Welcome to Doom9's Forum, THE in-place to be for everyone interested in DVD conversion.

Before you start posting please read the forum rules. By posting to this forum you agree to abide by the rules.

 

Go Back   Doom9's Forum > Capturing and Editing Video > Avisynth Usage

Reply
 
Thread Tools Search this Thread Display Modes
Old 20th October 2020, 07:28   #641  |  Link
LigH
German doom9/Gleitz SuMo
 
LigH's Avatar
 
Join Date: Oct 2001
Location: Germany, rural Altmark
Posts: 6,753
Seems that one calculates in Luma units and the other in Chroma units, hence half the number for YUV 4:2:x chroma subsampling... just a guess.
__________________

New German Gleitz board
MediaFire: x264 | x265 | VPx | AOM | Xvid
LigH is offline   Reply With Quote
Old 2nd November 2020, 22:41   #642  |  Link
real.finder
Registered User
 
Join Date: Jan 2012
Location: Mesopotamia
Posts: 2,587
Quote:
Originally Posted by tormento View Post
LWLibavVideoSource("E:\in\1_31 Tokyo godfathers\godfathers.lwi")
error: KNLMeansCL: INT10 require 'channels = YUV'!

LWLibavVideoSource("E:\in\1_31 Tokyo godfathers\godfathers.lwi", format="YUV420P10")
error: KNLMeansCL: INT10 require 'channels = YUV'!

LWLibavVideoSource("E:\in\1_31 Tokyo godfathers\godfathers.lwi", format="YUV420P16")
Working
Quote:
Originally Posted by real.finder View Post
KNLMeansCL has problems with 10 bit clip https://github.com/Khanattila/KNLMea...ment-469839637
Quote:
Originally Posted by tormento View Post
I saw there is a workaround setting YUV, is it possible to create a script variant for that specific case, as all the H.265 HDR material out there is 10 bit?
Quote:
Originally Posted by real.finder View Post
YUV in KNLMeansCL need 4:4:4 chroma, if your source is 444 then it should work with 10 bit, but in your case (4:2:0) 16 bit is better since going from 420 to 444 is not recommended
Quote:
Originally Posted by tormento View Post
Thanks man!

So if the source is 444, SMDegrain should work?
Quote:
Originally Posted by real.finder View Post
yes

Quote:
Originally Posted by tormento View Post
Just a question arise to my mind: Hi10P is 10 bit. You tell me to use 16 bit BUT at the end I have to dither to go back to 10 bit again to encode as proper format. Isn't better to find a way to stay in 10 bit with no dithering at the end?
Quote:
Originally Posted by real.finder View Post
more bits = better

anyway you can chose another prefilter other than 4 or ask KNLMeansCL developer to make it work with grayscale in any bit depth (seems he only against 420 with 10 bit which is not being used in SMDegrain)
Quote:
Originally Posted by tormento View Post
Yes unless the standard for UHD is 10bit and you need to dither at the end

That's why I was asking for a solution to have 10 bit chain
thanks to pinterf, KNLMeansCL soon will work with these cases

test build for now https://drive.google.com/file/d/1Yi8...ew?usp=sharing (it's not final version as pinterf said, for testing only)
__________________
See My Avisynth Stuff
real.finder is offline   Reply With Quote
Old 13th November 2020, 15:40   #643  |  Link
tormento
Acid fr0g
 
tormento's Avatar
 
Join Date: May 2002
Location: Italy
Posts: 2,542
Quote:
Originally Posted by real.finder View Post
thanks to pinterf, KNLMeansCL soon will work with these cases
Could you introduce some presets in SMDegrain to work with 8 bit, 10 bit and HDR arbitrary input/output combination? I.e. input 8 bit -> output 10 bit, input 10 bit -> output 8 bit, etc.
__________________
@turment on Telegram
tormento is offline   Reply With Quote
Old 13th November 2020, 16:56   #644  |  Link
real.finder
Registered User
 
Join Date: Jan 2012
Location: Mesopotamia
Posts: 2,587
Quote:
Originally Posted by tormento View Post
Could you introduce some presets in SMDegrain to work with 8 bit, 10 bit and HDR arbitrary input/output combination? I.e. input 8 bit -> output 10 bit, input 10 bit -> output 8 bit, etc.
you can only do input 8 bit -> output 16 bit with n16_out since it supported in mvtools, others I think there are no benefit from them (pinterf think same)

in any case you can always use convertbits() to do these (like input 8 bit -> output 10 bit) at the cost of some speed losing
__________________
See My Avisynth Stuff
real.finder is offline   Reply With Quote
Old 14th December 2020, 22:12   #645  |  Link
PoeBear
Registered User
 
Join Date: Jan 2017
Posts: 48
Deblock_QED_MT2 seems to crash 3.6.1 (and 3.6.2-test6). Works fine in 3.5.1

Any ideas what could be tweaked in the script to get it working again?
PoeBear is offline   Reply With Quote
Old 15th December 2020, 00:10   #646  |  Link
real.finder
Registered User
 
Join Date: Jan 2012
Location: Mesopotamia
Posts: 2,587
Quote:
Originally Posted by PoeBear View Post
Deblock_QED_MT2 seems to crash 3.6.1 (and 3.6.2-test6). Works fine in 3.5.1

Any ideas what could be tweaked in the script to get it working again?
update the used plugins
__________________
See My Avisynth Stuff
real.finder is offline   Reply With Quote
Old 15th December 2020, 01:58   #647  |  Link
PoeBear
Registered User
 
Join Date: Jan 2017
Posts: 48
Quote:
Originally Posted by real.finder View Post
update the used plugins
Doh! *facepalm*

DCTFilter was my culprit
PoeBear is offline   Reply With Quote
Old 26th December 2020, 03:44   #648  |  Link
hello_hello
Registered User
 
Join Date: Mar 2011
Posts: 4,823
real.finder,
Using the latest neo_dfttest (r7) here, despite the documentation, this produces an error.
neo_dfttest(fft_threads=2)
"neo_dfttest does not have a named argument fft_threads"
Therefore your sneo_dfttest function can only use dfttest. Unless it's an oddity specific to the XP version. I haven't tried Win7 yet.

And I know it's not your plugin, but is neo_dfttest normally slower than dfttest, or is that an XP thing? I haven't played with it on it's own much yet, but it's obvious it's output differs from dfttest quite a bit, so I'm wondering if that makes supporting it for QTGMC a good idea. neo_fft3d is also slower than FFT3DFilter. At least on my old XP PC.

Cheers.

Last edited by hello_hello; 26th December 2020 at 04:21.
hello_hello is offline   Reply With Quote
Old 26th December 2020, 04:52   #649  |  Link
kedautinh12
Registered User
 
Join Date: Jan 2018
Posts: 2,153
Please use this (compile by real.finder) have ncpu
https://forum.doom9.org/showpost.php...&postcount=153
kedautinh12 is offline   Reply With Quote
Old 26th December 2020, 05:39   #650  |  Link
real.finder
Registered User
 
Join Date: Jan 2012
Location: Mesopotamia
Posts: 2,587
Quote:
Originally Posted by hello_hello View Post
real.finder,
Using the latest neo_dfttest (r7) here, despite the documentation, this produces an error.
neo_dfttest(fft_threads=2)
"neo_dfttest does not have a named argument fft_threads"
Therefore your sneo_dfttest function can only use dfttest. Unless it's an oddity specific to the XP version. I haven't tried Win7 yet.

And I know it's not your plugin, but is neo_dfttest normally slower than dfttest, or is that an XP thing? I haven't played with it on it's own much yet, but it's obvious it's output differs from dfttest quite a bit, so I'm wondering if that makes supporting it for QTGMC a good idea. neo_fft3d is also slower than FFT3DFilter. At least on my old XP PC.

Cheers.
Except minideen and neo_Vague_Denoiser, I don't recommend use neo_ things for now, since at least neo_fft3d and neo_f3kdb both has bugs that not fixed yet, neo_dfttest may has bug too since there are output differs than dfttest

Regarding the speed, some people say neo_ is faster, IIRC msg7086 (MeteorRain) remove sse2 code and use sse as minimum so maybe this why is slower in your case
__________________
See My Avisynth Stuff
real.finder is offline   Reply With Quote
Old 26th December 2020, 08:43   #651  |  Link
hello_hello
Registered User
 
Join Date: Mar 2011
Posts: 4,823
Quote:
Originally Posted by kedautinh12 View Post
Please use this (compile by real.finder) have ncpu
https://forum.doom9.org/showpost.php...&postcount=153
I assume real.finder removed the ncpu argument from sneo_dfttest() in his latest shared functions collection (v1.143), but the fft_threads argument doesn't cause a problem with the version of neo_dfttest you linked to.

I think I'll take real.finder's advice and stick with the standard versions for the moment though.
hello_hello is offline   Reply With Quote
Old 26th December 2020, 08:58   #652  |  Link
hello_hello
Registered User
 
Join Date: Mar 2011
Posts: 4,823
real.finder,
Just a quick question if I may....

For functions/plugins that don't like YUY2, when it's stacked 16 bit, is there any reason not to simply convert to YV16 with Avisynth? This sort of thing:

ConvertToYUY2()
Dither_convert_8_to_16
ConvertToYV16()
Some_Planar_Only_Filter_That Supports_LSB()
DitherPost()
ConvertToYUY2()

It's just that I noticed your Dither Luma Rebuild function converts using the old Planar2Interleaved hack first.

Planar2Interleaved(uv==1)
Dither_YUY2toPlanar16()

or for AVisynth 2.6

Planar2Interleaved(uv==1)
ConvertToYV16()

I'm obviously missing something, but could you explain why that is for me?

Cheers.

Last edited by hello_hello; 26th December 2020 at 09:04.
hello_hello is offline   Reply With Quote
Old 26th December 2020, 16:37   #653  |  Link
real.finder
Registered User
 
Join Date: Jan 2012
Location: Mesopotamia
Posts: 2,587
Quote:
Originally Posted by hello_hello View Post
real.finder,
Just a quick question if I may....

For functions/plugins that don't like YUY2, when it's stacked 16 bit, is there any reason not to simply convert to YV16 with Avisynth? This sort of thing:

ConvertToYUY2()
Dither_convert_8_to_16
ConvertToYV16()
Some_Planar_Only_Filter_That Supports_LSB()
DitherPost()
ConvertToYUY2()

It's just that I noticed your Dither Luma Rebuild function converts using the old Planar2Interleaved hack first.

Planar2Interleaved(uv==1)
Dither_YUY2toPlanar16()

or for AVisynth 2.6

Planar2Interleaved(uv==1)
ConvertToYV16()

I'm obviously missing something, but could you explain why that is for me?

Cheers.
because Dither Luma Rebuild need Special hacked YUY2 that called Planar YUY2
__________________
See My Avisynth Stuff
real.finder is offline   Reply With Quote
Old 26th December 2020, 17:56   #654  |  Link
hello_hello
Registered User
 
Join Date: Mar 2011
Posts: 4,823
Thanks for the explanation. I might have to read it again later to make sure I absorbed it all.

Now I've looked again... I see I misread one of the functions. I thought it was doing this:

Interleaved2Planar(uv==1)
Dither_YUY2toPlanar16()

Instead of Planar2Interleaved, so now it makes more sense.

I've spent a while perusing some of your functions, mainly trying to understand more about QTGMC, and often it calls one of your shared functions, which call other functions and plugins and so on.... so trying to follow the flow can be tricky... and I'm easily confused.

Cheers.
hello_hello is offline   Reply With Quote
Old 1st January 2021, 18:22   #655  |  Link
Tempter57
Registered User
 
Join Date: Jan 2011
Location: Donetsk
Posts: 58
real.finder
Happy new year !!!
Adjustment of script YAHRmod.avsi is necessary. Additional functions EMask_dhh, DR_Radius_dhh, Camembert_dhh, GreyCenteredToMask_dhh and FastLineDarkenMOD3_dhh are necessary for function YAHR3 (). Earlier specified functions were a part of function DehaloHmod.avsi. Now in a new wording DehaloHmod.avsi the given functions miss

Last edited by Tempter57; 1st January 2021 at 18:42.
Tempter57 is online now   Reply With Quote
Old 1st January 2021, 23:19   #656  |  Link
real.finder
Registered User
 
Join Date: Jan 2012
Location: Mesopotamia
Posts: 2,587
Quote:
Originally Posted by Tempter57 View Post
real.finder
Happy new year !!!
Adjustment of script YAHRmod.avsi is necessary. Additional functions EMask_dhh, DR_Radius_dhh, Camembert_dhh, GreyCenteredToMask_dhh and FastLineDarkenMOD3_dhh are necessary for function YAHR3 (). Earlier specified functions were a part of function DehaloHmod.avsi. Now in a new wording DehaloHmod.avsi the given functions miss
Happy new year
now they are in Zs_RF_Shared.avsi and FastLineDarkMOD2.avsi, DehaloHmod.avsi still needed too
__________________
See My Avisynth Stuff
real.finder is offline   Reply With Quote
Old 1st January 2021, 23:22   #657  |  Link
real.finder
Registered User
 
Join Date: Jan 2012
Location: Mesopotamia
Posts: 2,587
Quote:
Originally Posted by LigH View Post
Seems that one calculates in Luma units and the other in Chroma units, hence half the number for YUV 4:2:x chroma subsampling... just a guess.
that true, I did update it in my git
__________________
See My Avisynth Stuff
real.finder is offline   Reply With Quote
Old 2nd January 2021, 06:37   #658  |  Link
Tempter57
Registered User
 
Join Date: Jan 2011
Location: Donetsk
Posts: 58
Quote:
Originally Posted by real.finder View Post
Happy new year
now they are in Zs_RF_Shared.avsi and FastLineDarkMOD2.avsi, DehaloHmod.avsi still needed too
No function
Code:
Function DR_Radius_dhh(clip i, int dr_rad, int count)
{
return count > dr_rad ? i : DR_Radius_dhh(i.mt_expand(), dr_rad, count+1)
}
Tempter57 is online now   Reply With Quote
Old 2nd January 2021, 07:48   #659  |  Link
real.finder
Registered User
 
Join Date: Jan 2012
Location: Mesopotamia
Posts: 2,587
do you have all scripts updated?

DR_Radius_dhh is removed now and I did made spasses() to did it job inside DeHaloH.avsi
__________________
See My Avisynth Stuff
real.finder is offline   Reply With Quote
Old 2nd January 2021, 12:40   #660  |  Link
Tempter57
Registered User
 
Join Date: Jan 2011
Location: Donetsk
Posts: 58
Quote:
Originally Posted by real.finder View Post
do you have all scripts updated?

DR_Radius_dhh is removed now and I did made spasses() to did it job inside DeHaloH.avsi
I am sorry, I have forgotten to update YAHRmod.avsi
Tempter57 is online now   Reply With Quote
Reply

Tags
avisynth, avs script, avs+, banding, generator, hbd, mod

Thread Tools Search this Thread
Search this Thread:

Advanced Search
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump


All times are GMT +1. The time now is 15:10.


Powered by vBulletin® Version 3.8.11
Copyright ©2000 - 2024, vBulletin Solutions Inc.