You see it on a pavement in the middle of the afternoon. Someone in an all-black outfit that clearly matched when they left the house, and now the jacket is faintly warm and brownish against trousers that have gone cool and slightly blue. Nothing has faded. Nothing was bought worn out. The two pieces sat side by side under a shop light an hour earlier and looked like one colour, and daylight has quietly pulled them apart.
This is the most common failure in the most popular outfit formula there is, and it gets blamed on the wrong things. People assume the cheaper piece has washed out, or that they misremembered. What actually happened has a name, a measurement standard, and a set of laboratory procedures built specifically to catch it before a garment ships.
Two blacks, one light source
Colour is not a property of an object on its own. What reaches your eye is the light source multiplied by whatever the surface reflects back at each wavelength. Two surfaces can hand your eye the same total signal while reflecting completely different mixtures underneath, in the way two different recipes can arrive at the same weight.
The international body that writes the rules for colour measurement, the CIE, defines that condition precisely. Two specimens are metameric when they carry identical tristimulus values for a given illuminant and a reference observer while their spectral properties differ. Change the illuminant, or change the observer, and the match falls apart. The CIE has published on this since at least 1989, when it issued a report on the mismatch that appears when one person with normal colour vision is swapped for another.
Black is the worst offender because a black is never one dye. It is a heavy mix built to absorb across the whole visible range, and there are many mixtures that get close enough to look like an absence of colour. Two of them absorbing slightly differently at the red end will look identical under a warm shop light and part company outdoors.
Fibres take different dyes, so blacks start apart
The reason your own wardrobe is full of these near misses is chemical. Dye classes are matched to fibre, not to colour. Cotton is usually coloured with reactive dyes that bond to cellulose. Polyester will not accept those at all and takes disperse dyes that migrate into the fibre under heat. Wool and nylon take acid dyes. Viscose, acrylic, and elastane each behave differently again.
So a black cotton shirt, a black wool trouser, and a black polyester lining were never coloured with the same chemistry. Their reflectance curves are unrelated by construction, and any agreement between them under one particular light is a coincidence rather than a match. Mills manage this deliberately when a single garment mixes fibres, which is why a jacket body and its sleeve lining are checked against each other before production runs. Nobody is checking your jacket against your trousers.
What the labs check that a shop floor cannot
The industry answer is to look twice, under different light, with instruments and with eyes. Colourists judge a pair under a daylight illuminant, then again under a shop or household light, and quantify the gap. ISO 18314-4, reissued in 2024, sets out the formalism for calculating that metamerism index for pairs of samples at a change of illuminant. On the visual side, AATCC publishes an evaluation procedure, EP9, that standardises how a trained observer assesses colour difference in textiles, including who is allowed to make the call after a colour vision screen.
None of that machinery exists on a shop floor. A fitting room typically has one lighting scheme, chosen to flatter, and increasingly an LED source with a spiky output that looks white while missing whole regions of the spectrum. It is a single-illuminant judgement, which is exactly the condition under which a metameric pair looks like a match.
How often a match quietly breaks
There is a measured answer to how likely this is, and it is more interesting than a plain warning. Working with fifty hyperspectral images of natural scenes, David Foster and colleagues reported in the Journal of the Optical Society of America A in 2006 that true metameric pairs are genuinely rare across all possible surface pairings, with a relative frequency somewhere around one in a million to one in ten thousand when the light shifts between two daylight conditions.
The second number is the one for a wardrobe. Restrict the count to pairs that already looked indistinguishable under the first light, and the conditional frequency jumps to roughly one in a hundred up to one in ten. The risk is concentrated precisely among the things you thought were the same colour, which is the whole set you assemble an all-black outfit from.
A two-light test you can run before paying
The professional method shrinks down to something you can do in three minutes without buying anything.
- Hold the two pieces edge to edge with no gap. Any daylight between them lets your eye recalibrate, and the difference stops being visible.
- Take them to the shop door or a window. Daylight is the second illuminant, and it is free. A pair that survives the doorway usually survives the street.
- Look at the fold, not the flat. The shadowed side of a fold exaggerates any warm or cool cast that the flat surface hides.
- Check them against the trousers you are actually going to wear, not against the shop's own sample. Carry the trouser leg or a cuff of it with you if the purchase matters.
- Note the fibre content of both. Same fibre and similar construction is the closest thing to a shortcut, because the dye chemistry is likely to be related.
The photograph test is the fastest of all. Your phone camera applies one white balance across the whole frame, so a shot of both pieces together in daylight flattens out the compensation your brain performs and often shows a split you were talking yourself out of.
Dressing in black without pretending it matches
The fix that works is not stricter matching. It is designing the outfit so an exact match was never load-bearing. Break the black with a deliberate separation: a belt, a shirt cuff, a knit collar, a bag that clearly sits in a different tone. Once the eye reads two intentional blacks it stops auditing them against each other.
Texture does the same job from the other direction. A wool trouser and a cotton jacket will never share a reflectance curve, but they also do not read as a failed attempt at one colour, because the surfaces are obviously different materials. The outfits that look wrong are the ones where two smooth flat fabrics of similar weight almost agree.
Practical order for the rest of the season: pick one anchor piece, usually the coat or the trousers, and buy future black items in the same fibre family where you can. Keep the doorway test for anything meant to pair with that anchor. And retire the pieces that have genuinely gone grey rather than trying to rescue them with dye, since a home dye bath adds a third chemistry to a problem caused by having two.