Why Two Greys Match in One Light and Clash in Another

The 2 greys are a metameric pair: different pigment mixtures that produce the same colour signal under 1 light and different signals under another. Why two greys match in one light and clash in another has nothing to do with a faulty batch and nothing to do with your eyes. It is what happens when 2 surfaces arrive at the same grey by different spectral routes.

Metameric matches are most common in near-neutral and dark colours, which is exactly why grey is where most people meet this for the first time. The colour of any surface is the product of its own reflectance curve and the spectrum of whatever is shining on it, so change the lamp and 2 curves that happened to agree stop agreeing.

Why do two greys match in one light and clash in another?

A colour match is a match of 3 numbers, not of a spectrum. Human eyes carry only 3 types of cone, so every spectrum reaching the retina is reduced to 3 tristimulus values, and a very large number of different wavelength mixtures can land on the same 3. Two greys that share those numbers under your kitchen lamp are metamers; the moment they part company under a different lamp is called illuminant metameric failure.

The split is measurable rather than a matter of opinion. Delta E is the colour difference the International Commission on Illumination defines between 2 samples, and the 1976 formula was later revised so that a value of about 2.3 stands for a just-noticeable difference. A metameric pair can measure near 0 under 1 illuminant and well past 2.3 under the next, without either sample changing in any way.

This is also why matching grey to grey is harder than matching a wood tone to a floor. Houseland's 6-piece shelf set is finished in warm honey pine under a clear matte lacquer, and the product page says plainly that natural pine varies in grain, knots and tone, so 2 sets are never identical. A material that never claims a match cannot fail one, which is a real advantage in a room with 4 different light sources in it.

How can you catch a metameric pair before you buy?

Look at the 2 samples together under at least 2 light sources that differ in spectrum, not just in brightness. Ten minutes of testing settles a question that a returns policy cannot, and the sequence below is the one professional colour work uses, scaled down to a living room.

1. Put the 2 samples edge to edge, touching. A gap of even 1 finger width lets the eye adapt separately to each and hides small differences.
2. Look at them in daylight within 3 ft of a window, with the sun off them. This is your closest domestic stand-in for the daylight illuminant colour science uses.
3. Switch off the daylight, close the curtains, and look again under the room lamp alone. Tungsten and warm LED sit near 2700 K, roughly half the colour temperature of noon.
4. Repeat under any fluorescent or under-cabinet strip in the house. These have the spikiest spectra in a home and split metameric pairs fastest.
5. Carry both samples into a second room facing a different direction. Northern light skews blue through the day and a western window swings orange at sunset.
6. Ask 1 other person before you decide, and read the next section for why their answer may honestly differ from yours.

Time the test as well as staging it. Sherwin-Williams quotes the neuroscientist Bevil Conway on chromatic bias: the brain is good at stripping the bias out of natural light, but works hardest during the transitions from dawn to early morning and from dusk to dark, when the light is changing fastest. Those are the 2 worst windows in the day for judging a pair.

Which lights in one home are genuinely different from each other?

Far more different than the packaging suggests. Standard illuminant A, which represents typical domestic tungsten lighting, is a Planckian radiator at about 2856 K; illuminant D65, which represents average daylight, sits at about 6504 K. Those 2 are the everyday extremes of a normal room, and they are more than 3600 K apart.

Spectrum matters more than colour temperature here, and the number on the box does not report it. Fluorescent lamps produce irregular, peaky output, which is why 2 materials can match under an incandescent source with a smooth curve and fail under a tube. The colour rendering index is a partial guide at best: Ra is the average of only 8 test samples, R9 for saturated red is excluded from it, and the metric is a weak predictor for sources below 5000 K, which covers most warm bulbs sold for homes.

Glass changes the answer too. The indoor daylight illuminants added by colour science are the outdoor ones filtered through window glass, which strips the ultraviolet and lifts the indoor colour temperature by roughly 100 K. That matters because optical brighteners in some paints, papers and textiles fluoresce under ultraviolet, so a pair that agrees indoors can separate on a balcony. If your greys shift with the season rather than with the lamp, the way shelf colours read in summer light is the closer explanation.

Can 2 people in the same room disagree about the same pair?

Yes, and neither of them is wrong. Observer metameric failure happens because the ratio of long-wavelength to medium-wavelength cones, the sensitivity profile of each cone type, and the amount of yellowing in the lens and macular pigment all differ from person to person, so 2 samples can match for one viewer and part for the next. Colour blindness is the obvious case, but it happens between people with ordinary colour vision too.

Two more failures are worth knowing because both are easy to trigger at home:

• Field-size failure — the mix of cone types changes from the centre of the visual field outwards, so colours that match as small patches viewed straight on can separate once they cover a whole wall. Industry uses large-field matches to set tolerances for exactly this reason, and it is why a 2 in chip lies about a wall.
• Geometric failure — 2 samples can match from 1 angle and not from another. Pearlescent car paint is the textbook case, and any metallic or micaceous finish on a lamp or a frame behaves the same way.

The practical consequence is that a match tested by 1 person, on a small swatch, from 1 chair, has been tested under the 3 conditions most likely to flatter it. The same reasoning explains why a grey paint can read purple on the wall when the chip looked neutral in the shop.

From the shop

6-Piece Pine Cube and Ledge Wall Shelves

Six natural pine modules under a clear matte lacquer in warm honey pine: 2 open square boxes at 10.6 in (27 cm) and 8.3 in (21 cm) plus 4 ledges reaching 35.4 in (90 cm) joined. Each is 3.5 in (9 cm) deep on a 0.6 in (1.5 cm) board, keyhole mounted, rated up to 11 lb (5 kg) per piece.

View the 6-piece pine cube shelf set — dimensions & current price

A spectral match vs a metameric match: which should you pay for?

A spectral match, whenever it is available to you. Two surfaces with the same reflectance curve stay matched under every illuminant, every observer and every angle, because there is nothing left for a change of light to pull apart.

The paint trade already works this way: colour matches there aim at the reflectance curve rather than at a tristimulus match, and the reliable route to it is using the same pigment and base composition in both. Translated to a room, that means buying the 2 grey things from 1 maker in 1 order, or accepting that they are 2 different greys and spacing them apart. Where they cannot be bought together, put a material with no match to fail between them, which is the job the open modules in the wall shelves collection do on a grey wall.

What I'd do

Stop trying to match greys across materials, and match them across brands never. Near neutrals are the hardest colours in the house to pair, because the range of possible metameric matches is widest exactly where saturation is lowest, so a grey fabric and a grey paint are 2 curves being asked to agree by accident.

When a pair has to work, I buy both from 1 supplier, test them touching under the room lamp and again by the window on a bright morning, and skip the 2 transition hours at each end of the day when the brain is working hardest on chromatic bias. If they hold across those 2 conditions they will hold across the rest.

When a pair does not have to work, I separate the greys with something that is not trying. The 6 Houseland modules run 3.5 in (9 cm) deep on a 0.6 in (1.5 cm) board and take up to 11 lb (5 kg) each, and the finish is sanded natural pine with visible corner seams rather than a colour chasing another colour. A warm break between 2 cool neutrals reads as a choice, which is the same trick as handling brass under warm light: pick the light you live in and let the pair be judged there.

Shop wall shelves from Houseland

-20%9-Piece Pine Hexagon Wall Shelf Setfrom $165.50$206.88-20%6-Piece Rounded Pine Box Wall Shelvesfrom $160.78$200.96-20%6-Piece Staggered Pine Wall Shelvesfrom $151.15$188.94-20%6-Piece Pine Honeycomb Wall Shelvesfrom $146.43$183.03-20%5-Piece Matte Black Wall Shelf Setfrom $155.59$194.49-20%9-Piece Matte Black Hexagon Wall Shelvesfrom $174.58$218.23

See all wall shelves →