Selur's Little Message Board

Full Version: Deoldify Vapoursynth filter
You're currently viewing a stripped down version of our content. View the full version with proper formatting.
Hi NASS,

  Can you provide a small sample and the script used.

Dan

(3 hours ago)Selur Wrote: [ -> ]updated Hybrid_havc_test, which should update the tool-tips.

Cu Selur

Now, it's Okay.

It is missing only the too-tip for "Retry (threshold/model)" you can add:  
"If Retry (threshold) > 0, this value will be used to detect frames with insufficient permanent-memory coverage and in this case the frame with missing good references will be colorized on the fly, suggested value: 0 (disabled) or 0.20-0.35 . The colorization models can be:  
HAVC (Deoldify + DDColor)  #  fast, but the quality is not good if compared to DiT models 
DiT fp4/int4  (depending on GPU used) # very good quality but can slow-down the inference speed by about 40%-50%".

Thanks,
Dan
Hi Dan. Thanks so much for your reply. Here's the script in dev version ( Motion Distortion + Color jump ):
# Imports
import sys
import os
import vapoursynth as vs
# getting Vapoursynth core
core = vs.core
# Import scripts folder
scriptPath = 'C:/Program Files/Hybrid/64bit/vsscripts'
sys.path.insert(0, os.path.abspath(scriptPath))
# loading plugins
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/MiscFilter/MiscFilters/MiscFilters.dll")
core.std.LoadPlugin(path="C:/Program Files/Hybrid/64bit/vsfilters/SourceFilter/LSmashSource/LSMASHSource.dll")
# Import scripts
import vsdeoldify as havc
import validate
# Source: 'C:\Users\NASS\Desktop\outestny_restoredcolor.mp4'
# clip current meta; color space: YUV420P8, bit depth: 8, resolution: 960x720, fps: 24, color matrix: 709, color primaries: Unspecific, color transfer: BT.709, yuv luminance scale: limited, scanorder: progressive, full height: true (Source)
# Loading 'C:\Users\NASS\Desktop\outestny_restoredcolor.mp4Ä using LWLibavSource
clip = core.lsmas.LWLibavSource(source="C:/Users/NASS/Desktop/outestny_restoredcolor.mp4", format="YUV420P8", stream_index=0, cache=0, prefer_hw=0)
frame = clip.get_frame(0)
# setting color matrix to 709.
clip = core.std.SetFrameProps(clip, _Matrix=vs.MATRIX_BT709)
# setting color transfer (vs.TRANSFER_BT709), if it is not set.
if validate.transferIsInvalid(clip):
clip = core.std.SetFrameProps(clip=clip, _Transfer=vs.TRANSFER_BT709)
# setting color primaries info (to vs.PRIMARIES_BT709), if it is not set.
if validate.primariesIsInvalid(clip):
clip = core.std.SetFrameProps(clip=clip, _Primaries=vs.PRIMARIES_BT709)
# setting color range to TV (limited) range.
prop_name = '_Range' if core.core_version.release_major >= 74 else '_ColorRange'
clip = core.std.SetFrameProps(clip=clip, **{prop_name: vs.RANGE_LIMITED})
# making sure frame rate is set to 24fps
clip = core.std.AssumeFPS(clip=clip, fpsnum=24, fpsden=1)
# making sure the detected scan type is set (detected: progressive)
clip = core.std.SetFrameProps(clip=clip, _FieldBased=vs.FIELD_PROGRESSIVE) # scan type: progressive
# adjusting color space from YUV420P8 to RGB24 for vsColorAdjust
clip = core.resize.Bicubic(clip=clip, format=vs.RGB24, matrix_in_s="709", range_in_s="limited", range_s="full")
# stabilize colors using ColorAdjust (HAVC)
clip = havc.HAVC_ColorAdjust(clip, BlackWhiteTune="light", BlackWhiteMode=6)
# adjusting output color from RGB24 to YUV420P8 for x264Model
clip = core.resize.Bicubic(clip=clip, format=vs.YUV420P8, matrix_s="709", range_in_s="full", range_s="limited")
# set output frame rate to 24fps (progressive)
clip = core.std.AssumeFPS(clip=clip, fpsnum=24, fpsden=1)
# output
clip.set_output()
# clip current meta; color space: YUV420P8, bit depth: 8, resolution: 960x720, fps: 24, color matrix: 709, color primaries: Unspecific, color transfer: BT.709, yuv luminance scale: limited, scanorder: progressive, full height: true (Meta)
# script was created by Hybrid 2026.05.22.1

It’s a shame because this is a really great feature, but Motion Distortion Issue + Color jump affects the final result. We definitely need to find a solution!

Thanks so much Dan

NASS

In the original version (Hybrid_2026.03.21.1_SETUP) and VapoursynthR73_torch_2026.02.26 with the old colormnet, there is no color-jumping issue, but the major distortion problem remains.

# Imports
import vapoursynth as vs
# getting Vapoursynth core
import sys
import os
core = vs.core
# Import scripts folder
scriptPath = 'C:/Program Files/Hybrid2/64bit/vsscripts'
sys.path.insert(0, os.path.abspath(scriptPath))
# loading plugins
core.std.LoadPlugin(path="C:/Program Files/Hybrid2/64bit/vsfilters/MiscFilter/MiscFilters/MiscFilters.dll")
core.std.LoadPlugin(path="C:/Program Files/Hybrid2/64bit/vsfilters/SourceFilter/LSmashSource/LSMASHSource.dll")
# Import scripts
import vsdeoldify as havc
import validate
# Source: 'C:\Users\NASS\Desktop\LA\restoredcolor.mp4'
# Current color space: YUV420P8, bit depth: 8, resolution: 938x698, frame rate: 24fps, scanorder: progressive, yuv luminance scale: limited, matrix: 709, transfer: bt.709, primaries: bt.709, format: AVC
# Loading ' C:\Users\NASS\Desktop\LA\restoredcolor.mp4Ä using LWLibavSource
clip = core.lsmas.LWLibavSource(source="C:\Users\NASS\Desktop\LA\restoredcolor.mp4 ", format="YUV420P8", stream_index=0, cache=0, prefer_hw=0)
frame = clip.get_frame(0)
# setting color matrix to 709.
clip = core.std.SetFrameProps(clip, _Matrix=vs.MATRIX_BT709)
# setting color transfer (vs.TRANSFER_BT709), if it is not set.
if validate.transferIsInvalid(clip):
clip = core.std.SetFrameProps(clip=clip, _Transfer=vs.TRANSFER_BT709)
# setting color primaries info (to vs.PRIMARIES_BT709), if it is not set.
if validate.primariesIsInvalid(clip):
clip = core.std.SetFrameProps(clip=clip, _Primaries=vs.PRIMARIES_BT709)
# setting color range to TV (limited) range.
clip = core.std.SetFrameProps(clip=clip, _ColorRange=vs.RANGE_LIMITED)
# making sure frame rate is set to 24fps
clip = core.std.AssumeFPS(clip=clip, fpsnum=24, fpsden=1)
# making sure the detected scan type is set (detected: progressive)
clip = core.std.SetFrameProps(clip=clip, _FieldBased=vs.FIELD_PROGRESSIVE) # scan type: progressive
# adjusting color space from YUV420P8 to RGB24 for vsColorAdjust
clip = core.resize.Bicubic(clip=clip, format=vs.RGB24, matrix_in_s="709", range_in_s="limited", range_s="full")
# stabilize colors using ColorAdjust (HAVC)
clip = havc.HAVC_ColorAdjust(clip, BlackWhiteTune="strong", BlackWhiteMode=6, Strength=1)
# adjusting output color from RGB24 to YUV420P8 for x264Model
clip = core.resize.Bicubic(clip=clip, format=vs.YUV420P8, matrix_s="709", range_in_s="full", range_s="limited") # additional resize to match target color sampling
# set output frame rate to 24fps (progressive)
clip = core.std.AssumeFPS(clip=clip, fpsnum=24, fpsden=1)
# output
clip.set_output()
# script was created by Hybrid 2026.03.21.1

Here are a few photos of the result with the distortion problem

https://drive.google.com/drive/folders/1...KwSv1D5ifD


Best

NASS
(2 hours ago)Dan64 Wrote: [ -> ]Hi NASS,

  Can you provide a small sample and the script used.

Dan


Hi! Can you provide a step-by-step workflow for using DITserver-based colorization? Also, can DITserver colorize new characters that appear in the same shot without altering the colors of the already colorized areas?