How Light Reveals Colour

Stand in a windowless room and switch the lamp off completely. The red sofa is not a darker red now - it is nothing. It has no colour at all, because colour is not stored in the sofa waiting to be switched on; it is made, from scratch, out of the light that falls on it. Remove the light and you remove the colour with it. This is the plainest possible statement of the idea the whole of Module 3 rests on: light is the source, and every colour you have ever admired was on loan from a light source.

Recall the chain from Lesson 0.1. White light is not colourless - it is every wavelength at once, from long-wave red through to short-wave violet. When that light meets a surface, the surface's pigments absorb some wavelengths and reflect the rest, and the leftover light is what your eye receives and your brain reads as colour. Every link in that chain begins with the light. The pigment can only subtract from what it is given; it cannot invent a wavelength the light never contained. So if the light is missing its reds, a red cushion has no reds to hand back, and it goes dull and brownish - not because the cushion changed, but because there was nothing red for it to reflect.

This is why colour is fundamentally unstable in a way that trips up beginners and burns experienced designers who get complacent. A material's pigment is fixed, but the colour it shows is a collaboration between that pigment and whatever light is present. Two things follow immediately, and both matter on every project. First, you can never fully judge a colour until you know its light - a chip approved under a showroom's cool fluorescent is simply not the same decision as that chip on a north wall or under a warm dining pendant. Second, and more usefully, the light is something you can often choose. Which means colour is not only a pigment problem; it is a lighting problem, and the two must be designed together or not at all.

If light is the source, then the recipe of the light decides everything downstream - and no two kinds of light share the same recipe. What we loosely call 'white' is really a blend of wavelengths, and the exact blend differs enormously from one source to the next. Midday sun, an overcast sky, a candle, a warm LED, a cheap cool LED, a fluorescent tube and an old street sodium lamp each emit a different mixture of energy across the spectrum.

The technical name for that mixture is the source's spectral power distribution, or SPD: a simple graph of how much energy the light emits at each wavelength across the visible band. The SPD is the light's fingerprint, and once you can picture it, the behaviour of colour under different lights stops being mysterious.

Daylight has a broad, fairly continuous SPD - energy present, in reasonable measure, at every wavelength - which is why colours look full and natural in it and why it remains the reference against which lamps are judged. An incandescent bulb or candle has a smooth SPD too, but heavily weighted toward the reds and oranges with very little blue, which is why everything under it looks warm and cosy, and why blues look muddy. Now the awkward one: many LEDs and fluorescents have spiky, gap-toothed SPDs. A typical inexpensive LED makes its 'white' by pairing a strong blue spike with a yellow phosphor, leaving a real dip in the cyan-greens and, crucially, weak output in the deep reds. The light meter still reads 'white', and to a casual glance the room looks fine - but any colour that needed those missing wavelengths to look right will fail.

> A light meter can call two very different SPDs 'white' at the same brightness. Your eye - and your client's skin, and a red cushion - will not agree.

This is the root of almost every 'the colour looks wrong here' complaint. It is not superstition and it is not the painter's fault. It is a mismatch between what the surface needs and what the light's SPD actually offers.

Put the two halves together and you have the working model for this entire module. The colour you see is the light's SPD multiplied by the surface's spectral reflectance - what fraction of each wavelength the surface throws back. Wherever the light has little energy, the surface has little to reflect, no matter how richly pigmented it is. A brilliant crimson lacquer under a light with no deep-red output cannot look crimson; the pigment is starved.

Take a real example a designer meets constantly: a warm terracotta. Under full daylight it reads as its true self - a rich, slightly earthy orange-red, exactly as sampled. Move it under a warm incandescent-style pendant and it becomes even richer and more glowing, because that light is pouring extra reds and oranges onto a surface built to reflect them; the colour is flattered. Now move the same terracotta under a cheap, blue-heavy LED that is thin on deep red, and it collapses into a flat, greyed brown - lifeless, as if someone had thrown a veil over it. Three lights, one pigment, three genuinely different colours. Nothing about the paint changed.

The same mechanism runs in reverse for cool colours. A clear teal or a slate blue can look crisp and clean under daylight or a cool lamp, then go dull and greenish-grey under warm incandescent light, because the warm source is thin exactly where the teal needs energy to sparkle. This is why a scheme that leans on blues and greens can quietly die in a warmly-lit room, and why a warm, earthy scheme can feel muddy under harsh cool light.

The practical rule writes itself: match the light to the colours you care about, or match the colours to the light you are stuck with - but never pretend the light is neutral. A designer who chooses a jewel-toned blue for a room lit only by warm 2700K lamps has, in effect, chosen a different colour than the one on the chip. The pigment is a promise the light either keeps or breaks.

If colour is made from light, then designing colour well begins before you open a fan deck. It begins with a clear picture of the light every surface will actually receive - and there is usually more of it, and more variety, than people assume. A single living room might be lit by strong side daylight in the morning, flat overcast light on a monsoon afternoon, warm pendant light at dinner, and cool task light at a desk in the corner. A colour has to survive all of them, and it will not read the same in each. Your job is not to find a colour that looks identical everywhere - that colour does not exist - but to choose one whose range of appearances you are happy to live with.

There are three habits this leads to, and they run through the rest of the module. First, audit the light before the colour. Note orientation, window size, time-of-day use, and every lamp's character. A north-facing study used mostly by day is a different colour problem from a south-facing one, and both differ from a windowless bathroom lit by a single downlight. Second, judge every colour under its real light, at real size. A chip on a desk under office tubes tells you almost nothing; a large sample on the actual wall, seen morning and night, tells you the truth. Third, specify the light as deliberately as the paint. The lamp's warmth and its ability to render colour honestly (the subjects of Lessons 3.3) are part of the colour decision, not a separate trade's problem to be solved later.

This reframing is quietly powerful. Amateurs pick a colour and hope the room cooperates. Professionals design the conditions - light first, then pigment, then finish, then neighbours - so that the colour people experience is the colour that was intended. The rest of this module is the detail of that craft: how daylight shifts through the day and by orientation (3.2), how to read and choose artificial light so colours stay true (3.3), and the traps of colours that match under one light and diverge under another (3.4). It all grows from the idea you now hold: colour is not a thing you pick, it is light you have learned to control.