but there's HDRTools by Jean Philippe that definitely does support HDR HLG, HDR PQ, Linear BT709 with three tone-mapping algorithms (Hable, Mobius, Reinhard) and Linear BT2020 and it works in 8/10/12/14/16bit planar. https://forum.doom9.org/showthread.php?t=174248

There's also DGTonemap by Donald Graft.

And last but not least, there are my manually written matrices of linear transformation (LUTs): https://forum.doom9.org/showthread.php?t=176091


Pick your favorite one.

Cheers,
Frank.

[Groucho2004]

I noticed you updated recently ColorMatrix but it doesn't support HDR, what is the easiest way to change matrix in HDR nowadays? I know HDRTools but it requires ConvertToDoubleWidth().I don't know much about the HDR stuff, I suggest you try out the options outlined by FranceBB.