See a picture as protanopia, deuteranopia, tritanopia or achromatopsia shows it. Compare and save PNGs, on your device.
A color blindness simulator shows how a picture looks to a person with protanopia, deuteranopia, tritanopia or achromatopsia, by converting the colour of every pixel with a fixed matrix in linear light. Each pixel's red, green and blue are first decoded from sRGB to linear light. A 3 by 3 matrix then mixes them, and the result is encoded back to sRGB. The matrices for protanopia, deuteranopia and tritanopia are the severity 1.0 matrices of Machado, Oliveira and Fernandes, 2009, in IEEE Transactions on Visualization and Computer Graphics 15(6). Achromatopsia uses relative luminance, 0.2126 red plus 0.7152 green plus 0.0722 blue in linear light, for all three channels. White and greys come out unchanged. Pure red 255, 0, 0 becomes 109, 95, 0 in protanopia and 163, 144, 0 in deuteranopia, pure blue 0, 0, 255 becomes 0, 107, 150 in tritanopia, and in achromatopsia pure red, green and blue become greys of 127, 220 and 76.
Protanopia is the absence of working red-sensitive cones, deuteranopia the absence of green-sensitive cones, and tritanopia the absence of blue-sensitive cones. Protanopia and deuteranopia are the red-green types: they confuse reds, greens, browns and oranges that look different to most people, so check both. Tritanopia mainly confuses blues with greens and yellows with pinks. Achromatopsia is the absence of colour vision, where only lightness remains, so the result is a grey picture whose grey levels follow how bright each colour is. That is one simple model of it, and people with the condition also tend to have other differences in sight that no colour filter shows.
Strength 100 percent is the full simulation above. Lower values blend the simulated picture with the original, pixel by pixel, in linear light, which gives a milder effect. This is a simple blend. It is not the same as Machado and colleagues' separate matrices for each degree of partial colour weakness, such as protanomaly or deuteranomaly, and this tool does not include those. Real vision also varies from person to person, and the result depends on your screen's own colours, so treat the preview as a guide to which colours get confused rather than a measurement.
Use it to check charts, maps, status colours and logos: if two colours that carry different meanings look the same in the red-green types, add a label, a pattern or a difference in lightness. These simulations are approximations of average perception, not a diagnosis, and no picture here can tell you whether someone is colour blind. The preview shows the original and all four types side by side, or one at a time. Saved files are PNG, one for each type you tick, made from the original at its full size, up to 50 images per batch. A picture too large for your device's canvas is made smaller. The photo details stored in the original, such as location, are not copied.
Add the image and look at the Protanopia and Deuteranopia pictures in the preview, which show the two red-green types. Reds and greens both move toward yellows and browns, so objects that differ only in red against green become hard to tell apart. Choose One at a time to see one type large, with Normal vision to compare.
Check protanopia and deuteranopia together, because both confuse red with green and brown. Then tritanopia if your design relies on blue against green or yellow against pink, and achromatopsia as a quick test of whether your colours differ in lightness, since a grey version shows whether two colours still separate.
No. It filters a picture the way a model of colour blindness predicts, so you can judge your design. It cannot test anyone's eyes. A person who wants to know whether they are colour blind needs a test from an eye care professional.
Simulators use different methods and colour spaces. This one decodes sRGB to linear light and applies the Machado, Oliveira and Fernandes (2009) matrices at severity 1.0. Another tool using Brettel, Viénot or a different space can give slightly different colours for the same picture.
Not exactly. Strength blends the full simulation with the original, so 50 percent is halfway between normal and full deuteranopia. The published model has its own matrix for each degree of colour weakness, and this tool does not include those, so use it as a milder preview.
No. The simulation runs on a canvas inside this tab, so no server receives your picture. The page counts visits with Google Analytics using rough groups such as 1 to 10 MB, never a file name or an image. Your choice of view, strength and types to save is remembered on this device.