Below is a series of generative works created with simple algorithms built around vectors and evolving grid systems. Each piece develops both functionally and aesthetically through these rules. For the best viewing experience, please set all videos to HD in the YouTube video settings.
The direction of each branch is controlled entirely by a Perlin noise flow field. This
creates smooth and coordinated motion that mimics smoke or fluid. The branching remains
randomized, but the flow field causes the branches to drift outward in unison and then peel
apart when the field curves. The trails left behind form a secondary image within the piece. This
piece is inspired by Ben Laposky’s Oscillon series from 1952.
I begin with 100 branches at the center, each with a lifespan of 400
frames and a Perlin noise influence. Branches near the outer edges die sooner than those near the center. New branches constantly emerge, so the piece is always in motion.
Again, I begin with 100 branches at the center, each with a lifespan of 400
frames. Branches near the outer edges die sooner than those near the center. New branches constantly emerge, so the piece is always in motion. This piece exhibits subtler Perlin noise influence than the previous.
This work is inspired by François Morellet’s 4 Grids 0° 22.5° -45° -67.5° (1958). In that early generative piece, Morellet plotted four grids at four different rotations,creating ornamental polygonal shapes that almost become circular in certain areas. I recreated his setup by generating four grids at the same angles, then pushed the idea forward by having them rotate – two clockwise, two counterclockwise – around the center. As they move, the overlapping lines generate shifting polygons and a pulsing of plaid-like squares that change size.
In the next iteration, I kept the same set of grids but added oscillation to the spacing between the horizontal and vertical lines. The plaid-like pulsing from the previous example is still there, but now it seems to rotate and jump around the composition, almost like shifting from one spot to another on a map. When the spacing gets very small, it resembles TV static and the squares
appear to spin.
Here, real-time branching emerges from the center. The system grows as new branches are continually added to an array around a central star. Multiple branches are drawn per frame to keep the growth smooth. Branch length, number of children, and slight jitter are all randomized to make the structure feel more organic and less symmetrical. This piece resembles Casey Reas’ Process 6 (Image 4) from 2005.
