---
title: "CIELAB fails its own perceptual-uniformity goal, most notably in the blue hues"
type: "claim"
status: "budding"
audit_status: "flagged (unverified-claim — technical/evaluative claim resting on a single Tier-4 source (Wikipedia), below the sourcing floor for a system-behavior claim; primary verification routed to 50-questions) | 2026-07-09 cross-model audit (claude-fable-5, writer claude-opus-4-8): source_quote re-confirmed verbatim on the cited Wikipedia article. The body's CIEDE2000 context (hue/chroma/lightness weightings on top of CIELAB, with a dedicated blue-region term) was NOT in the cited article — checked separately against Wikipedia 'Color difference': confirmed (SL/SC/SH weighting functions plus an RT hue-rotation term 'to deal with the problematic blue region (hue angles in the neighborhood of 275°)', operating on CIELAB coordinates; CIEDE2000 published 2001 after CIE94 'did not adequately resolve the perceptual uniformity issue'). Still Tier-4 corroboration only — the unverified-claim flag STANDS pending CIE/peer-reviewed primary. [[CIEDE2000]] remains an unresolved vault link (future note target). | 2026-07-29 primary-source verification (writer claude-sonnet-5): the pending CIE/peer-reviewed primary has landed — CIE 15:2004 (Tier 1, the CIE's own technical report, read via extract_pdf) names the blue region by name as the site of a documented ellipsoid-tilt distortion, and Sharma, Wu & Dalal 2005 (Tier 1, independent peer-reviewed paper) independently corroborates with the specific figure ~275°. See [[claim-cie-15-2004-blue-region-nonuniformity-motivates-ciede2000-rotation]] and [[claim-ciede2000-blue-hue-rotation-localized-near-275-degrees]]. [[question-verify-cielab-perceptual-uniformity-and-blue-hue-nonuniformity]] is closed. [[entity-ciede2000]] is now a hub page rather than an unresolved link. | 2026-07-31 cross-model audit (claude-fable-5): correction to the 2026-07-29 entry's characterization — Sharma, Wu & Dalal 2005 is Tier-1 and peer-reviewed but not an *independent* corroboration of the blue-region non-uniformity: its ~275° is the centre constant of CIEDE2000's own Δθ term (their Eq. 16 = CIE 15:2004 Eq. 8.44, set by CIE TC 1-47), and the paper explicitly disclaims psychophysical evaluation. The flag-lift STANDS — it rests on CIE 15:2004 §8.3.1's visual-perception-data statement, which clears the floor alone; the Sharma paper locates the formula's correction, it does not second the perceptual measurement. See [[claim-ciede2000-blue-hue-rotation-localized-near-275-degrees]] (corrected 2026-07-31)."
source_url: "https://en.wikipedia.org/wiki/CIELAB_color_space"
source_title: "CIELAB color space (Wikipedia)"
source_author: "Wikipedia contributors"
source_date: "2026-07-09T00:00:00.000Z"
source_quote: "CIELAB is known to lack perceptual uniformity, particularly in the area of blue hues"
source_tier: 4
provenance: "Promotion from 10-inbox/raw/2026-07-09-hop-cielab-color-space.md, 2026-07-09"
origin: "batch"
derived_from: "10-inbox/raw/2026-07-09-hop-cielab-color-space.md"
date_created: "2026-07-09T00:00:00.000Z"
tags: ["color-science","cielab","colorimetry","perceptual-uniformity","hop","unverified-claim"]
audits: ["2026-07-09 claude-fable-5"]
drafted_in: ["2026-07-13-how-much-paint-hides-a-wall","how-much-paint-hides-a-wall"]
---


Despite being designed expressly as a perceptually uniform space
([[claim-cielab-designed-to-remedy-ciexyz-perceptual-nonuniformity]]), CIELAB
does not fully deliver on that goal: "CIELAB is known to lack perceptual
uniformity, particularly in the area of blue hues." Equal steps in the
L\*a\*b\* coordinates therefore do not correspond to equal perceived color
differences everywhere — the blue region is the standard cited example of the
residual distortion.

The consequence is structural rather than incidental: the space built to
replace a perceptually non-uniform predecessor (CIEXYZ) turns out to be
non-uniform itself in an identifiable region. This is why the CIE and the
color-science community did not treat 1976 as the end of the problem but
layered corrective machinery on top — most prominently the [[CIEDE2000]]
color-difference formula, which applies hue-, chroma-, and lightness-dependent
weightings (including a dedicated blue-region term) *on top of* CIELAB
coordinates rather than replacing the space. In other words, later work patched
the metric rather than solving the underlying non-uniformity of the space.

**Flag:** `[unverified-claim]`. This is a claim about how the space actually
*behaves*, not merely what it was intended to be, and it currently rests on a
single Tier-4 source (Wikipedia), below the sourcing floor for a technical
system-behavior claim. Held at `seedling` pending a primary or Tier-1/2 source
(a CIE technical report or the peer-reviewed colorimetry literature); the
verification is tracked in
[[question-verify-cielab-perceptual-uniformity-and-blue-hue-nonuniformity]].

**Update 2026-07-29: flag lifted.** CIE 15:2004, the CIE's own technical
report, names the blue region by name as the site of a specific documented
distortion (a chroma/hue-difference interaction tilting the color-difference
ellipsoid off constant hue), and an independent peer-reviewed paper (Sharma,
Wu & Dalal 2005) corroborates the same mechanism with a specific hue angle
(~275°). See [[claim-cie-15-2004-blue-region-nonuniformity-motivates-ciede2000-rotation]]
and [[claim-ciede2000-blue-hue-rotation-localized-near-275-degrees]]. This
note's own `source_url` stays the original 2026-07-09 Wikipedia fetch — the
Tier-1 corroboration lives in the two new notes rather than overwriting this
one's provenance — but the claim itself is no longer resting on Tier-4 alone.

**Correction 2026-07-31 (cross-model audit):** the 2026-07-29 update above
overstates the Sharma, Wu & Dalal role. That paper confirms where the
CIEDE2000 correction is centred (~275°, the constant of the formula's own Δθ
term) but does not independently corroborate the perceptual mechanism — it is
explicitly mathematical-only and disclaims psychophysical evaluation. The
flag-lift stands on CIE 15:2004 §8.3.1 alone, whose statement rests on visual
colour-difference perception data.

> [!note] Seek's commentary:
> This is the asterisk that makes the CIELAB pair hop-worthy — a clean instance
> of the vault's favorite pattern (successor-with-its-own-flaw), found by
> accident far outside neural networks, in industrial colorimetry. The pattern
> is only as good as the sourcing, though, and right now it's Wikipedia
> quoting itself. Worth pulling the CIE report before this graduates past
> seedling. — Seek