How to match a hex colour to a paint colour

You have a hex value. It came off a product photograph or a palette generator, and it looks exactly right on your monitor. The question is what to hand the paint counter. A hex is a genuinely useful starting point and a genuinely bad instruction, and the reasons are physical rather than anyone failing to try.

What a hex actually is

A hex code is three numbers describing how hard to drive the red, green and blue emitters of a display. It is an additive model: the three lights add together, and all of them at full is white. It describes an instruction to a screen, not a material. As one designer summarised it, “Hex codes and RGB codes are for inks and screens”. Another was shorter about handing one to a paint mixer: “Hex no worko”.

Paint works the other way round. It is subtractive. Pigments absorb parts of the spectrum and reflect the rest, and what you see is whatever is left over after that subtraction, bounced off a surface, under whatever light happens to be in the room. Two systems, two different physics. There is no lossless conversion between them, only an approximation with a lot of context stripped out.

Why it does not survive the trip to a wall

Four things change between the hex on your screen and the finished wall.

Pigment. Two paints can look identical on a chip and be mixed from entirely different colourants. That difference is invisible until the light changes, and then it is not.

Sheen. Matt scatters light in every direction and reads flat and deep. Eggshell and gloss reflect more of it directly, so the same tint appears lighter, shinier and far more variable across the day. Sheen is not a finish decision made after the colour decision; it is part of the colour.

Substrate. New plaster, lining paper, previously painted timber and MDF all absorb differently. The primer underneath changes the final reading, particularly with the thin, chalky, low-opacity colours designers like most.

Light. A north-facing room in Edinburgh in February and a south-facing room in June are not the same environment. Add a 2700K bulb in the evening and the wall you approved at noon is a different colour by seven.

Metamerism, explained plainly

Metamerism is the reason two colours can match under one light and refuse to match under another. Colour, as you experience it, is your eye responding to the mix of wavelengths reaching it. Two surfaces made from different pigments can produce the same response under daylight, because their differences happen to cancel out at those wavelengths, and then come apart completely under a warm lamp that contains a different mix of wavelengths.

The practical consequence is the one everybody has lived through: the joinery that matched the wall in the showroom and does not match it in the hall. Nobody got it wrong. The two finishes were metameric. It is also why you check a paint and a fabric together, in the room, under both daylight and the artificial light the room will really use.

Can you colour match a fabric swatch?

Partly, and less reliably than people expect. The spectrophotometer at a paint counter is designed to read a flat, opaque, matt surface. A textile is none of those. Weave and pile scatter light in directions the instrument is not expecting, a slub or a print gives it several colours at once to average, and a thin or loosely woven material lets whatever is behind it influence the reading. Sheen on a satin or a velvet throws it further. A designer relayed the response from a paint counter directly: that you “can’t color match from fabric because the machine won’t pick it up”.

The workaround is the old one. Take the swatch to the fan deck, hold candidates against it in daylight, pick three, and sample them properly. A trained eye comparing physical samples still beats a scan of an uncooperative surface.

The workflow that actually works

1. Use the hex to shortlist. Get to the right family and the right rough value fast. Three or four candidates, not one. This is the part digital colour is good at.

2. Order the physical samples. Peel-and-stick sample sheets or painted boards, never a small patch straight onto the wall, because the existing colour around it will distort your judgement. Use two coats and the sheen you intend to specify.

3. View them in the actual room, at three times of day. Morning, mid-afternoon, and after dark under the lighting the room will really have. Move the boards to different walls. A colour that survives all three is a specification; a colour that only works at noon is a complaint waiting to happen.

4. Then decide. On the physical sample, in the room, with the fabric and the flooring present. Never from the screen, and never from the hex.

Brand fan decks remain the reliable route through all of this, and they are not a legacy habit. A fan deck is a set of physical drawdowns made with the actual pigments in an actual finish, produced under controlled conditions. It is a sample library that fits in a bag, and it is the only colour reference in the process that is not an approximation of something else. Keep it current, and keep it filed with the rest of your material and sample library.

Where digital colour genuinely helps

None of this makes hex values useless. It makes them a retrieval tool rather than a decision tool, and retrieval is a real problem worth solving.

You remember a chair by its colour and nothing else. You need everything in the library within reach of a chalky green, quickly, so you can shortlist. That is what searching a saved library by colour or hex is for, and it works precisely because it is not being asked to decide anything. It narrows a hundred saved products down to five worth ordering samples of.

The same logic runs into the board. When you are composing a moodboard from what you have saved, digital colour is doing grouping work, showing whether a scheme hangs together. Sign-off still happens with samples in the room. For the wider method, there is a piece on building a colour palette that survives a whole project.

Screen colour never replaces a physical sample. It just makes sure you order the right three.

Can you colour match a fabric swatch at a paint counter?

Sometimes, and less reliably than with a flat painted chip. Weave, pile and sheen scatter light, and thin or loose material lets the backing influence the reading. One paint counter put it plainly, that you cannot colour match from fabric because the machine will not pick it up. Take the swatch anyway, but expect to choose by eye from a fan deck rather than trust a scan.

Will a hex code from a product photo give me the right paint?

It will give you the right family. A product photograph has been through a camera, a retoucher and your screen, so the hex describes the image rather than the object. Use it to shortlist three or four candidates, then order samples.

Why does the same paint look different in two rooms?

Light. Daylight direction, bulb colour temperature, and the colours bouncing off floors and neighbouring walls all change what reaches your eye. This is why a sample has to be viewed in the actual room rather than in the shop or on a screen.

How many times of day should I look at a sample?

Three. Morning, mid-afternoon and after dark under the lighting the room will actually have. A colour that holds up in all three is a safe specification; one that only works at noon will be regretted in November.

Does sheen change the colour?

It changes the colour you perceive, which is what matters. Higher sheen reflects more light specularly, so the same tint reads lighter and more variable at eggshell or gloss than it does at matt. Always sample in the finish you intend to use.