How We See Colour

At the back of each eye, spread across the retina, sit two families of light-sensitive cell. Rods handle dim light and see no colour - which is why a moonlit garden reads in greys. Colour is the work of the cones, and in most human eyes there are three kinds. We name them by the part of the spectrum each is most sensitive to: L cones (long wavelengths, the reddish end), M cones (medium, the greens) and S cones (short, the blues and violets).

It is tempting to imagine these as a red detector, a green detector and a blue detector, but that is not quite how they work. Each cone responds to a broad band of wavelengths, and the bands overlap heavily - the L and M curves in particular sit almost on top of each other. A single wavelength of light does not switch on one cone; it stirs all three to different degrees. Pure yellow light, for example, excites the L and M cones strongly and almost equally, and the S cone barely at all. Your brain does not receive 'yellow' - it receives three numbers, a strength reading from each cone type, and yellow is its interpretation of that particular pattern.

This has a strange and important consequence. Because colour is a pattern of three responses, completely different mixtures of light can produce the identical pattern - and therefore look identical. A pure spectral yellow and a careful mix of red and green light hit the three cones in the same ratio, so you cannot tell them apart, even though the actual light is physically different. This is not a flaw in your eyes; it is the loophole every screen and printer exploits. They never reproduce the real spectrum of a scene. They just trigger your three cones in the right proportion with a handful of primaries, and your brain does the rest.

So the raw material of colour is only ever three numbers per point in your visual field. Everything else - the millions of shades you can name, the subtle warmth of a north-lit wall - is built from that three-channel signal. Understanding this is the first step to understanding why colour behaves the way it does.

The fact that normal human colour vision rests on three channels is called trichromacy, and it quietly shapes the entire practice of colour. Three numbers may sound impoverished, but the arithmetic is generous: if each cone can register many levels of response, the combinations run to millions of distinguishable colours. That is why we can argue about whether a grey has a green or a violet cast - the eye is a remarkably fine comparator, especially for colours seen side by side.

Trichromacy also explains the whole idea of primary colours, a concept students often find slippery. There is nothing sacred about red, green and blue, or about red, yellow and blue. Primaries are simply a small set of colours that, mixed in the right way, can stimulate the three cones across a wide enough range to fake most other colours. Because there are three cone types, three well-chosen primaries do the job - which is why screens use three (red, green, blue light) and why the painter's world settles on a small handful too. We will separate the light-mixing primaries from the pigment-mixing ones carefully in the next lesson; for now the point is that 'three primaries' is a fact about your eyes, not about the colours themselves.

It is worth pausing on how much of colour is therefore construction rather than measurement. Your visual system does not report the physical spectrum the way a scientific instrument would; it compresses a rich, continuous band of wavelengths into three numbers and then interprets them in context. Two of the most powerful pieces of that interpretation - judging a colour by its neighbours, and holding a colour roughly constant as the light changes - are the subject of the rest of this lesson. Both are brilliant survival adaptations. Both are, for the colour designer, a daily source of surprises. The professional habit is not to fight the way vision works, but to know it well enough to predict where it will trip a client, a sample or a scheme.

Because the brain reads colour by comparison, no colour is ever judged on its own - it is judged against whatever surrounds it. This is simultaneous contrast, and once you have seen it you cannot unsee it. Place one identical grey on a blue field and again on an orange field: the grey on blue drifts warm and slightly ochre, the grey on orange cools toward blue-grey. The pigment never moved. Your brain, busy separating each patch from its background, pushes the grey away from its neighbour - away from blue toward blue's opposite, away from orange toward orange's opposite.

The effect works on all three dimensions of colour. A mid-grey looks darker on white and lighter on black (value contrast). A muted rose looks positively vivid ringed by grey and turns drab beside a saturated red (chroma contrast). A beige reads pink beside green and greenish beside pink (hue contrast). The stronger and larger the surround, the stronger the shift - which is exactly why a colour chip held against a white catalogue page is close to useless as a prediction of how that colour will feel across a whole wall, next to a timber floor, under a warm lamp.

This is not a party trick; it is the mechanism behind a great deal of real design. It is why a small accent looks more intense the more neutral its setting, why an all-mid-tone room feels muddy and inert (nothing has anything to react against), and why the same 'safe' greige can look fresh in one home and dirty in another. Skilled colourists use contrast deliberately: they place a quiet ground so an accent sings, or surround a strong colour with relatives to calm it.

The working discipline that follows is simple to state and hard to remember under pressure: never approve a colour in isolation. Judge it at something near its real size, next to the actual materials it will touch, under the actual light it will get. A colour is a relationship, not an object, and simultaneous contrast is the proof.

Not everyone builds colour from three channels. In colour-vision deficiency (often loosely called colour-blindness), one cone type is missing or shifted, so the three-number signal collapses toward two. The commonest forms are red-green deficiencies, where the L and M responses - already close - become hard to separate. The numbers are striking: roughly one in twelve men and about one in two hundred women have some form of it, a consequence of the genetics involved. In a lecture hall, an office, or a client's family, someone almost certainly sees colour differently from you.

What this changes is not that these viewers see 'less' colour in some pitiable way - most navigate the world perfectly well - but that certain distinctions you find obvious can vanish for them. A red-on-green warning, a red line and a green line on a plan, a status dot that means 'go' when green and 'stop' when red: these can become the same muddy tone. The figure shows a red and a green signal that are unmistakable to most viewers collapsing into near-identical olive under a simulated red-green deficiency.

The design response is a rule you will meet again in Module 8, and it is worth adopting now as a habit of mind: never let colour be the only thing carrying a piece of information. Back it up with a second cue - a shape, a symbol, a text label, a position, or a difference in value (light versus dark, which survives almost every kind of colour vision). A stop sign works even in greyscale because it is also an octagon with a word on it. A wayfinding scheme that codes floors by colour should also number them. A drawing that distinguishes services by colour should also label or dash them.

This is not a constraint on beautiful colour; it is a reminder of what colour is for. Colour enriches, delights and organises - but because it is constructed differently in different eyes and different lights, it makes a fragile sole messenger. Design so the message survives when the colour does not, and your work includes more people at no cost to its beauty.