♟
How it works.
The Checker Shadow illusion, created by MIT professor Edward Adelson in 1995, is arguably the most compelling demonstration of lightness constancy ever devised. Square A (appearing dark) and Square B (appearing light, inside the cylinder's shadow) are physically identical shades of grey — pixel for pixel, they reflect the same amount of light to your eye.
Yet no matter how many times you verify this, you cannot see them as the same. Your visual system has already decided that Square B must be a lighter-colored square temporarily darkened by shadow, so it 'corrects' for the shadow and reports the perceived lighter surface color. This correction is usually helpful — it's why a white piece of paper looks white in candlelight and noon sunlight alike.
The brain's lightness computation doesn't work on raw pixel values. It works on relational contrasts — how much lighter or darker is this region compared to its neighbors? Square A's neighbors are light-colored squares, making A appear dark. Square B's neighbors are darkened by the shadow overlay, making B appear light. The identical grey values produce opposite perceptual reports.
Adelson's illusion holds a special place in neuroscience because it shows that 'seeing' involves unconscious inference. You are not perceiving the image on your retina; you are perceiving your brain's best hypothesis about what caused the image. Toggle 'Reveal truth' above to see the connecting lines — and notice you still can't unsee the difference.
Science fact The visual cortex runs a lightness computation before conscious perception: V4 and lateral occipital cortex encode 'surface reflectance' (perceived color) while V1 encodes raw luminance. Adelson's squares activate V1 identically but V4 differently.