🔲 Worley Noise
Cellular texture
Mode: F1
Output
Parameters
Controls
Stats
Points
0
Grid
0
Frame ms
0
Status
Ready
Info & Theory

Worley noise (cellular noise), introduced by Steven Worley in 1996, scatters feature points across the plane and colours each pixel by its distance to the nearest ones.

F1, F2 and combinations

  • F1 — distance to the nearest point: smooth cells.
  • F2 — distance to the second nearest.
  • F2 − F1 — thin ridges along cell borders: cracks.
  • F1 · F2 — a softer organic blend.

Distance metric

Euclidean gives round cells, Manhattan gives diamonds and Chebyshev gives squares. The metric reshapes the whole field at once.

Spatial grid (fast)

Points are bucketed into a grid. Each pixel only checks its own cell and the 8 neighbours, so the cost is roughly O(pixels) instead of O(pixels × points).

Voronoi link

Each point owns the pixels closest to it — a Voronoi cell. So F2 − F1 simply draws the Voronoi edges.

Frequently asked questions

What is Worley noise?

Worley noise, also called cellular noise, scatters feature points across the plane and colours each pixel by its distance to the nearest points. Steven Worley introduced it in 1996 for procedural stone, water, cell and crack textures.

What do F1 and F2 mean?

F1 is the distance from a pixel to its nearest feature point, F2 to the second nearest. Using F1 gives smooth cells; F2−F1 highlights the borders between cells as cracks; F1·F2 and other combinations give different organic looks.

How is this related to a Voronoi diagram?

Each feature point owns the region of pixels closest to it — that is exactly a Voronoi cell. The F2−F1 mode draws the Voronoi edges, so Worley noise is essentially a smooth, shaded Voronoi diagram.

What does the distance metric change?

Euclidean distance gives round, organic cells. Manhattan (taxicab) distance gives diamond-shaped cells aligned to the axes. Chebyshev distance gives square cells. Switching the metric reshapes every cell at once.

Why use a spatial grid?

Checking every pixel against every point would be very slow. Instead the feature points are bucketed into grid cells, and each pixel only checks the points in its own cell and the eight neighbours — making the cost roughly constant per pixel.

What does point density control?

Density sets how many feature points fall in each grid cell on average. Higher density means smaller, more numerous cells; lower density means large open cells with long cracks.

What is the animation doing?

When animation is on, each feature point drifts slowly around its home position, so the cells gently wobble and flow. The pattern is fully recomputed each frame from the moved points.

What are the palettes for?

The palette maps the distance value at each pixel to a colour. Grayscale shows the raw field; the stone, ember and ice palettes tint it to suggest rock, lava or frozen textures.

Is the texture seamless or tileable?

This demo renders a single window of the noise field and is not made tileable. Production Worley noise can be wrapped at the edges so it tiles, but that is left out here for clarity.

Where is Worley noise used in practice?

It is widely used in computer graphics for procedural textures, cloud and water shaders, caustics, cracked surfaces, and as a building block layered with Perlin noise in game and film rendering.