Posts: 10.553
Threads: 57
Joined: May 2017
Quote:Line -> Resizer -> NNEDI3 GPU resize WORKS, but prescreening: level 0,1,2 no more there. It only have option "new' and "old". Is this a bug?
No.See: https://github.com/HomeOfVapourSynthEvol.../README.md
Quote:Yes, i attempt to open a folder with LUTs as well. I can press button "Open" when folder is selected. But after that folder didn't opened and filter keep show same path /Library/Frameworks/VapourSynth.framework/lib/vapoursynth/ColorFilter/TimeCube
will try to look at it tomorrow
Cu Selur
----
Dev versions are in the 'experimental'-folder of my GoogleDrive, which is linked on the download page.
Posts: 785
Threads: 16
Joined: Mar 2020
Ok, great! Seems it was really nice update with a lot of critical and useful bugfixes...
I will test other plug-ins today and will add reports so you may look tomorrow or somewhere in future.
Thanks for your great work and support!
Posts: 785
Threads: 16
Joined: Mar 2020
I always wonder why tabs in "Config" tab made like this?
- Those tabs don't match to other Hybrid UI tabs look and logic.
- They are too small.
- Active tab is not highlighted and hard to understand location (there is a text hint at far right, but visually it don't works well)
- "Output" submenu is even more confusing and hard to navigate.
- Most often used "Output -> Container Setting" hidden deeply in submenu.
Sure, "Config" tab is not used too often, and it is not critical to change things there. But if possible, just for improved usability i really prefer if it was replaced with single row of normal unified tabs.
---
Here are also few more ideas about tabs arrangement (for discussion only because i am not 100% sure about this):
1. What if move "Config" tab to the far right? (it may have same separation logic as current position of "Misc" tab in far right)
... | Filtering | Chapter | Jobs | Log | Config
Cons: From some other point of view "Jobs" and "Log" tabs feels like another totally different group separated from Filters and Config, so no need to change current positions.
2. What if flip "Sharpen" and "De-Halo" tabs placements? Seems like logical "de-something" group of tabs, next sharpen and other tabs...
| Color | Deblocker | DeNoise | DeGrain | DeHalo/-Ring | Sharpen | Artefacts | Line | Other | Custom | Misc |
Cons: "Sharpen" filters feels more like belong to "DeNoise/DeGrain" group, and "De-Halo" is logically closer to artifact remove group, so no need to change current positions.
Posts: 785
Threads: 16
Joined: Mar 2020
Color -> ColorMatrix -> From/To: Rec.2020 ERROR
When i select transform From: Rec.2020 or To: Rec.2020 VS Preview window opens as empty window.
Also looking to colors, seems rendered file seems don't transformed to Rec2020 color space as well.
Quote:"This is strange because both libsangnom.dylib and libsangnommod.dylib are installed. AntiAliasing -> MAA and AAF give same error."
Both work fine here, so it's something with the dylibs you use.
I try SangNom-r40 and SangNom-r41 from l33tmeatwad's folder https://www.mediafire.com/folder/wvdlnja...k2cr8w4qdd and both versions give me same error with MAA and AAF filters "AttributeError: There is no function named SangNom."
Maybe you use some newer custom compiled version?
All other AA filters (including Nedi3AA) seems works well now with GPU OpenCL.
Hey Selur,
Meatwad mentioned that getting ffmbc compiled for mac may not be worth the trouble. He said it was meant to support DNxHD and Prores better which is already in FFMpeg.
Posts: 785
Threads: 16
Joined: Mar 2020
AntiAliasing -> Santiag WORKS if i choose "Type: nnedi3 or eedi2 or eedi3" inside it
but gives me ERROR if i choose "Type: Sangnom" inside it. (same AttributeError: There is no function named SangNom as MAA and AAF filters)
Failed to evaluate the script:
Python exception: There is no function named SangNom
Traceback (most recent call last):
File "src/cython/vapoursynth.pyx", line 2244, in vapoursynth.vpy_evaluateScript
File "src/cython/vapoursynth.pyx", line 2245, in vapoursynth.vpy_evaluateScript
File "/Volumes/temp/Hybrid Temp/tempPreviewVapoursynthFile10_06_05_407.vpy", line 33, in <module>
clip = havsfunc.santiag(c=clip, type="sangnom", opencl=True, device=1)
File "/Library/Frameworks/VapourSynth.framework/lib/python3.8/site-packages/havsfunc.py", line 218, in santiag
c = santiag_dir(c, strh, typeh, fwh, fhh)
File "/Library/Frameworks/VapourSynth.framework/lib/python3.8/site-packages/havsfunc.py", line 150, in santiag_dir
c = santiag_stronger(c, strength, type)
File "/Library/Frameworks/VapourSynth.framework/lib/python3.8/site-packages/havsfunc.py", line 172, in santiag_stronger
c = santiag_stronger(c, strength - 1, type)
File "/Library/Frameworks/VapourSynth.framework/lib/python3.8/site-packages/havsfunc.py", line 195, in santiag_stronger
return c.sangnom.SangNom(order=field, aa=aa)
File "src/cython/vapoursynth.pyx", line 1934, in vapoursynth.Plugin.__getattr__
AttributeError: There is no function named SangNom
Also ERROR if i choose "Type: selectHV" inside it
Failed to evaluate the script:
Python exception: santiag() got an unexpected keyword argument 'typevh'
Traceback (most recent call last):
File "src/cython/vapoursynth.pyx", line 2244, in vapoursynth.vpy_evaluateScript
File "src/cython/vapoursynth.pyx", line 2245, in vapoursynth.vpy_evaluateScript
File "/Volumes/temp/Hybrid Temp/tempPreviewVapoursynthFile10_12_26_439.vpy", line 33, in <module>
clip = havsfunc.santiag(c=clip, typeh="sangnom", typevh="sangnom", opencl=True, device=1)
TypeError: santiag() got an unexpected keyword argument 'typevh'
Posts: 785
Threads: 16
Joined: Mar 2020
Filtering -> Other -> VFM ERROR:
From my understanding VFM is some sort of special detelecine filter? So maybe it give me error because require some specific input to work? I test it with different sources including telecined animation examples (try to switch deinterlace on and off) and it still give me error. When i launch filter it also give this pop-up window message:
"Got an empty value for 'vsOVFMMChroma', please report to Selur!"
Failed to evaluate the script:
Python exception: VFM: Invalid mi threshold specified
Traceback (most recent call last):
File "src/cython/vapoursynth.pyx", line 2244, in vapoursynth.vpy_evaluateScript
File "src/cython/vapoursynth.pyx", line 2245, in vapoursynth.vpy_evaluateScript
File "/Volumes/temp/Hybrid Temp/tempPreviewVapoursynthFile04_55_28_928.vpy", line 18, in <module>
clip = core.vivtc.VFM(clip=clip, order=0, blockx=4, blocky=4)
File "src/cython/vapoursynth.pyx", line 2069, in vapoursynth.Function.__call__
vapoursynth.Error: VFM: Invalid mi threshold specified
Posts: 785
Threads: 16
Joined: Mar 2020
Small not too critical visual problem in filters presets list: "AntiAliasing - Nedi3aa" located before global "AntiAliasing"
Seems just name mistake An itAliasing
Posts: 785
Threads: 16
Joined: Mar 2020
UPDATE. seems chroma edges artifacts where caused by MPEG1 source. Seems Hybrid reads MPEG1 in some strange way. When i convert source file to 10bit ProRes422HQ and import again in hybrid, chroma edges have no artifacts when i apply 601 to 709 color matrix. But anyway other problems with color matrix still exists.
Color -> ColorMatrix PROBLEM
It seems this important filter may need some serious quality control or total redesign. In addition to broken Rec.2020 seems it have global problems with quality.
Here is example of VCD MPEG1 4:2:0 352x288 upscaled to 1440x1080.
As you can see ColorMatrix transform produce strange ugly artifacts with chroma edges.
I try to experiment with Filtering-> Misc -> Color Resizer Methods and with Crop/Resize Resize Methods but it don't help a lot.
Here is comparison with quality color managed video player output on macOS (on the right):
DeHalo YAHR ERROR
When VS preview already open and i turn on DeHalo YAHR, i got Pop-up message:
"Got an empty value for 'vsYAHRBlur', please report to Selur!"
Sometimes problem can't be reproduced...
Posts: 785
Threads: 16
Joined: Mar 2020
Also in ColorMatrix SMPTE 240M may be broken as well.
In all sources it is described that SMPTE 240M should look similar to BT.709 https://forum.doom9.org/showthread.php?t=133982 https://avisynth.org.ru/docs/english/ext...matrix.htm
But in Hybrid ColorMatrix SMPTE 240M looks same as Rec.601
|