Color Blindness Simulator
See an image or a palette the way people with red-green or blue-yellow colour blindness see it. 100% free, no signup. Everything runs in your browser.
Loading the tool…
Around one in twelve men and one in two hundred women have some form of red-green colour deficiency, which means that in any room of fifty people, somebody cannot tell the red line on your chart from the green one. Designers know this and still ship the chart, because without seeing it the way that person does, the problem is abstract. I built this to make it concrete. Drop an image, a chart, a screenshot of an interface, and it renders four simulated versions beside the original: deuteranopia and protanopia, the common red-green forms; tritanopia, the rare blue-yellow one; and achromatopsia, no colour at all.
The simulation uses the matrices from Machado, Oliveira and Fernandes' 2009 model, which is the one Chrome's developer tools use for the same purpose and the closest thing the field has to a standard. There is also a palette checker: type up to five colours and the page shows them under each simulation and names the pairs that collapse into each other, which is the question that actually matters when you are picking colours for a status badge or a map. Everything runs on your device. When a pair fails, the usual fix is a second cue, and the colour contrast checker confirms the text still meets the WCAG ratio once you have changed it.
How to use
- For a palette, click each swatch at the top and set its colour. The table below updates with each simulation and lists any pairs that become hard to tell apart.
- For an image, drop a chart, screenshot or photo onto the box, or click to choose one.
- Compare the five versions side by side. Look for lines, labels or regions that were distinct in the original and are not in a simulation.
- Download any simulated version to attach to a design review or a bug report.
- Fix problems by adding a second channel: a label, a pattern, a shape or a brightness difference, rather than by finding a different red and green.
Why use our color blindness simulator?
A published model, not a guess. The three dichromacy matrices are the severity 1.0 values from Machado et al. 2009, which model the physiology of the missing cone type rather than just desaturating red and green. They are the same matrices behind Chrome DevTools' vision-deficiency emulation, applied the same way, so what you see here matches what a developer would see there.
Tells you which colours fail, not just that something looks off. The palette checker measures the distance between every pair of colours after simulation and flags pairs that were clearly different for typical vision and are nearly identical for the simulated one. A chart with six series and two of them merging under deuteranopia is a specific, fixable finding rather than a vague worry.
Every type, including the one nobody tests. Most checkers stop at red-green. Tritanopia is rare but real, and it breaks blue-against-yellow schemes that pass every red-green test. Achromatopsia is the strictest check of all: if a design still works in greyscale, it works for everyone, which is why it is worth looking at even though almost nobody has it.
Runs on your device. The image is drawn to a canvas and each simulation is computed pixel by pixel in the browser; nothing is uploaded. For an unreleased interface or a confidential dashboard, that is the difference between checking it and leaking it.
Fast enough to use while designing. Images are scaled to 900 pixels wide for the preview, so even a large screenshot renders all four simulations in well under a second. Change a colour in the palette and the table updates immediately. The colour palette extractor pulls the colours out of an existing design so you can test them here.
Who is this tool for?
Charts and dashboards. Line charts, pie charts and heatmaps lean on colour harder than anything else on the web, and red-versus-green is the default in half the charting libraries. Screenshot the chart, drop it here, and see whether the series survive. Usually the fix is direct labels or distinct line styles.
Status colours in an interface. Red for error, green for success, amber for warning: the universal convention, and invisible to one in twelve men. The palette checker shows whether your specific red and green survive, and the usual answer is to add an icon so the colour is reinforcement rather than the only signal.
Maps and data visualisation. Choropleth maps with a red-to-green scale are a classic failure. Simulating the map shows whether the scale reads as a gradient or as mud, and a single-hue scale from light to dark almost always reads better for everyone.
Reviewing a brand palette before it is locked in. Five swatches in the checker, four simulations, and a list of pairs that merge. Catching that the primary and the accent collapse under protanopia before the style guide is printed saves an argument later. The OG image generator is where that palette ends up next.
Frequently asked questions
As accurate as the published model, which is good and not perfect. The Machado matrices simulate full dichromacy, the complete absence of one cone type. Many people have milder anomalous trichromacy and see more colour than the simulation suggests, so if a design works here it works for them too. Applying the matrices in sRGB rather than linear light is a simplification the model's own users, including Chrome, accept.
Both are red-green deficiencies. Protanopia is the absence of the long-wavelength cone and makes reds look darker as well as shifting them; deuteranopia is the absence of the medium-wavelength cone and is the more common of the two. In practice they fail the same designs, and a palette that survives one usually survives the other.
Red and green are the obvious pair, but also green and brown, blue and purple, and light green and yellow. The palette checker will tell you for your specific colours. The more reliable approach is to avoid relying on hue alone: vary lightness as well, since differences in brightness survive every form of colour blindness including achromatopsia.
No. The image is read into a canvas in your browser and each simulation is computed there. Nothing is sent anywhere. Download links are generated on your device too. The EXIF remover is the tool to run first if the image is a photo that will be shared afterwards.
Roughly 8% of men and 0.5% of women of Northern European descent have some red-green deficiency, with lower rates in other populations. Tritanopia affects well under one in ten thousand. Achromatopsia is rarer still. The red-green figure is the one that matters for design: on any site with meaningful traffic, that is a lot of people.
It addresses one dimension. Contrast, text size, focus states and screen reader support are separate requirements. The colour contrast checker covers the ratio between text and background, which colour blindness simulation does not measure.

