What is a colour blindness simulator?

It is a tool that re-renders an image the way someone with a colour-vision deficiency (CVD) would perceive it. By transforming each pixel through a model of the missing or shifted cone response, it lets people with normal colour vision see which colours become hard to tell apart — for example reds and greens collapsing toward the same muddy tone under deuteranopia.

Which types of colour blindness does it simulate?

Four. Protanopia (missing red cones), deuteranopia (missing green cones — the most common form), tritanopia (missing blue cones, which is rare) and achromatopsia (total colour blindness, perceived as greyscale). The original image is always shown alongside for reference.

How accurate is the simulation?

It uses well-established linear-light dichromacy matrices applied in proper linear-RGB space (sRGB is first converted to linear light, transformed, then converted back). This matches the approach used by professional accessibility tooling. No model is a perfect substitute for testing with real colour-blind users, but it reliably reveals the problem areas in a design.

Is my image uploaded anywhere?

No. The whole simulation runs locally in your browser using the HTML canvas. Your image never leaves your device, there is no server round-trip, and nothing is stored after you close the tab. That makes it safe for confidential mockups, medical imagery or private photos.

Why does the severity slider matter?

Most colour-blind people are not full dichromats — they have anomalous trichromacy, a milder, partial form. The severity slider blends the full-strength simulation with the original so you can model, say, a 50% deuteranomaly. Set it to 100% for the worst case and to a lower value to approximate a milder deficiency.

Can I download or export the result?

Yes. Each simulation tile has a download button that exports that view as a PNG, and the large preview has its own Download this view button. The exported file keeps the same pixel dimensions as the processed image (very large uploads are downscaled to a maximum edge of 1400px for speed).

Image Color Blindness Simulator

An image colour-blindness simulator that lets you upload any photo, screenshot, chart or UI design and instantly preview it through four kinds of colour-vision deficiency — protanopia, deuteranopia, tritanopia and achromatopsia. It is built for designers, developers, data-visualisation authors and accessibility reviewers who need to check that their work stays readable for the roughly 1 in 12 men and 1 in 200 women who see colour differently. Everything runs in your browser; your image is never uploaded.

How it works

Colour-vision deficiency happens when one of the three cone types in the retina is missing or shifted. To simulate it faithfully you cannot just nudge the raw sRGB numbers — those values are gamma-encoded. This tool first converts every pixel from sRGB into linear light, applies the appropriate dichromacy transform, then converts back to sRGB for display.

For protanopia, deuteranopia and tritanopia it multiplies each linear-RGB triple by a standard 3×3 simulation matrix that collapses the colour space along the axis the missing cone would normally separate. For achromatopsia it computes the Rec.709 luminance of each pixel, so the result is a perceptually weighted greyscale rather than a naive average. A severity slider then blends the full-strength result with the original, letting you model partial (anomalous) colour vision as well as full dichromacy.

Because the heavy work is per-pixel, large uploads are downscaled to a maximum edge of 1400px before processing — enough detail to judge a design while keeping the preview responsive on a phone. You can click any thumbnail to enlarge it and export every variant as a PNG.

Example

Suppose you build a dashboard where “good” rows are green and “alert” rows are red. Upload a screenshot and switch the large preview to deuteranopia: the two states collapse toward the same brownish tone and become nearly indistinguishable — a real accessibility failure. Drop the severity to 40% and they separate slightly, showing that a mild deuteranomalous user might still cope, but a full dichromat would not. The fix is to add a non-colour cue (an icon or label), which you can re-test the same way.

Load the built-in sample card to see the effect immediately: a row of six labelled colour bands. Under protanopia the red and orange bands merge; under tritanopia the blue and green bands shift dramatically while reds barely move — a quick, intuitive primer on how each deficiency reshapes the palette.

Every pixel is processed locally in your browser. Nothing is uploaded, and nothing is stored once you close the tab.