Color Blindness Simulator
See a picture the way people with protanopia, deuteranopia, tritanopia or achromatopsia see it.
Runs in your browser Your input never leaves your device.
Drop a picture here, , or .
The picture is processed in this page and never uploaded.
No red cones: reds look dark and greenish, about 1% of men
No green cones: the most common kind, about 1% of men
No blue cones: blues look green and yellows pinkish, very rare
No colour at all: only light and dark, very rare
A simulation is an approximation: real colour vision varies between people. Do not rely on colour alone to carry meaning; add shapes, labels or patterns.
Report a problem with this tool
About this tool
About one man in twelve and one woman in two hundred have some form of colour vision deficiency, which makes some colours hard to tell apart. This tool shows your picture as the main types see it: protanopia (no working red cones), deuteranopia (no green cones, the most common), tritanopia (no blue cones, rare) and achromatopsia (no colour). The severity slider goes from mild to complete, which models the common partial forms. Use it to check charts, interface designs, maps and brand colours: if two things look the same in a simulation, add a difference that does not depend on colour, such as labels, patterns, shapes or brightness. The simulation uses the model of Machado, Oliveira and Fernandes (2009) on linear colour values, the same model that Python's colorspacious library uses; we compared 3,036 colours at four severities with that library and the results were identical to the last digit. A simulation is an approximation of an average viewer, not of every person, and screens differ. You can download each simulated picture as a PNG. The picture is processed in your browser and never uploaded.
How to use it
- Drop a picture, or use the colour chart.
- Set the severity.
- Compare the simulations with the original.
- Download any simulation if you need it for a report.
Common problems
- Two colours look the same in a simulation
- That is the problem the tool is meant to find. Change one of them in brightness, add a label or a pattern, or choose colours that differ on the blue-yellow axis as well as red-green.
- The partial severity does not look like my case
- Severity follows the published steps of the model and is an average. Colour vision varies between people; ask someone who has it when it matters.
- The picture is large and slow
- The simulation is computed on a copy of at most 900 pixels on the long side, so previews stay quick.