Color Blindness Simulator

They're approximations based on widely-cited linear RGB transforms. Real human vision varies; treat the previews as 'how this might look', not exact.

Deuteranomaly (mild green weakness) is by far the most common — about 5% of men. Protanomaly and protan/deutan dichromacy together account for most CVD.

Complete color blindness — the world is seen in shades of gray. Very rare, but a good worst-case check: if the design still works in pure grayscale, it'll work for everyone.

It isn't — the simulation runs locally with Canvas. The image and the matrices both stay in your browser.

Enough to measure. Pure red and pure green sit 361 units apart in RGB. Under the protanopia matrix red becomes (145, 142, 0) and green becomes (110, 113, 62) — 77 units apart, a 79% loss. Deuteranopia takes them to (159, 179, 0) and (96, 77, 77), 143 apart, a 60% loss. That is the whole argument against colour-only status indicators in one number: a pair your eye reads as maximally different can arrive four times closer than you designed it.

It shifts the mid-tones. Physiological simulation models such as Brettel's and Viénot's are defined on linear RGB, and applying a matrix straight to the gamma-encoded bytes makes the result somewhat darker and more saturated than a linearised pipeline would. It is done that way here because it is what essentially every web CVD simulator does, so the output is comparable with the tools you would check against. Treat it as a reliable answer to 'can these be told apart' and not as a colorimetric prediction of what someone sees.