crystallize effect

SourceEffect01/ ORIGIN
The structure is a Voronoi diagram, named for Georgy Voronoy who formalised it in 1908, though the idea is older and Descartes sketched something like it in 1644. Assign a set of seed points, then give every location to whichever seed is closest. It describes crystal growth, cell packing, mobile phone coverage and how a forest divides light, which is why the result reads as organic rather than designed despite being pure arithmetic.
02/ SWEEP
FACET SIZE sweep



03/ SPEC
This effect partitions the image into irregular polygonal cells and fills each with a single averaged colour, which is Voronoi tessellation rather than a regular grid. FACET SIZE runs from 1 to 12 pixels with a default of 2 and controls seed spacing, so it sets how many facets the image is broken into: small values keep detail across many cells, large values reduce a portrait to a few dozen shards. Because cell boundaries are irregular, edges do not align to a grid and the result reads as fracture rather than as pixelation. In kott it renders in natural colour.
it can move to the music
Every control of this effect can follow a track. Pick a signal in the studio’s audio panel (the beat, the downbeat, a build, the bass, or an EQ band you set yourself), point it at facet size or any other control here, and set how far it moves. The sound comes from your microphone, a browser tab or a file you load, and it is analysed on your own machine.
how the audio routing works04/ FAQ
- What is a crystallize effect?
- A crystallize effect breaks an image into irregular polygonal cells of flat colour, using a Voronoi partition where every point belongs to its nearest seed rather than to a grid square.
- What is the difference between crystallize and pixelate?
- Pixelation uses a regular grid, so every cell is the same square. Crystallize uses irregular polygons of varying size and shape, which is why it reads as fracture rather than as low resolution.
- What FACET SIZE keeps a face recognisable?
- Stay at 1 or 2. Because facets are irregular, a single large cell can swallow an eye entirely, so coarse settings lose likeness faster than an equivalent pixelation would.