Bridges Conference Helsinki, 2022 – Phillip C. Reiner

Aperiodic Gradient is a proof-of-concept for a continuous gradient built by culling points in a non-repetitive, non-random way. A large point cloud was generated. Machine learning (e.g. clustering by proximity) subdivides the cloud into sub-clouds of roughly equal size. A curve function is mapped as an intensity attribute to each sub-cloud. Each point's distance to an attractor is remapped onto the intensity range of its sub-cloud; a threshold on that value decides whether the point is culled. The result is a gradient that avoids both strict periodicity and randomness. The output was printed for the 2022 Bridges Conference.

Research:
Point clouds were subdivided by ML clustering on relative proximities. Intensity was assigned per sub-cloud from a chosen curve; per-point intensity was then set by attractor distance and remapping. The culling threshold produced the final pattern. The pipeline was tested for gradient continuity and visual coherence across different curve and attractor choices.