CIELAB (1976) was designed as a perceptually uniform color space to remedy CIEXYZ's perceptual non-uniformity

CIELAB (also written L*a*b*), the color space published by the International Commission on Illumination (CIE) in 1976, was created to give the earlier CIEXYZ space a perceptually uniform successor: "CIELAB was intended as a perceptually uniform space, where a given numerical change corresponds to a similar perceived change in color." The design goal is that equal Euclidean distances in the L*a*b* coordinates should correspond to roughly equal perceived color differences — the property CIEXYZ notably lacks, where a fixed numerical step can be visually negligible in one region and obvious in another.

This is the standard motivation for a whole family of transforms that re-map a device- or tristimulus-based space into one shaped to human perception. CIELAB's sibling CIELUV was recommended by the CIE in the same year for the same reason, and the persistence of the underlying problem is what later color-difference work — CIEDE2000 — was built to address.

The claim here is definitional/historical (what CIELAB was for), which the sourcing floor allows to rest on a Tier 3–4 encyclopedia. It was held at seedling because it rested on a single Tier-4 source; the primary home is the original CIE 15 recommendation, and confirming the design-intent wording against it was tracked in question-verify-cielab-perceptual-uniformity-and-blue-hue-nonuniformity. The companion claim — that CIELAB does not actually achieve this goal — is claim-cielab-lacks-perceptual-uniformity-in-blue-hues.

Update 2026-07-29: the tracked verification is closed. CIE 15:2004, Technical Report: Colorimetry (3rd edition), confirms the design intent in the CIE's own primary wording — see claim-cie-15-2004-recommends-cielab-for-perceptual-uniformity, which now carries the Tier-1 citation this note was waiting on. This note's own source_url is left as the original 2026-07-09 Wikipedia fetch (per the vault's one-source-per-note discipline); the primary corroboration lives in the new note rather than overwriting this one's provenance.