Thank you. The issue with using 64 colors is the way I divided the byte-like genes in the procedural generation into multiple color levels. I think I used -2, -1, neutral, 1, 2 color levels for each color, so that leaves only 256 / 5 = 51 colors to work with.