Are you color blind?

Your eye has three kinds of cone cells, tuned roughly to long (red), medium (green) and short (blue) wavelengths. Your brain compares their signals to work out color. If one type of cone is missing or shifted, some pairs of colors that look clearly different to most people look almost the same to you.

Each plate uses exactly that. The number is drawn with dots from one set of colors and the background with dots from another. The two sets are matched in brightness and only differ in hue, along a line of colors that a particular kind of color blindness mixes up. If your cones separate those hues, the number stands out. If they don't, the plate looks like a random field of dots, or you see a different number.

The test mixes three kinds of plates:

• Control plates that almost everyone can read, including most people with color blindness. They check that your screen and your answers are working.
• Red-green plates, including some built to lean toward protan (red cone) or deutan (green cone) differences.
• Blue-yellow plates for the rarer tritan type.

The plates are generated by our own code, not copied from a printed book. They follow the same principle Shinobu Ishihara published in 1917, which is still the most widely used color vision screen in eye clinics.

At the end you get one of a handful of outcomes. Each one links to a page with more detail.

• Normal color vision: you read the plates the way people with typical color vision do. A single miss is common and still counts as normal.
• Red-green deficiency: you missed several red-green plates. We estimate whether it looks more like a red cone difference (protanomaly or protanopia) or a green cone difference (deuteranomaly or deuteranopia). Deutan types are the most common form of color blindness.
• Blue-yellow deficiency: you missed several blue-yellow plates, which suggests tritanomaly or tritanopia. This is rare, and screens vary most in how they show blue, so treat it as a reason to retest and get checked.
• No clear pattern: a few misses spread across plate types. Most often this is lighting, screen settings or a slip of the finger. Retake the test before reading anything into it.
• Missed the control plates too: this can point to very limited color vision, another eye problem or a display issue. This one is worth a proper eye exam.

The strength meter shows how many plates on the affected axis you missed. "-anomaly" names the milder form, where one cone type is shifted, and "-opia" the stronger form, where it is missing or barely works. An online test can't measure your cones directly, so the name is an estimate, not a diagnosis.

Most people who are born color blind never need treatment, and many find out as adults. But some situations call for a proper test with an optometrist or ophthalmologist:

• Your result shows a deficiency and color matters for your work. Pilots, electricians, some police and military roles and some commercial driving licenses have color vision standards, often with their own approved tests.
• Your color vision has changed. Inherited color blindness stays the same your whole life. A new change, or one that affects one eye more than the other, can be a sign of an eye or nerve condition, or a side effect of some medicines. Get that checked soon.
• A child struggles with colors at school. Our kids test is a good first look, and a pediatric optometrist can confirm it.
• You want a definite type. A clinic can use a full plate book under standard lighting, an arrangement test like the Farnsworth D-15, or an anomaloscope to tell protan from deutan and measure severity.

Take your result with you. It won't replace their test, but it gives them a useful starting point.