Why does the same colour look darker and richer in gloss than in matt?

Surface reflection is the same few per cent, redistributed

Fresnel's equations give the fraction of light reflected where two media meet. For a binder, lacquer, glaze or plastic with a refractive index near 1.5, about 4% is reflected at the surface at normal incidence, rising steeply at grazing angles. That light never meets the pigment, so it carries the colour of the light source rather than of the material. Gloss does not change how much is reflected; it changes where it goes. A smooth surface concentrates it into the mirror direction, producing a sharp highlight and an image of the room. A rough, matt surface spreads the same few per cent over the whole hemisphere, so a veil of white is laid over the body colour wherever you stand.

The table on this page gives the normal-incidence figure for several refractive indices: water about 2%, glass-like binders about 4%, polyester about 5%, and the titanium-dioxide pigment inside white paint around 20%.

Dark colours change most; pale colours hardly at all

Adding a few per cent of white light matters in proportion to how little light the colour itself returns. A navy that reflects around 3% of light has its lightness pushed up markedly by a 4% veil; a pale grey that reflects 70% barely notices it. In Colourwise's finish model a dark navy moves from about L* 30 in matt to about L* 17 in gloss seen away from the highlight — a difference of more than ten CIELAB units — while a pale warm grey changes by less than one. That is why black and deep jewel colours are specified in gloss when depth matters, why deep colours in flat paint read chalky, and why the finish choice for a white room is mostly about cleanability rather than colour.

The highlight is where gloss looks worst

The price of gloss is that the undiluted colour is only visible away from the mirror direction. In the highlight, the concentrated surface reflection swamps the colour entirely: a glossy dark cabinet under a row of downlights shows white streaks, and a glossy wall opposite a window shows the window. A gloss surface also shows every ripple in the substrate, because the reflected image distorts across it, and it reveals orange peel, brush marks and poor filling that a matt finish would disguise. Satin and eggshell finishes are compromises: some of the surface reflection is still scattered, so dark colours lift a little, but highlights are softened and imperfections hidden.

How gloss is measured, and why it is not colour

Gloss is measured with a glossmeter that shines light at a fixed angle and reads the mirror reflection against a polished glass reference defined as 100 gloss units. ISO 2813 and ASTM D523 use three geometries: 60° for most surfaces, 20° to separate high-gloss surfaces (above about 70 units at 60°) and 85° to separate matt ones (below about 10 units at 60°). Gloss units are not a percentage of light and do not map onto the paint trade's words 'eggshell' or 'satin', which vary between manufacturers. Because gloss changes measured colour, spectrophotometers report specular-included and specular-excluded readings; comparing a gloss and a matt sample of the same paint is only meaningful with the geometry stated.

Gloss changes with wear, cleaning and time

Gloss is a property of the top micrometres, so it moves more easily than pigment colour. Matt surfaces polish up where they are rubbed — handrails, the edges of worktops, the seat of a sofa — and those shiny tracks look darker and more saturated than the surrounding matt area. Gloss surfaces dull as they are scratched or weathered: outdoor coatings lose gloss before they fade, and the dulled surface looks paler and chalkier even if the pigment is unchanged. A colour complaint about an old surface is often a gloss change; wiping a damp cloth across it is a quick test, because a wet film temporarily restores the smooth surface and the colour returns.

The data

What goes wrong, and what to do

How to check it yourself

1. Hold the sample so a window reflects in it, then tilt it until the reflection moves away; the colour you see away from the highlight is the body colour.
2. Compare matt and gloss versions of the same colour side by side at the same angle, away from the light.
3. Wipe a damp cloth across a dull old surface: if the colour deepens while wet, the change is mostly gloss.
4. Try the finish simulator with your colour to see the direction and size of the effect.

What the evidence says

Each statement is labelled by kind — established fact, a standard’s requirement, observed market data, a convention, or Colourwise’s own interpretation or analysis — with the strength of the evidence behind it.

• FactModerate evidence

  At normal incidence the fraction of light reflected at an interface between air and a material of refractive index n is ((n − 1)/(n + 1))², about 4% for n = 1.5.

  Source: HyperPhysics (optics, atmospheric optics and vision pages)

• StandardModerate evidence

  Specular gloss is measured at 20°, 60° or 85° under ISO 2813 and ASTM D523, in gloss units referenced to a polished glass standard, with 60° readings above about 70 units remeasured at 20° and below about 10 units at 85°.

  Caveat: Summarised from an instrument maker's explainer rather than the standards' text, which Colourwise has not reproduced.

  Source: What is gloss and how is it measured?

• Colourwise analysisModerate evidence

  In Colourwise's surface-finish model a dark navy body colour is more than ten CIELAB lightness units darker in gloss (seen away from the highlight) than in matt, while a pale grey changes by less than one ΔE00 unit.

  Based on: Calculated with src/lib/colour/surface-finish.ts: 4% Fresnel surface reflection (n = 1.5) distributed with illustrative diffuse shares for each finish, added to the body colour in linear sRGB and converted to CIELAB.

  Caveat: The diffuse shares for matt, satin and gloss are illustrative, not measured; real finishes and lighting vary.

  Source: HyperPhysics (optics, atmospheric optics and vision pages); CIE (International Commission on Illumination) publications

• Colourwise interpretationLimited evidence

  Many apparent colour changes on ageing surfaces are gloss changes — polishing of matt areas or dulling of gloss ones — rather than changes to the pigment.

  Based on: Follows from the surface-reflection mechanism above: the same pigment reads darker when the surface is smoother and paler when it is rougher.

  Caveat: Fading and yellowing also occur; a damp-cloth test helps tell them apart.

Related questions

Why do black cars and pianos look blacker than black paint on a wall?

They are high-gloss: the surface reflection is confined to sharp reflections of the room, so between them you see the pigment with almost nothing added. A flat black wall has a few per cent of white spread over its whole area, lifting it towards dark grey.

Does gloss paint contain a different pigment from matt paint?

Usually the same pigments are used; the difference is the binder content and particle packing that make the dried surface smooth or rough. That is why the same colour name can look noticeably different across sheens from one manufacturer.

Is a gloss meter reading a percentage of reflected light?

No. Gloss units compare the mirror reflection with a polished glass reference set to 100 at that angle. Mirrors and metals can read well above 100, and the number does not tell you total reflectance or colour.

Related topics

Go further

Sources

• HyperPhysics (optics, atmospheric optics and vision pages) — Georgia State University, Department of Physics and Astronomy (Academic)
• What is gloss and how is it measured? — Rhopoint Instruments (Manufacturer technical documentation)
• CIE (International Commission on Illumination) publications — CIE (Official standard or government)
• List of refractive indices — Wikipedia (Reference work)
• Colourwise editorial synthesis — Colourwise (Colourwise editorial)

Reviewed 29 September 2026. Colourwise summarises its sources in its own words and does not reproduce standards text or proprietary colour data. Spotted an error? Tell us.