What the simulator shows
Choose An image or A palette, then pick a vision type. In image mode, drag across the picture to compare typical vision on the left with the simulation on the right, and the strip underneath shows all seven types at once. In palette mode, every color is shown as each type sees it, and pairs that become hard to tell apart are listed.
The vision types
Most people have three kinds of cone cells, tuned to long (L, "red"), medium (M, "green") and short (S, "blue") wavelengths. Color vision deficiencies come from one type missing or shifted.
| Type | Cone affected | Roughly how common |
|---|---|---|
| Deuteranomaly | M shifted | About 5% of men |
| Protanomaly | L shifted | About 1% of men |
| Deuteranopia | M missing | About 1% of men |
| Protanopia | L missing | About 1% of men |
| Tritanomaly, tritanopia | S affected | Rare in both sexes |
| Achromatopsia | No working cones | About 1 in 30,000 |
Red-green types are inherited on the X chromosome, which is why around 1 in 12 men but only about 1 in 200 women of Northern European ancestry have one. Figures vary between populations.
The "-anopia" types are the extreme case. The "-anomaly" types are simulated here at moderate strength; real severity ranges from barely noticeable to almost the same as the full form.
The model behind it
The simulation uses the matrices published by Machado, Oliveira and Fernandes (2009), a physiologically based model widely used in accessibility tools. Each pixel is converted from sRGB to linear light, multiplied by the matrix for the chosen type and converted back. Achromatopsia is shown as relative luminance only, the same brightness measure used for WCAG contrast.
No simulation reproduces anyone's experience exactly. People with the same diagnosis differ, and the brain adapts in ways a matrix cannot capture. Use the results to find problems, not to decide how a particular person sees.
Checking a palette
Paste HEX codes separated by spaces, commas or new lines (up to 16). For the selected type, the tool compares every pair of colors with CIEDE2000, a standard measure of visible color difference:
- A pair is flagged when it is clearly distinct for typical vision (ΔE 10 or more) but drops below ΔE 10 in the simulation.
- The listing shows both numbers, for example "ΔE 34 → 4.2", so you can see how severe the collapse is.
Around ΔE 10 colors start to look like variations of one another when they sit apart, as they do in chart legends.
Fixing the problems you find
- Vary lightness, not just hue. Red and green with the same lightness merge for deuteranopes. A dark red against a light green stays readable.
- Add a second cue. Labels, patterns, icons or direct annotations make charts readable without color.
- Avoid the classic traps: red/green status colors, green/brown, blue/purple and pink/gray are frequent confusions.
- Check in grayscale. If a design survives the achromatopsia view, lightness alone carries the information.
Then check text separately with the contrast checker. WCAG contrast is based on luminance, which stays a useful guide for most vision types, though people with protan types see saturated reds darker than the ratio suggests.
Images and privacy
Images are processed on your device, scaled to at most 2,400 pixels on the long side for speed, and never uploaded. Download PNG saves the simulated version at that size, useful for design reviews and bug reports.
Related tools
The color blind test screens your own color vision with generated dot plates, and the hue test measures how finely you order similar hues. To build an accessible palette from scratch, start in the color palette generator, and use the grayscale image converter to check lightness structure.
FAQ
Is red-green color blindness really two different things?
Yes. Protan types involve the long-wavelength cones and make reds look darker. Deutan types involve the medium-wavelength cones and leave reds about as bright. Both confuse reds, oranges, yellows and greens, but in slightly different ways.
Why does the simulated image look only slightly different?
Many photos have few colors along the confusion lines, such as skies, skin and wood. Try a picture of fruit, a map, a chart or traffic lights to see bigger changes.
Which type should I test designs against?
Start with deuteranopia and protanopia, which cover the most common problems. Then check tritanopia and grayscale.