https://github.com/erkaman/glsl-worley
Worley noise implementation for WebGL shaders
https://github.com/erkaman/glsl-worley
cell-noise cellular-noise glsl javascript noise shader shaders voronoi webgl worley-noise
Last synced: 5 months ago
JSON representation
Worley noise implementation for WebGL shaders
- Host: GitHub
- URL: https://github.com/erkaman/glsl-worley
- Owner: Erkaman
- License: other
- Created: 2016-05-06T20:15:37.000Z (over 9 years ago)
- Default Branch: master
- Last Pushed: 2016-05-06T20:30:38.000Z (over 9 years ago)
- Last Synced: 2025-04-06T23:14:12.513Z (6 months ago)
- Topics: cell-noise, cellular-noise, glsl, javascript, noise, shader, shaders, voronoi, webgl, worley-noise
- Language: GLSL
- Homepage: http://erkaman.github.io/glsl-worley/
- Size: 369 KB
- Stars: 90
- Watchers: 3
- Forks: 7
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# glsl-worley
From this module, a GLSL implementation of Worley Noise written by [Stefan
Gustavson](http://webstaff.itn.liu.se/~stegu/GLSL-cellular/GLSL-cellular-notes.pdf) can be imported.
A demo can be found [here](http://erkaman.github.io/glsl-worley/)
## Usage
This module provides four functions, and they can be exported as
```glsl
#pragma glslify: worley3D = require(glsl-worley/worley3D.glsl)
#pragma glslify: worley2x2x2 = require(glsl-worley/worley2x2x2.glsl)
#pragma glslify: worley2D = require(glsl-worley/worley2D.glsl)
#pragma glslify: worley2x2 = require(glsl-worley/worley2x2.glsl)
```
And then they can easily be used to generate a texture in a shader by doing something like:
```glsl
vec2 F = worley3D(vPosition, 1.0, false);
float F1 = F.x;
float F2 = F.y;
gl_FragColor = vec4(vec3(F2-F1), 1.0);
```
`worley3D` is defined as `vec2 worley3D(vec3 P, float jitter, bool manhattanDistance)`. It returns a `vec2`
where `x` is `F1` and `y` is `F2`(it is assumed that the reader knows the meaning of these two).
`P` is the input point, `jitter` is the amount of jitter in the pattern, and if `manhattanDistance` is true,
then a manhattan distance is used to generate the pattern, instead of the usual Euclidean distance(this basically makes
the noise appear more "jagged").
The remaining three functions take the same arguments, except that in the case of `worley2D` and `worley2x2`, `P` is a
`vec2`.
`worley2x2x2` is basically a faster version of `worley3D`. But be aware that it has some artifacts. In the same manner,
`worley2x2` is basically a faster version of `worley2D`, but with some potential artifacts.