Huely/Color blindness simulator
Around one man in twelve has some form of color vision deficiency. Paste your colors below to see them under deuteranopia, protanopia, tritanopia and achromatopsia — and, more usefully, to be told which pairs stop being distinguishable rather than left to judge it by eye.
| Type | Missing | Roughly |
|---|---|---|
| Deuteranopia | Green cones | ~6% of men — by far the most common |
| Protanopia | Red cones | ~2% of men; reds also look darker |
| Tritanopia | Blue cones | Rare, and affects all genders equally |
| Achromatopsia | All color | Very rare; also a good proxy for greyscale printing |
Red-green deficiency is not one condition but a family. Deuteranopia and protanopia both collapse the red-green axis, but protanopia additionally darkens reds, so a red that merely looks brownish under one can look almost black under the other. Both are shown separately here for that reason.
Each color is converted to linear RGB, multiplied by the Machado, Oliveira & Fernandes (2009) transform matrix for that deficiency, and re-encoded to sRGB. The linearisation is the step that gets skipped most often: applying the matrices directly to gamma-encoded sRGB makes dark colors simulate far too light, which is precisely the region where palettes fail in practice.
Achromatopsia is computed as Rec. 709 luminance in the same linear space, which is why it doubles as a check for how the palette will photocopy or print in greyscale.
The collision warnings compare the perceptual distance between every pair before and after simulation. A pair is flagged only when it was clearly distinguishable to begin with and is not afterwards — so the list is actionable rather than a wall of near-misses.
The instinct is to swap one of the two hues. That usually just moves the problem: the red-green axis is compressed as a whole, so another red will fail the same way. Three fixes that hold up:
Deuteranomaly and deuteranopia — reduced or absent green cone function. Together with protan types they account for the great majority of cases, affecting roughly 8% of men and 0.5% of women of northern European descent. Blue-yellow (tritan) deficiency is far rarer.
Red against green is the classic failure, but so are green against brown, blue against purple, and any pair of mid-tone colors that differ only in hue. The safest structural fix is to make sure any two colors carrying different meanings also differ in lightness.
No. The simulations here model full dichromacy at severity 1.0, which is the strongest case. Milder anomalous forms — deuteranomaly, protanomaly — are more common and less severe, so treat a passing result as encouraging rather than as proof, and a failing one as definitely worth fixing.
Simulate the palette first, since a design's colors come from its palette. For a full interface view, the palette builder applies your colors to a working dashboard layout and simulates that too.
No. The transform runs in your browser. Your colors appear only in this page's URL, so a simulation can be shared as a plain link.