colour tool
Colour blindness simulator
This colour blindness simulator shows your image, or a list of colours, the way it looks with protanopia, deuteranopia, tritanopia and achromatopsia, side by side with the original. It uses the Machado 2009 model, so you can see which colours collapse together before a chart or slide ships.
Loading the tool…
- Runs on your device, nothing is uploaded
- Free, no sign in
- No watermark, no file limit
How it works
- 1
Choose image or palette
Click An image to check a chart, slide or photo, or A palette to paste colours as text.
- 2
Drop the file or paste colours
Choose an image or drop it on the box. For a palette, type one colour per line, with a name after a comma if you want one.
- 3
Compare the five views
Typical vision sits first, then protanopia, deuteranopia, tritanopia and achromatopsia. For a palette, the tool lists every pair that falls below a CIEDE2000 difference of 8.
- 4
Download a view
Each simulated image has its own Download PNG button, so you can drop the evidence into a review.
What the four views show
Protanopia and deuteranopia are the two red green types: the red sensitive or the green sensitive cones are missing, so reds, oranges, greens and browns slide towards the same muddy yellow. Together they affect about 1 in 12 men and 1 in 200 women of northern European descent. Tritanopia, the blue yellow type, is rare and confuses blue with green and yellow with violet. Achromatopsia removes colour entirely and leaves only lightness, which is also a fair preview of a page printed on a black and white printer.
How the simulation works
Each pixel is converted to linear light and multiplied by a 3 by 3 matrix from Machado, Oliveira and Fernandes, published in 2009, which models how the cone responses shift when one cone type is absent. The result is converted back to sRGB. Achromatopsia uses the Rec. 709 luminance weights instead. Transparency is left alone. Working in linear light matters: applying the same matrices to the stored values would darken mid tones and overstate the change.
Checking a palette before it reaches a chart
The palette view is the faster check for chart series and status colours. Paste the colours with their meanings, such as loss, profit and target, and the tool compares every pair under each view using CIEDE2000, the colour difference measure used in print. A pair that is clearly distinct in typical vision but drops below 8 under one view is flagged by name, which is usually the red and green pair everyone already suspected.
Fixing what the views reveal
WCAG success criterion 1.4.1 says colour must not be the only way information is conveyed. When two series merge, the fixes that always work are direct labels on the lines, different marker shapes or dash patterns, and a lightness difference large enough to survive the achromatopsia view. Swapping red and green for blue and orange helps the two common types, and checking the result here again takes seconds.
Limits worth knowing
This is a simulation, not a diagnosis: it shows the full strength form of each type, and real colour vision varies from person to person, so many viewers will see less change than the preview shows. Screens also differ. Nothing leaves this page: the image is decoded and recoloured in your browser, nothing is uploaded and nothing is stored, ever.
Limits
- A simulation, not a diagnosis. It shows the most severe form of each type, and severity varies between people, so many viewers see less change than shown.
- Previews are scaled to 900 pixels on the long side; downloads are scaled to 4096.
- HEIC photos open only in browsers that can decode them,. Elsewhere, convert them to JPG first.
What it does not do
- It does not test your own colour vision. An optometrist does that with plates and an anomaloscope.
- It does not recolour an image to be safer. It shows the problem; the fix is a design decision.
- It does not simulate milder anomalous trichromacy, only the full dichromat forms and achromatopsia.
Questions people ask
Which type of colour blindness is the most common?
Deuteranomaly, a weakened green response, is the most common by far, followed by protanomaly and then the full dichromat forms. Between them the red green types account for almost all colour vision deficiency. Tritan types are rare, and complete achromatopsia, seeing no colour at all, is rarer still. Designing for the red green types covers the large majority of readers.
Is a red and green chart always a problem?
Usually, unless something other than hue separates the series. A dark green against a light red can survive because the lightness differs, and the achromatopsia view will show whether it does. Bright red against mid green at similar lightness merges for most people with protan or deutan vision. Labels on the lines solve it regardless of the colours.
Why do the simulated images look washed out?
Losing one cone type removes one axis of colour, so every colour is projected onto the two that remain, mostly blues and yellows. Saturated reds and greens lose most of their distinctiveness and look duller. That flattening is the point of the preview: it is what makes two chart series or two status badges impossible to tell apart.
Can I check a whole web page or slide deck?
Take a screenshot of the page or export the slide as a PNG, then drop the image here. The tool works on any picture, so a screenshot shows text, charts and buttons together. For a long deck, check the slides that rely on colour to carry meaning: charts, maps, status tables and anything with a coloured legend.
Does the simulation work on transparent PNG logos?
Yes. Only the red, green and blue values are changed and the transparency of each pixel is kept, so a logo stays cut out and the downloaded PNG can be placed over a background to see it in context. Checking a logo on both a light and a dark background is worth doing, since lightness carries most of the difference.
Why are there no settings for milder colour blindness?
The tool uses the Machado matrices at full severity only. Milder anomalous forms sit between typical vision and the views shown, so if a design survives the full strength views it will survive the milder ones too. Showing the worst case keeps the check simple and errs on the side of the reader.
Build a deck with charts that read for everyone
Take what you made here into presentations.
Create a presentation with OneCraftMore tools like this
- Contrast checkerSet a text colour and a background and read the WCAG ratio on a live specimen, with a pass or fail attached to every size from 30px down to 13px.
- Colour contrast matrixColour contrast matrix: paste 2 to 12 colours and see every pair's WCAG ratio with AA and AAA marks, and which pairs are safe for body text.
- Colour palette generatorColour palette generator: pick one seed colour and get monochrome, analogous, triadic and split palettes with hex, RGB, HSL and a WCAG text check.
- Check image size onlineCheck image size online: width, height, aspect ratio, file size, format and colour depth for up to 20 images, plus which flyer sizes each one fills.
- Colour picker from imageDrag a magnifier across a picture and read the pixel under it as hex, RGB, HSL and OKLCH, with sample averaging, arrow key nudging and a history of picks.
Background reading: the question page on this.
Written and checked by the OneCraft team. Last checked .