How Color Blindness Works: The Science of Color Vision

Color vision starts with three cones

Your retina has three types of cone photoreceptor, each tuned to a different part of the spectrum: the L cone (long wavelengths, peak near 564 nm, loosely "red"), the M cone (medium, around 534 nm, "green") and the S cone (short, around 420 nm, "blue"). Normal color vision needs all three pulling their weight.

No single cone reads "color" on its own; the eye infers it from the ratio of how strongly the three respond. The weak point is the L and M cones. Their sensitivity curves overlap so heavily that telling red from green comes down to a small difference between two nearly identical signals. That narrow margin is why red-green vision is the first thing to break, and why it accounts for almost all color blindness.

Your brain doesn't use red, green and blue directly

The visual system never passes the raw L, M and S signals straight to the brain. In the retina it recombines them by subtraction into three "opponent" channels: brightness (L + M), red-green (L − M), and blue-yellow (S − (L + M)).

Only two of those channels carry color: red-green and blue-yellow. This is not a convenience we made up for testing; it is how the visual system represents hue, and both behavior (opponent-process theory) and the wiring of retinal cells back it up.

Why color blindness is always red-green or blue-yellow

A deficiency is a fault in one cone type, and it shows up as a broken opponent channel. Damage the L or M cone and the red-green channel (L − M) can no longer be computed cleanly. That is red-green color blindness: protan when the L cone is affected, deutan when it is the M cone. Damage the S cone instead and the blue-yellow channel is the one that fails, which is tritan.

Two color channels means two families. There is no separate "red-blue" channel to lose, so "red-blue color blindness" is not a real category. Inherited color vision deficiency falls on the red-green axis or the blue-yellow one. (Total color blindness, where the cone system barely works at all, is a different and far rarer problem.)

The colors that get confused

When an axis is weak, any two colors that differ only along that axis drift toward looking identical. Plot every such pair and they fall along the classic "confusion lines". In practice a red-green deficiency blurs red with green, green with brown, and red with black, while a tritan one blurs blue with green and yellow with pink.

The three types, and why red-green is so common

Red-green color blindness affects about 1 in 12 men and 1 in 200 women, and the gap between those two numbers is genetic. The L and M cone genes sit right next to each other on the X chromosome and are nearly identical, so they swap and mutate easily. A man has only one X chromosome, so a single faulty copy is enough to show; a woman usually needs both copies affected, which is far less likely.

Tritan deficiency is much rarer, on the order of 1 in 10,000. Its gene sits on chromosome 7 rather than the X, mutates less often, and passes to men and women alike, so there is no sex skew.

How a color blindness test detects it

A pseudo-isochromatic plate is the dotted Ishihara-style image. It hides a number or shape in dots colored so the figure and the background differ only along one confusion axis. Normal vision separates them and the figure is obvious. A weakness on that axis lets figure and background merge into one field of dots, and nothing stands out.

Difficulty comes down to how wide that color gap is. An easy plate keeps figure and background far apart along the axis, so almost everyone reads it. A hard plate narrows the gap toward the confusion line until only an intact cone system can resolve it. Generating fresh plates each session has a second benefit: there is no fixed answer key to memorize beforehand.

How severity is measured

Severity is your threshold: the smallest color difference you can still resolve before recognition fails. Someone with a mild deficiency only misses the hardest, near-threshold plates. Someone with a strong one needs a wide color gap and fails even the easy plates. A test therefore steps each axis from easy to hard and records where reading breaks down. The axis that fails tells you the type; how far you got before failing tells you the severity.

What an online test can and can't tell you

On an ordinary screen, an online test is reliable for one thing: whether a color vision deficiency is present at all. It also gives a reasonable estimate of the type and a relative severity. What it cannot give is a calibrated, absolute measurement. Your screen is not color-managed to a standard, room lighting shifts the colors, and mild cases sit close enough to the red-green boundary to be misread one way or the other. Treat the type and severity as a screening result you can repeat over time, not a diagnosis.

When the answer actually matters, for a job, a licence, or a medical question, confirm it with an eye-care professional on calibrated equipment such as an anomaloscope or a CAD test.