IT lexicon › Programming › Worley noise

Worley noise

Programming På svenska → Updated: 2026-10-24

Cellular noise, introduced by Steven Worley in 1996, that colours each point by its distance to the nearest scattered feature points — the look of cells, cracks and scales.

You scatter feature points (for speed, often one per grid cell) and, at each position, compute the distance to the nearest (F1), the second-nearest (F2) and so on. F1 alone gives bubbly cells; F2−F1 draws out the cracks between them — in effect the edges of a Voronoi diagram. The choice of distance metric, Euclidean, Manhattan or Chebyshev, changes the shape of the cells.

Worley described it in a 1996 paper, and it is also called cellular or Voronoi noise. It is used for stone, reptile scales, water caustics and for partitioning a world into biomes. Unlike Perlin noise and simplex noise, which are smooth waves, Worley noise builds on discrete proximity and so yields hard edges almost for free. It works nicely as a height source for heightmaps.

← Back to the lexicon