Color Blindness Simulator

See any image the way a dichromat sees it. Useful for checking whether a chart, map, dashboard or UI still reads when one cone axis is gone.

How the Simulation Works

Each pixel is converted from sRGB into LMS space, the response of the three cone types, using the matrix from Viénot, Brettel and Mollon (1999). The coordinate belonging to the missing cone is then thrown away and reconstructed from the other two, which is exactly what a dichromat's retina does. The result is converted back to sRGB and drawn.

That makes this a simulation of dichromacy: protanopia, deuteranopia and tritanopia, where a cone type is absent. It is the honest worst case. The far more common anomalous forms, deuteranomaly above all, keep partial discrimination, so a real deuteranomalous viewer generally sees somewhere between the first panel and the third.

Your monitor is not color-calibrated and neither is the room you are in. Use this to find color pairs that collapse into each other, not to judge exact shades. Same engine, incidentally, that validates the plates in the color blind test, see types of color blindness for what each axis means.

Designing So It Survives the Simulation

• Never let color be the only signal. Pair it with shape, position, label or texture.
• Vary lightness as well as hue, protan and deutan collapse hue, not brightness.
• Red on green, and green on brown, are the two pairings that fail hardest.
• Blue against orange survives red-green deficiency almost intact, and is the safest two-color pair.
• Check your error and success states. Red/green status dots are the classic failure.