See the Web in Full Color

Why I rebuilt the test from scratch

The old test was based on comparing color swatches — the kind of "do these two squares look the same?" quiz you see all over the web. It worked, but it had a fundamental weakness: most swatch pairs differ in brightness as well as color, which means even someone with severe color blindness can often get the right answer by spotting the brightness difference rather than the color difference. That inflates scores and hides mild deficiencies entirely.

The new test measures something far more precise: your actual chromatic discrimination threshold along the exact color axes where protan and deutan deficiencies live. Luminance cues are masked with noise, so the only way to succeed is through true color vision. The result is a continuous, quantitative measurement of severity rather than a simple pass/fail — which directly translates into better, more personalized correction strength for your eyes.

📚 The science behind it

The new test is built on a method called confusion-axis chromatic discrimination with adaptive staircase — a combination of two major lines of vision research going back more than a century.

The idea of confusion lines — specific paths through color space where people with protan or deutan deficiency can't tell colors apart — traces back to Hermann von Helmholtz and Arthur König in the late 1800s, and was later formalized mathematically with the copunctal point coordinates published by Glenn Fry in 1992. The extension uses exactly those coordinates to generate its test colors.

The adaptive staircase procedure — where test difficulty adjusts based on your responses to home in on your threshold — was formalized by Harry Levitt in 1971 and is now the gold standard for measuring perceptual thresholds of any kind, from hearing tests to contrast sensitivity.

Combining these two ideas gave rise to the modern clinical color vision tests used today: the Cambridge Colour Test (CCT), developed by J. D. Mollon and colleagues at the University of Cambridge, and the Colour Assessment and Diagnosis (CAD) test, developed by Professor John Barbur's team at City, University of London. The CAD test is used by the UK Civil Aviation Authority to certify pilots, and similar methods are used in research labs, military vision screening, rail and maritime certification, and studies on the genetics of color vision.

ClearColorVision brings that same methodology — confusion-line stimuli, luminance-masked targets, adaptive staircases, catch trials, and quantitative thresholds — into a free browser extension. It won't replace a calibrated lab instrument, but the underlying method is the same one trusted for pilot certification.

🎯 A deliberate decision: focusing on what I can measure well

As part of this update, I've removed tritan (blue-yellow) testing from the extension. This wasn't an easy call, but it was the right one. Tritan deficiency is extremely rare (affecting roughly 1 in 10,000 people compared to about 1 in 12 men for red-green deficiency), and accurately testing for it on an uncalibrated consumer display is genuinely difficult — the blue-yellow axis is especially sensitive to display white point, backlight spectrum, and ambient lighting, all of which I can't control from inside a browser.

Rather than ship a tritan test I wasn't confident in, I chose to concentrate fully on the protan and deutan axes — which cover the overwhelming majority of color vision deficiency cases — and make those tests as accurate as I possibly can. Fewer features, done properly, beats more features done poorly. I may revisit tritan testing in the future once I can solve the calibration problem honestly.