Nail Science

Metamerism: why two shades can match in one light and differ in another

Articles saved on this device
Nail polish bottles displayed beside framed photographs.
Thematic photograph. cottonbro studio / Pexels · Pexels licence

Metamerism is the effect in which colours with different light spectra look the same to the eye. For two surfaces it means they can match under one light and separate under another, because the eye reduces light to three kinds of response and a change of illuminant can pull apart surfaces that gave equal responses before. It offers one explanation for why two polishes that match under shop lighting can differ at home, but the sources define the effect and do not measure how often it happens with polishes. [1]

KEY TAKEAWAYS

  • Metamerism occurs when different spectra produce the same visual response. Two surfaces can match under one illuminant and become distinguishable under another. [1]
  • The cited optical study explains the role of three cone-response dimensions; it does not measure how often nail-polish pairs are metameric. [1]
  • The DOE shows that apparently similar white lights can have different spectral power distributions. A shared colour-temperature label does not make their spectra identical. [2]
  • General CRI, Ra, summarises fidelity across eight standard samples. It is not a test of whether two particular nail shades match. [2][3]
  • A visual mismatch alone does not establish metamerism. The spectral and viewing conditions of the specific pair matter. [1]

What metamerism is

In a 2006 paper in the Journal of the Optical Society of America A, metamerism is introduced as lights looking the same to the eye while differing in their spectral radiant power distributions across the visible spectrum. The same paper applies the idea to surfaces: surfaces whose spectral reflectances differ can produce equal sensor responses under some illuminant, and the effect becomes a problem when the light they reflect becomes distinguishable after the illuminant changes. In plain terms (an editorial gloss), a spectrum is how much light there is at each wavelength, and a spectral reflectance is the share of light a surface sends back at each wavelength. [1] Related reading: light absorption and scattering.

The reason is a limit in the sensor. The paper counts the sensor system's degrees of freedom as three for the cones of the normal human eye or a typical camera, so, in plain terms, many different spectra can collapse into the same three responses. As a rough analogy of our own, two baskets of different goods can cost the same under one price list and different amounts under another. [1]

Why the light source matters, even when both lights look white

The definition starts with lights: two lights can look alike while their spectra differ. A US Department of Energy fact sheet on LED colour makes the point with a chart. Its Figure 1 caption says that, because the human visual system is trichromatic, white light of equal appearance can be created with different spectral power distributions, and the chart shows four light sources (halogen, fluorescent and two kinds of LED) all producing 2700 K white light with a CRI above 80. It follows (our reading) that a shop and a home can both look white-lit and still deliver different spectra to a polish. [2]

A lamp's colour temperature does not settle the question. The fact sheet explains that correlated colour temperature, given in kelvin, characterises the colour of the emitted light, not the colour of illuminated objects, and that two sources with the same colour temperature can look different, one appearing greenish and the other pinkish. [2]

Hypothetical example: two workstations both use lamps labelled warm white. Their similar apparent light colour does not establish that the lamps illuminate a pair of swatches with the same spectral distribution. The label is a useful description of the lamp, not a guarantee of the shade match. [2]

What colour temperature and colour rendering figures do and do not tell you

The general colour rendering index, Ra, summarises a source’s fidelity to a reference using eight standard test colours. DOE distinguishes this fidelity question from whether colours look appealing or can be readily distinguished. Image Engineering’s technical note also distinguishes general Ra from special indices for individual colours. A high general score is therefore not a direct result for every pigment or polish. [2][3]

Keep the two labels separate: correlated colour temperature describes the apparent colour of the emitted light; general CRI summarises rendering against reference samples. Neither supplies a matched spectral comparison of two particular nail shades. This is the practical limit of using a lamp label to predict a shade match. [2][3]

What three colour and lighting terms describe, and what each leaves open
TermWhat it describesWhat it does not tell you
Metamerism [1]Colours or lights that look the same while their spectra differ; surfaces can become distinguishable when the illuminant changes.How often a given pair of polishes separates; the paper's frequency estimates are for natural scenes.
Correlated colour temperature (CCT) [2]The colour of the emitted light on a scale in kelvin.The colour of illuminated objects; two sources with the same CCT can look different, one greenish and one pinkish.
General colour rendering index (Ra) [2][3]Fidelity across eight standard samples relative to a reference illuminant.A matched comparison of the exact two nail shades, or a complete description of appeal and discrimination.

Comparing two shades when metamerism is possible

For an editorial comparison, record the actual light used and compare the same pair under a second, different light. Keep the two swatches and viewing arrangement otherwise as consistent as possible. This is a simple comparison plan derived from the illuminant-dependent definition, not a diagnostic test that proves metamerism. [1] Related reading: building a swatch catalogue that stays traceable.

Hypothetical example: two beige swatches look alike at a counter but differ beside a window. That observation is consistent with illuminant-dependent metamerism, but it does not prove that their spectral reflectances are the cause. Describing what was seen and under which lights is more defensible than labelling the pair metameric from a single impression. [1] Related reading: gloss and colour as different parts of appearance, why swatch photos differ with the image setup.

The optical paper concerns natural scenes, the DOE fact sheet explains lighting, and the Image Engineering note describes a rendering index. None reports a matched experiment on nail-polish pairs. The account explains a possible mechanism without assigning it a frequency or a confirmed diagnosis for any pictured swatch. [1][2][3]

Article Sources

  1. Journal of the Optical Society of America A 2006;23(10):2359-2372 (Foster, Amano, Nascimento and Foster), via PubMed Central - Frequency of Metamerism in Natural Scenes

    Primary source

    'Frequency of Metamerism in Natural Scenes' (J Opt Soc Am A 2006;23(10):2359-2372), Introduction: metamerism as lights appearing the same to the eye with different spectral radiant power distributions; surfaces with different spectral reflectances producing equal sensor responses with some illuminant; the problem arising when reflected lights become distinguishable with an illuminant change; three degrees of freedom for the cones of the normal human eye or a typical camera. The abstract's frequency findings (50 hyperspectral images of natural scenes) are natural-scene results and are not used.

  2. US Department of Energy, Building Technologies Program - LED Color Characteristics (Solid-State Lighting Technology Fact Sheet)

    Primary source

    Pages 1-2 of 'LED Color Characteristics', Solid-State Lighting Technology Fact Sheet, Building Technologies Program, US Department of Energy (no date printed on these pages): Figure 1 and its caption (white light of equal appearance from different spectral power distributions; four sources, labelled halogen, fluorescent, LED (RGB) and LED (PC), at 2700 K with CRI above 80); the Color Temperature section (CCT characterises the emitted light, not the colour of illuminated objects; two sources with the same CCT can look different, one greenish and one pinkish); the Color Rendition and Color Rendition Metrics sections (CRI as a measure of fidelity that does not address appeal or discrimination; eight standard colour samples).

  3. Image Engineering - Color Rendering Index (CRI): Understanding Light Quality

    Commercial source material

    Technical note 'Color Rendering Index (CRI): Understanding Light Quality' (undated, no author shown): the definition of CRI against a reference illuminant, the Ra of up to 100 statement, and the criticisms that only 8 colour samples were compared and that the CRI is not correlated to human vision.

  4. X-Rite - The Munsell System: History and Color Solutions

    Commercial source material

    History section of 'The Munsell System: History and Color Solutions' (posted 17 October 2022, modified 4 November 2025): Munsell painting scenery by day and noticing that the colours painted at night did not match those painted during the day.

    Source date:

Products and resources

A NailMarket department for display tips used to organise shade samples. A catalogue category only; no stock or colour-accuracy claim is implied. NailMarket nail colour charts

Brand resource only; no technical, launch, testing or certification claim is made. BELOM equipment project