Skip to main content

floatColors

The floatColors shader module provides shared utilities for semantic colors in shader code. It centralizes byte-to-float normalization and common alpha helpers so applications and shader modules can reuse one consistent color contract.

Props​

useByteColors?: boolean​

When true, semantic colors passed through floatColors_normalize() are interpreted as byte-style 0..255 values and divided by 255.0 in shader code. When false, semantic colors are interpreted directly as floats and are returned unchanged.

This is intended for semantic colors only. Raw texel storage such as Uint8Array texture uploads remains byte-oriented.

Shader Functions​

floatColors_normalize(vec3 inputColor)​

Normalizes byte-style RGB colors to 0..1 when useByteColors is enabled. Float colors pass through.

floatColors_normalize(vec4 inputColor)​

Normalizes byte-style RGBA colors to 0..1 when useByteColors is enabled. Float colors pass through.

floatColors_premultiplyAlpha(vec4 inputColor)​

Returns vec4(inputColor.rgb * inputColor.a, inputColor.a).

floatColors_unpremultiplyAlpha(vec4 inputColor)​

Returns the unpremultiplied RGBA color when alpha is greater than zero, otherwise returns transparent black.

Usage​

import {floatColors, lighting} from '@luma.gl/shadertools';

const shaderInputs = new ShaderInputs({floatColors, lighting});

shaderInputs.setProps({
  floatColors: {useByteColors: false},
  lighting: {
    useByteColors: false,
    lights: [{type: 'ambient', color: [1, 0.9, 0.8], intensity: 0.2}]
  }
});
vec4 surfaceColor = floatColors_normalize(vec4(255.0, 128.0, 64.0, 255.0));
 surfaceColor = floatColors_premultiplyAlpha(surfaceColor);

Remarks​

  • floatColors is for semantic colors, not packed image data.
  • Packed vertex colors still work well with normalized vertex formats such as unorm8x4.
  • Phase 1 compatibility defaults are preserved on the APIs that adopt useByteColors.