How to see more colour

Three kinds of cone

Colour vision starts with the cone cells in the retina. Estimates of how many each eye holds vary between sources, from around five million to six or seven million.¹² They come in three types, named for the part of the spectrum they respond to best: long, medium and short wavelengths. People often call them red, green and blue cones, which is handy but a little misleading. The long and medium cones actually peak quite close together, in the yellow-green part of the spectrum, while the short cones peak in the blue at around 445 nanometres.¹² Every colour you have ever seen is the brain's reading of how strongly those three types are firing compared with each other.

The cones are not spread evenly. They are packed most tightly in the fovea, a tiny rod-free patch at the centre of the retina about 0.3 millimetres across. Further out, cone density drops and rods fill the gaps between them.¹² The periphery still has cones, but it samples colour far more coarsely than the centre. That is why the patch of the world you are looking straight at is the only part you see in sharp, reliable colour.

The brain fills in the rest

It does not feel that way. Most of us would say we see a full field of colour all the way out to the edges. A 2020 study at Dartmouth College tested this. Michael Cohen, Thomas Botch and Caroline Robertson put nearly 180 people into virtual reality headsets with eye tracking and let them look around 360 degree scenes of real places. As people looked around, the software kept colour only where they were looking and drained it from the periphery. Most participants did not notice. In the most extreme version, almost a third of observers failed to spot it when less than five per cent of their visual field was still in colour.³⁴ Robertson's reading of the result was that the brain "is likely filling-in much of our perceptual experience".³

For a painter that matters. What you think you saw out of the corner of your eye is partly memory and expectation. If you want to know what colour something really is, you have to look at it directly, and then look again.

Colour depends on its neighbours

Josef Albers built a teaching method on the idea that colour cannot be judged on its own. He joined the Bauhaus as a student in 1920 and taught there until the Nazis closed it in 1933. He then taught at Black Mountain College in North Carolina until 1949, and from 1950 he ran the design department at Yale.⁵ His book Interaction of Color, published by Yale University Press in 1963, grew out of his classroom exercises. Students were given sheets of coloured paper and asked to make one colour look like two different colours, to make two different colours look alike, or to make three colours look like two.⁶ Albers summed up the point of the book in one line: "All color perception is really illusional."⁶

Shadows are a good everyday example. They are rarely the neutral grey we assume. On snow at sunset, the lit surface gets both golden sunlight and white skylight, while the shadow gets only skylight. The shadow reads as violet-blue. Shine a red-filtered lamp on a white wall and the shadow it casts looks pale green. With a green lamp, it looks pink. In each case the shadow appears roughly complementary to the light falling around it.⁷

Practice makes a measurable difference

Colour discrimination can improve with training. In a 2016 study, Alexandra Grandison and colleagues at the University of Surrey had people practise a fine colour judgement over eight days. Their thresholds dropped, so they could tell apart colours that had been too close to separate before. The gains were narrow, though. They held for hues near the ones people had trained on but did not carry over to a different hue region or to a different part of the visual field.⁸ Seeing colour better seems to be learned one patch of the spectrum at a time.

That suggests a few simple ways to train your own eye. Gather several white papers, such as printer paper, cartridge paper, watercolour paper and the back of a business card, and lay them side by side in daylight. They will not match. Some lean warm and some cool. Many papers contain optical brightening agents, which absorb ultraviolet light and give it off again as blue, so they look cleaner and bluer under daylight than under a lamp with little UV.⁹ Then try the same with black objects around the house. Another habit is to pick one thing a day and spend half a minute trying to describe its colour precisely, past "brown" or "green", until the words match what is actually in front of you. Or take an Albers exercise and try to make one colour look like two by changing what sits next to it.

---

1. Georgia State University, Rods and Cones, HyperPhysics.
2. Brian A. Wandell, Foundations of Vision, chapter 3, Sinauer Associates, 1995.
3. Technology Networks, VR Study Participants Fail To Notice Color Removal From 95% of Their Visual World, Technology Networks, 2020.
4. Michael A. Cohen, Thomas L. Botch and Caroline E. Robertson, The limits of color awareness during active, real-world vision, Proceedings of the National Academy of Sciences, 2020.
5. Whitney Museum of American Art, Josef Albers, Whitney Museum of American Art.
6. Eric Gershon, The book on color that never fades, Yale News, 2023.
7. Susan Houde-Walter and Greg Pierce, Colored Shadows and Retinex Vision, Optics & Photonics News, 1992.
8. Alexandra Grandison, Paul T. Sowden, Vicky G. Drivonikou, Leslie A. Notman, Iona Alexander and Ian R. L. Davies, Chromatic Perceptual Learning but No Category Effects without Linguistic Input, Frontiers in Psychology, 2016.
9. ChromaChecker, OBAs, ChromaChecker.