The first question was why one grid was wrong.

One Rule 110 grid, dying cells only — I built that. It works. 003 B-side candidates came from it. But a single grid has one rate of change, one density pattern, one kind of silence. The dying cells are evenly distributed across the runtime. You hear the method too clearly.

Two grids changes what you hear. Not because it adds complexity — the method is the same, each grid produces dying cell events as tones — but because the two patterns are never aligned. Different seeds produce different geometries. Different rates produce different rhythms. The piece becomes the interference between them, not the method itself.