Marius Comper Română

Colour vision test · about 3 minutes

Colour blindness test: red, green and blue, measured on screen

A ring of dots has a gap. You say which side it is on, a few dozen times, while the colour difference shrinks until it disappears. At the end you get your thresholds on the three axes of colour vision, beside the published limits for normal vision, and the world as each kind of colour blindness sees it.

Start the testNo account. Your answers stay on your device.

The lens shows the plate through the eyes of a

Inside the small circle, the same plate through the eyes of a deuteranope. The number disappears: its dots differ from the background only in the signal of the M cones, which a deuteranope lacks. Move the lens over the plate.

The test

On each plate a ring of dots has a gap: up, down, left or right. Press the arrow on that side, an arrow key, or the plate itself where the gap is. When you can no longer see it, choose “Can’t see” and the test will pick at random.

Before you start

  • Screen brightness at its usual level, not the minimum.
  • Night Shift, True Tone, “warm light” modes and blue-light filters off: they change exactly the colours being measured.
  • No tinted lenses. Ordinary glasses are fine.
  • The screen at arm’s length, the plate as large as possible.

How this colour blindness test works

Each plate is a ring of dots with a gap, like the letter C. The ring differs from the background only in colour: the brightness of the dots varies at random, so the gap cannot be found by how light or dark a dot is.

The eye has three kinds of cones, sensitive to long (L), middle (M) and short (S) wavelengths. The colours of the ring are chosen to change the signal of one kind of cone while leaving the other two untouched. Someone who lacks that cone, or has an altered one, sees the ring melt into the background.

The test follows three axes at once: L (the protan axis), M (deutan) and S (tritan). After a right answer the colour difference shrinks; after a mistake it grows. Where it settles is the threshold on that axis. If a threshold comes out above the usual limit, the test checks it once more with six plates shown exactly at the limit: five out of six passes. Four control plates, which differ only in brightness, show whether the answers can be trusted.

The method follows the Cambridge Colour Test (Regan, Reffin and Mollon, 1994), and the thresholds are in its units: the distance from grey in the CIE u′v′ colour diagram, multiplied by 10,000. Young adults with normal colour vision stay below 100 on the protan and deutan axes and below 150 on the tritan axis; typical values at 20 are around 40, 40.5 and 53 (Paramei and Oakley, 2014).

How reliable is a colour vision test on a screen

The procedure was run 2,080 times on simulated observers with normal vision, protanopia, deuteranopia, tritanopia, anomalous trichromacy and monochromacy, on four differently tuned screens. No observer with typical normal vision was given a colour-blind result, not even those answering at random on one plate in 12. An observer at the upper edge of normal, with thresholds like the weakest 10% of young people with normal vision, came out “normal” in 97% of runs; the rest were flagged on the blue-yellow axis. No dichromat came out “normal”. For observers with normal vision, dichromacy, monochromacy and marked anomalous trichromacy, the type was right in 97% of cases. For mild forms the type is less certain, and the grade, mild or marked, is rougher.

The simulation checks the logic of the test, not people’s eyes. With a real person the screen, the room light and attention also matter. The blue-yellow axis is the most sensitive to screen settings: Night Shift, True Tone, blue-light filters and “vivid” modes change exactly the colours on it.

What the world looks like with other kinds of colour vision

Drag the line in the middle: on the left the original picture, on the right the same picture with the colours the chosen kind of vision confuses. A simulation shows which colours end up alike. A deuteranope sees the world with one colour axis fewer.

Kind of vision
Picture

Protanopia and deuteranopia are simulated with the Viénot, Brettel and Mollon (1999) projection. Tritanopia uses the same construction, with one plane instead of the two in Brettel’s (1997) model. Anomalous trichromacy is a blend between typical vision and dichromacy, by severity; achromatopsia is rendered as luminance, and blue cone monochromacy as the S-cone signal mixed with luminance. Everything after the first two is an approximation.

What colour blindness is, and its types

Colour blindness is the everyday name for colour vision deficiencies. Most people are trichromats: they have three kinds of cones. If one kind is missing, the person is a dichromat. If one kind has its sensitivity shifted towards another, the person is an anomalous trichromat: they confuse the same colours, but less.

Types of colour vision deficiency. The figures for men come from a study in Norway: 9,049 men screened with plates, and the 725 found classified with an anomaloscope (Waaler, 1927, in Birch, 2012).
TypeCone affectedWhat gets confusedHow common
DeuteranomalyM, shiftedred, green, brown, olive; milder5.01% of men
DeuteranopiaM missingred with green, without red getting darker1.03% of men
ProtanomalyL, shiftedred with green, red slightly darker1.04% of men
ProtanopiaL missingred with green and black, red much darker0.88% of men
Tritanopia, tritanomalyS missing or weakblue with green, yellow with violet1 in 15,000–50,000 people
Complete achromatopsiano working conesall colours; strong sensitivity to lightabout 1 in 30,000 people
Blue cone monochromacyonly S cones workalmost all coloursvery rare

How common colour blindness is in men and women

Among people of Northern European descent, about 8% of men and 0.4% of women have an inherited red-green deficiency (Birch, 2012): roughly one man in 12.

The difference is genetic. The genes for the L and M pigments, OPN1LW and OPN1MW, sit on the X chromosome. A man has one X chromosome, so one copy of these genes; in a woman the deficiency usually appears only if the genes are altered on both X chromosomes. The tritan form comes from the OPN1SW gene on chromosome 7 and is equally common in both sexes.

Prevalence varies a lot between populations. The Unseen Spectrum shows the figures by country and what colour blindness means at school and at work.

Where the name comes from: John Dalton and the eye kept for 150 years

The English chemist John Dalton described his own vision in 1794, in “Extraordinary facts relating to the vision of colours”. He believed the vitreous humour of his eyes was tinted blue and held back red. After his death in 1844 his eyes were examined and found to be clear.

One eye was kept. In 1995, DNA from it showed that Dalton was a deuteranope: he lacked the M-cone pigment (Hunt and colleagues, Science, 1995).

Can an online colour blindness test give a diagnosis?

No. A screen test can show that an examination is worthwhile; the diagnosis is made by an eye specialist, with printed tests or an instrument. Ishihara plates screen only for red-green deficiencies. HRR plates and arrangement tests, the Farnsworth D-15 and the Farnsworth-Munsell 100 hue, also cover the blue-yellow axis. The anomaloscope, in which a yellow light is matched with a mixture of red and green, is the most accurate way to confirm and classify a red-green deficiency.

For a job with colour vision requirements, what counts is the test set by the employer or the licensing authority, not an online result. The limits of normal vision also rise with age: after 60 the usual thresholds are higher than the ones used here, which are those of young adults.

What helps in everyday life

Phones and computers have colour filters that restore some lost differences, in the display accessibility settings: Color Filters on iPhone, iPad and Mac, Color correction on Android, Color filters on Windows 11.

Anyone making charts, maps or interfaces can check in the lab above whether two important colours end up alike. This page’s interface uses only blue, yellow and greys: the pair that stays distinct for people with red-green deficiencies.

Sources

  1. Regan, B. C., Reffin, J. P., Mollon, J. D. (1994). Luminance noise and the rapid determination of discrimination ellipses in colour deficiency. Vision Research 34(10), 1279–1299. doi.org/10.1016/0042-6989(94)90203-8
  2. Paramei, G. V., Oakley, B. (2014). Variation of color discrimination across the life span. JOSA A 31(4), A375 (Table 2; limits from the Cambridge Colour Test handbook). doi.org/10.1364/JOSAA.31.00A375
  3. Fernandes, T. M. P., Santos, N. A., Paramei, G. V. (2020). Cambridge Colour Test: reproducibility in normal trichromats. JOSA A 37(4), A70. doi.org/10.1364/JOSAA.380306
  4. Birch, J. (2012). Worldwide prevalence of red-green color deficiency. JOSA A 29(3), 313–320 (Tables 2 and 4). doi.org/10.1364/JOSAA.29.000313
  5. Smith, V. C., Pokorny, J. (2003). Color matching and color discrimination. In: The Science of Color, 2nd ed., ch. 3, pp. 138–140. sites.socsci.uci.edu/~kjameson/SmithPokorny2003.pdf
  6. Smith, V. C., Pokorny, J. (1975). Spectral sensitivity of the foveal cone photopigments between 400 and 500 nm. Vision Research 15(2), 161–171. doi.org/10.1016/0042-6989(75)90203-5
  7. Viénot, F., Brettel, H., Mollon, J. D. (1999). Digital video colourmaps for checking the legibility of displays by dichromats. Color Research & Application 24(4), 243–252. vision.psychol.cam.ac.uk/jdmollon/papers/colourmaps.pdf
  8. Brettel, H., Viénot, F., Mollon, J. D. (1997). Computerized simulation of color appearance for dichromats. JOSA A 14(10), 2647–2655. doi.org/10.1364/JOSAA.14.002647
  9. Kaernbach, C. (1991). Simple adaptive testing with the weighted up-down method. Perception & Psychophysics 49(3), 227–229. doi.org/10.3758/BF03214307
  10. Hunt, D. M., Dulai, K. S., Bowmaker, J. K., Mollon, J. D. (1995). The chemistry of John Dalton’s color blindness. Science 267(5200), 984–988. doi.org/10.1126/science.7863342
  11. Nathans, J., Thomas, D., Hogness, D. S. (1986). Molecular genetics of human color vision: the genes encoding blue, green, and red pigments. Science 232(4747), 193–202. doi.org/10.1126/science.2937147
  12. MedlinePlus Genetics: Color vision deficiency; Achromatopsia; OPN1SW gene. medlineplus.gov/genetics/condition/color-vision-deficien…
  13. Cleveland Clinic: Color blindness test. my.clevelandclinic.org/health/diagnostics/24845-color-bl…
  14. Apple Support: Change display colors on iPhone (Color Filters). support.apple.com/guide/iphone/change-display-colors-iph…
  15. Google: Use color correction on Android. support.google.com/accessibility/android/answer/11183305
  16. Microsoft Support: Use color filters in Windows. support.microsoft.com/en-us/accessibility/windows/use-co…
  17. CIE. CIE 1931 colour-matching functions, 2 degree observer (data table). doi.org/10.25039/CIE.DS.xvudnb9b

Photographs from Wikimedia Commons