Think about how a normal camera works. When you snap a picture, the final file isn't magic—it's just a specific arrangement of colored pixels laid out on a flat grid.
If you take a camera, lock down the exact number of pixels it can record, and set a hard limit on the specific shades of colors those pixels can show, you get a finite sandbox. By running a loop that forces those pixels to go through every single math combination possible, you eventually output every single image that specific camera could ever capture.
A regular camera handles millions of pixels. To make this concept testable, we scale the canvas down to fixed dimensions (like a 3x3 grid or a 100x100 matrix). This locks the structural boundaries of our sandbox.
Instead of using billions of subtle color gradients, we limit the precision—for example, mapping out exactly 7 or 8 shades per channel. This keeps the total number of permutations trackable while maintaining clear visual data.
Because we fixed the grid size and the color palette, the engine can systematically cycle through every layout state. Consequently, any photo you could ever take on this specific camera setup is guaranteed to show up eventually.
This project was created to turn these abstract data patterns into real, running engines. The simulations allow you to watch the generation happen frame by frame, test out the absolute baseline structures with a 3x3 loop, run larger matrices, or even upload your own custom files to see exactly where they sit inside the grid sequence.