Short answer
Because ΔE formulas model average judgements of two large, flat, adjacent samples on a neutral background under standard light. Change the size, the surround, the texture, the gloss, the light or the distance between the samples and the colour that looks closest can change too. The nearest colour by number is the best starting candidate, not a guaranteed visual match.
CIE94 and CIEDE2000 were fitted to visual judgements made under defined reference conditions for the light, the viewing arrangement and the samples themselves — broadly, daylight-like illumination and flat, uniform samples seen together against a neutral background. The CIE reports state these conditions and provide parametric factors for departures from them precisely because judgements change outside them; most real comparisons depart from them in several ways at once. A ΔE computed for a small screen swatch against a paint chip on a white wall is being applied far from where it was calibrated.
The same colour looks lighter on a dark surround and darker on a light one, and takes on a tint opposite to a coloured surround (simultaneous contrast). Two samples judged on different backgrounds can therefore look further apart, or closer, than their ΔE says. Size matters too: large areas look more saturated and are judged with more of the retina, which is why the 10° observer exists and why a colour chosen from a small chip often looks stronger on a wall. Separation matters most of all: side-by-side comparison is far more sensitive than comparing samples with a gap between them or from memory, so a pair that fails a tolerance at the edge may be invisible on opposite sides of a room.
ΔE assumes flat, homogeneous samples. Textiles, stone, textured paint and printed halftones mix light and shadow at a small scale, and the eye integrates them differently from an instrument's aperture. A glossy and a matt sample with equal measured colour (in SCI) look different, and equal SCE values can still look different at other angles. Metallic, pearlescent and fluorescent materials depart furthest: their appearance depends on angle and on UV in the light. For these, a 'nearest colour' match by number across different materials is often a poor visual match even at small ΔE.
Colourwise's matching tools use CIEDE2000 to rank candidates, which is the best general-purpose choice. Treat the top few results as a shortlist: compare them physically, side by side, on the real surface and in the real light, with the surrounding colours that will be present. If two candidates are within about one ΔE00 of the target, the choice between them should be made by eye, and if the materials differ (a fabric against a paint, a print against a screen), expect to adjust. Where the light will change, check candidates under more than one light for metamerism.
Why: The comparison was made on screen or as a small chip; size, surround and room light change appearance.
Fix: Test the top candidates as large samples on the wall, beside the fixed materials, in the room's light.
Why: Different materials: texture, gloss and spectral curves differ, so the match may be metameric.
Fix: Compare under several lights and accept that cross-material matches need visual adjustment.
Why: The difference lies in hue or near-neutral chroma, to which the eye is especially sensitive.
Fix: Choose by eye among close candidates; check the hue component of the difference.
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.
StandardStrong evidence
CIE colour-difference formulas are defined together with reference viewing conditions and include parametric factors to adjust for conditions of use that differ from them.
Source: CIE 116-1995 Industrial colour-difference evaluation; CIE 142:2001 Improvement to industrial colour-difference evaluation
FactStrong evidence
Human sensitivity to colour differences varies with saturation, hue and lightness — lowest for highly saturated colours and highest around mid lightness — which weighted formulas such as CIEDE2000 model only on average.
Source: Precise Color Communication; The development of the CIE 2000 colour-difference formula: CIEDE2000
FactModerate evidence
A colour's appearance depends on its surround and size, so two samples with a small measured difference can look further apart when viewed on different backgrounds or at different scales.
Caveat: The magnitude of the effect depends strongly on the specific colours and viewing arrangement.
Source: CIE 015:2018 Colorimetry, 4th edition; Precise Color Communication
Colourwise interpretationModerate evidence
A nearest-colour result by ΔE00 is best treated as a shortlist to be confirmed physically in the real surround, material and light, particularly when the target and candidate are different materials.
Based on: Follows from the reference conditions of the CIE formulas and the known effects of surround, size, texture and illuminant on appearance.
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.