Selur's Little Message Board

Hi guys,

I am new here.

Two Things i wanna ask.

1)How to fix jaggy lines of anime clip attached herewith.

2) I cant able to find scunet and basic vsr++ in hybrid installed in vs filters section. why?

Regarding 1: those are dow crawlers, try DeCrawl->LUTDECrawl
With default settings:

try tweaking the settings.

Regarding 2: BasicVSR++ requires you to install the torch-addon. SCUNet is available through the torch- and the vs-mlrt-addon. => you need to install the addons to use these. (both addons are way to large to ship them directly with Hybrid)

Cu Selur

Side note: the file extension on that file wrong. It's not a .mp4, but a .mkv file, feeding Hybrid mislabed files might cause problems.

Sorry my mistake. Changed the extension to. mp4 because i couln't able to upload here. Forgot to change back before adding to rar.

---

Thanks.
LUTDECrawl helps.
Any other suggestions to clean the video and to stop video shaking.
Can you guide me how to install addon. Is there any separate thread?

Selur sir please help me.

How to fix unstable lines and color bleeding marked in below image.

https://ibb.co/VWv9bBMJ

Clip also added.

Quote: Can you guide me how to install addon. Is there any separate thread?

Read the README.md in the GoogleDrive share.

For both clips:
To reduce shaking, use Stab with range=3 and crop after that to get rid of black bars, Stab added to compensate for the movement.
(2 pixels at top and left worked fine here)
There is also some distortion in the lower quarter of the image sometimes, which you probably won't get rid of.
I would also add sRestore with Frate=23.976 (and Omode=6) to get rid of some of the ghostings.

Santiag can help with the aliasing.
aWarpSharp with Chroma on chroma planes only should help with the chroma bleed. Alternatively 1x_BleedOut_Compact_300k_net_g either through VSGAN or vsMLRT for the color bleed.
Depending on your liking, you can also add BasicVSR++ (maybe just masked to filter the edges)
I would also add DeHalo_Alpha to get rid of most of the white halos.
Note that is you filtering this way is slow.
Script: https://pastebin.com/2reuBbz4
Files: https://www.mediafire.com/folder/nq8yx4m...lltogether
Additionally, using something like SMDeGrain or TemporalDegrain with DCT=1 might additionally help to stabilize the image.

Cu Selur

Thanks a lot.

But BasicVSR++ showing error

Failed to evaluate the script:
Python exception: basicvsrpp: model file has not been downloaded. run `python -m vsbasicvsrpp` to download all models, or set `auto_download=True` to only download the specified model
Traceback (most recent call last):
File "src\\cython\\vapoursynth.pyx", line 3387, in vapoursynth._vpy_evaluate
File "src\\cython\\vapoursynth.pyx", line 3388, in vapoursynth._vpy_evaluate
File "C:\Users\green\AppData\Local\Temp\tempPreviewVapoursynthFile19_53_33_125.vpy", line 44, in
clip = BasicVSRPP(clip=clip, model=4)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Program Files\Hybrid\64bit\Vapoursynth\Lib\site-packages\torch\utils\_contextlib.py", line 116, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "C:\Program Files\Hybrid\64bit\Vapoursynth\Lib\site-packages\vsbasicvsrpp\__init__.py", line 148, in basicvsrpp
raise vs.Error(
vapoursynth.Error: basicvsrpp: model file has not been downloaded. run `python -m vsbasicvsrpp` to download all models, or set `auto_download=True` to only download the specified model

Fixed that yesterday iirc. either redownload the torch addon or open an admin terminal inside the Vapoursynth folder and call:
to fix it.

Cu Selur

Error coming even after reinstalling hybrid and replacing VS folder.

Failed to evaluate the script:
Python exception: basicvsrpp: model file has not been downloaded. run `python -m vsbasicvsrpp` to download all models, or set `auto_download=True` to only download the specified model

Traceback (most recent call last):
File "src\\cython\\vapoursynth.pyx", line 3387, in vapoursynth._vpy_evaluate
File "src\\cython\\vapoursynth.pyx", line 3388, in vapoursynth._vpy_evaluate
File "C:\Users\green\AppData\Local\Temp\tempPreviewVapoursynthFile23_02_18_623.vpy", line 38, in
File "C:\Program Files\Hybrid\64bit\Vapoursynth\Lib\site-packages\torch\utils\_contextlib.py", line 116, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "C:\Program Files\Hybrid\64bit\Vapoursynth\Lib\site-packages\vsbasicvsrpp\__init__.py", line 148, in basicvsrpp
raise vs.Error(
vapoursynth.Error: basicvsrpp: model file has not been downloaded. run `python -m vsbasicvsrpp` to download all models, or set `auto_download=True` to only download the specified model
2025-01-31 23:04:22.100
[VSE Server]: incoming connection
[VSE Server]: ConnectedState
[VSE Server]: socket is ready to be read
[VSE Server]: connection open: true
[VSE Server]: connection readable: true
[VSE Server] - Message received: changeTo ### C:\Users\green\AppData\Local\Temp\tempPreviewVapoursynthFile23_02_29_774.vpy ###
2025-01-31 23:04:22.365
Failed to evaluate the script:
Python exception: basicvsrpp: model file has not been downloaded. run `python -m vsbasicvsrpp` to download all models, or set `auto_download=True` to only download the specified model
Traceback (most recent call last):
File "src\\cython\\vapoursynth.pyx", line 3387, in vapoursynth._vpy_evaluate
File "src\\cython\\vapoursynth.pyx", line 3388, in vapoursynth._vpy_evaluate
File "C:\Users\green\AppData\Local\Temp\tempPreviewVapoursynthFile23_02_29_774.vpy", line 38, in
clip = BasicVSRPP(clip=clip, model=7)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "C:\Program Files\Hybrid\64bit\Vapoursynth\Lib\site-packages\torch\utils\_contextlib.py", line 116, in decorate_context
return func(*args, **kwargs)
^^^^^^^^^^^^^^^^^^^^^
File "C:\Program Files\Hybrid\64bit\Vapoursynth\Lib\site-packages\vsbasicvsrpp\__init__.py", line 148, in basicvsrpp
raise vs.Error(
vapoursynth.Error: basicvsrpp: model file has not been downloaded. run `python -m vsbasicvsrpp` to download all models, or set `auto_download=True` to only download the specified model

You did use: VapoursynthR70_torch_2025.01.30.7z right? (that is the updated torch where this should be fixed,..)