Colour blindness test
Put the colours in order so each sits between its two nearest neighbours. The first and last are fixed. Tap one, then tap another to swap them.
Tap a colour to pick it up.
Scored in your browser · nothing is uploaded
A hue-arrangement exercise: put the colours in order and get an error score. It is a screening indication only — screen calibration affects the result, and only an optometrist can tell you anything definite.
How to use the colour blindness test
This is a hue arrangement rather than an Ishihara plate, and that is a deliberate and fairly important choice. Ishihara plates — the dotted circles with a number hidden in them — depend on precise printed pigments under controlled lighting. Reproduced on an uncalibrated screen at whatever brightness you happen to have, they test your monitor at least as much as your eyes, which is why an online "Ishihara test" can tell a person with normal vision that they are colour blind and vice versa.
Arranging hues by similarity is the basis of the Farnsworth–Munsell tests, and it degrades more gracefully. The score is the standard total error: for each cap, how far its position is from its two neighbours in the true sequence. A perfect arrangement scores zero, and because the caps form a ring, starting at a different point or going the other way round is not an error — only genuine mis-ordering counts.
The two end caps are fixed as anchors. Without them there is no reference for which direction the ring runs, and a perfectly-arranged ring in reverse would be indistinguishable from a random one.
What this cannot do is tell you which type of colour vision deficiency you might have. Distinguishing protanopia from deuteranopia needs specific confusion axes tested under known conditions — an anomaloscope, in a clinic. What an arrangement score gives you is a rough signal of whether your discrimination is in the usual range. If it is not, turn the brightness up and try again before concluding anything, and then see an optometrist.
Questions
They depend on exact printed pigments under controlled light. On an uncalibrated screen they test the monitor as much as the eye.
The total error: how far each cap sits from its true neighbours. Zero is perfect. It is a rough signal, not a measurement.
No. Distinguishing protanopia from deuteranopia needs an anomaloscope in a clinic.
Without anchors there is no reference for which way the ring runs, and a perfect reverse order would look random.
Turn the screen brightness up and try again first. If it persists, see an optometrist, who can test properly.