StackGenVis: Alignment of Data, Algorithms, and Models for Stacking Ensemble Learning Using Performance Metrics https://doi.org/10.1109/TVCG.2020.3030352
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
StackGenVis/frontend/node_modules/glsl-specular-cook-torrance/README.md

67 lines
1.9 KiB

4 years ago
# glsl-specular-cook-torrance
Computes the specular power in the Cook-Torrance model.
# Example
```glsl
#pragma glslify: cookTorranceSpec = require(glsl-specular-cook-torrance)
uniform vec3 eyePosition;
uniform vec3 lightPosition;
uniform float roughness, fresnel;
varying vec3 surfacePosition, surfaceNormal;
void main() {
//Light and view geometry
vec3 viewDirection = normalize(eyePosition - surfacePosition);
vec3 lightDirection = normalize(lightPosition - surfacePosition);
//Surface properties
vec3 normal = normalize(surfaceNormal);
//Compute specular power
float power = cookTorranceSpec(
lightDirection,
viewDirection,
normal,
roughness,
fresnel);
gl_FragColor = vec4(power,power,power,1.0);
}
```
# Usage
Install with npm:
```
npm install glsl-specular-phong
```
Then use with [glslify](https://github.com/stackgl/glslify).
# API
```glsl
#pragma glslify: cookTorrance = require(glsl-specular-cook-torrance)
```
##### `float cookTorrance(vec3 lightDir, vec3 eyeDir, vec3 normal, float roughness, float fresnel)`
Computes the specular power in the Cook-Torrance
* `lightDir` is a *unit length* `vec3` pointing from the surface point toward the light
* `eyeDir` is a *unit length* `vec3` pointing from the surface point toward the camera
* `normal` is the *unit length* surface normal at the sample point
* `roughness` is the surface roughness parameter, between 0 and 1. 0 means surface is perfectly smooth, 1 means surface is matte
* `fresnel` the Fresnel exponent. 0 = no Fresnel, higher values create a rim effect around objects
**Returns** A `float` representing the specular power
**Note** Unlike the usual Cook-Torrance model, the light power is not scaled by 1/(normal . lightDirection). This avoids an unnecessary numerically unstable division, but requires modifying how diffuse light is calculated.
# License
(c) 2014 Mikola Lysenko. MIT License