Convolution3D

Latest revision as of 09:57, 13 March 2021

[edit] Description

Convolution3D is an AviSynth filter that will apply a 3D - temporal and/or spatial - convolution to a frame

[edit] Requirements

• [x86] AviSynth+ or AviSynth 2.6
• [x64] AviSynth+
• Supported color formats: Y8, YV12, YV16, YV24, YV411

[edit] Syntax and Parameters

Convolution3D (clip, int "matrix", int "ythresh", int "cthresh", int "t_ythresh", int "t_cthresh", float "influence", int "debug")

clip   =

Input clip.

int  matrix = 0

Matrix:

• 0 : Original matrix; useful for normal video (not anime) because it keep more details.

• 1 : Full 1 matrix; useful with anime or bad quality sources because it blurs a little more (so it removes more noise).

int  ythresh = 3

int  cthresh = 4

Spatial treshold for luma and chroma.

int  t_ythresh = 3

int  t_cthresh = 4

Temporal treshold for luma and chroma.

float  influence = 3.0

This parameter is used especially to speed up a little this filter and to avoid using temporal information when not needed (scene change, fade, ...).

First a limit is built = Temporal Luma Threshold * Temporal influence

For each 2 pixel computed, this is checked:

if
  (Abs (Y0 - Y0[Previous frame]) +
  Abs (Y0 - Y0[Next frame]) +
  Abs (Y1 - Y1[Previous frame]) +
  Abs (Y1 - Y1[Next frame])) > limit
then
  do Spacial work (only 3*3 matrix)
Else
  do Spacial and Temporal work (3*3*3 matrix)

The lower it is -> the faster the filter will be but compressibility should be lower

The higher it is -> the slower the filter will be but compressibility should be higher

If temporal influence is set to -1 then only spatial work is done (high speed).

Written by Wilbert.

int  debug = 0

Not documented.

[edit] Examples

Convolution3D with all default values:

AviSource("Blah.avi")
Convolution3D (matrix=0, ythresh=3, cthresh=4, t_ythresh=3, t_cthresh=4, influence=3.0)

[edit] Changelog

Version      Date            Changes

v1.1         2021/03/11      - Migrate from VC6 to Visual Studio 2019, v142 toolset
                             - Drop all (buggy beta) inline SSE assembler
                             - Rewrite routines in SSE2 SIMD
                             - Turn into Avisynth 2.6 interface
                             - Preserve frame properties when Avisynth+ v8 interface
                             - Add YV16, YV24 and YV411 format support besides YV12
                             - known bug: left and right side calculation can theoretically cause crash

[edit] Archived Downloads

[edit] External Links

• avisynth.nl - original documentation.
• Doom9 Forum - original 2003 discussion thread on Doom9.
• Doom9 Forum - original 2002 discussion thread on Doom9.

Back to External Filters ←