chromatic

A comprehensive color science library for Dart. Color space conversions, perceptual color difference (Delta E), chromatic adaptation, and more.

Port of Python's colormath with identical math, validated against 690+ Python-generated test fixtures and the Sharma 2005 CIEDE2000 reference dataset.

🎨 Live demo: https://enraio.github.io/chromatic/

Features

10 color spaces: sRGB, Adobe RGB, Apple RGB, BT.2020, XYZ, xyY, Lab, LCH(ab), Luv, LCH(uv), HSL, HSV, CMY, CMYK, IPT
4 Delta E formulas: CIE76, CIE94 (graphic arts + textiles), CIEDE2000 (the gold standard), CMC l:c
Chromatic adaptation: Bradford, von Kries, XYZ scaling — convert between any illuminant (D50, D65, A, C, F-series, etc.)
Proper sRGB EOTF: piecewise transfer function (not simple power gamma)
Proper BT.2020 transfer: ITU-R BT.2020 piecewise OETF
Hex parsing: #ff8040, ff8040, #f84 shorthand
All immutable: every color type is a pure value class

Validated against Python colormath

Every conversion and Delta E formula is tested against 690+ fixtures generated from Python colormath, plus the 34-pair Sharma 2005 reference dataset for CIEDE2000. Max error against Sharma's published values: 4.9e-5.

Usage

import 'package:chromatic/chromatic.dart';

// Parse and convert
final red = SRgbColor.fromHex('#ff0000');
final lab = sRgbToLab(red);         // Lab(53.24, 80.09, 67.20)
final lch = labToLchAb(lab);        // LCH(53.24, 104.55, 39.99)
print(red.hex);                      // #ff0000

// Perceptual distance — use CIEDE2000
final blue = sRgbToLab(SRgbColor(0.0, 0.0, 1.0));
print(deltaE00(lab, blue));          // ~52.88 (clearly different)
print(deltaE00(lab, lab));           // 0.0

// Chromatic adaptation between illuminants
final xyzC = XyzColor(0.5, 0.4, 0.1, illuminant: 'c');
final xyzD65 = adaptXyzColor(xyzC, targetIlluminant: 'd65');

// Or directly
final (x, y, z) = chromaticAdapt(
  0.5, 0.4, 0.1,
  sourceIlluminant: 'c',
  targetIlluminant: 'd65',
  method: 'bradford',      // or 'von_kries', 'xyz_scaling'
);

Performance vs Python colormath

Benchmarks on M-series Mac, 100K iterations each:

Operation	chromatic (Dart)	colormath (Python)	Speedup
sRGB → Lab	4,072,988 ops/s	87,224 ops/s	46.7×
sRGB → XYZ	6,002,041 ops/s	130,585 ops/s	46.0×
sRGB → HSL	5,448,700 ops/s	201,759 ops/s	27.0×
sRGB → HSV	6,059,504 ops/s	200,664 ops/s	30.2×
sRGB → LCH	3,432,651 ops/s	70,992 ops/s	48.4×
sRGB → Luv	5,288,207 ops/s	80,895 ops/s	65.4×
Lab → sRGB	1,502,720 ops/s	28,874 ops/s	52.0×
Delta E 1976	99,304,866 ops/s	272,325 ops/s	364.7×
Delta E 1994	27,173,913 ops/s	43,891 ops/s	619.1×
Delta E CIEDE2000	4,507,144 ops/s	10,700 ops/s	421.2×
Delta E CMC	16,784,156 ops/s	33,854 ops/s	495.7×
Bradford adaptation	2,950,027 ops/s	48,610 ops/s	60.7×
Batch (100K colors → Lab)	16.3 ms	1036.3 ms	63.6×

Reproduce:

dart run benchmark/benchmark.dart
python3.11 benchmark/benchmark.py

Supported color spaces

Space	Forward	Inverse
sRGB	`rgbToXyz`	`xyzToSRgb`
Adobe RGB	`rgbToXyz`	`xyzToAdobeRgb`
Apple RGB	`rgbToXyz`	`xyzToAppleRgb`
BT.2020	`rgbToXyz`	`xyzToBt2020`
XYZ	`xyzToLab`, `xyzToLuv`, `xyzToXyy`, `xyzToIpt`	←
xyY	`xyyToXyz`	`xyzToXyy`
Lab	`labToLchAb`, `labToXyz`	←
LCH(ab)	`lchAbToLab`	`labToLchAb`
Luv	`luvToLchUv`, `luvToXyz`	←
LCH(uv)	`lchUvToLuv`	`luvToLchUv`
HSL	`rgbToHsl`	`hslToRgb`
HSV	`rgbToHsv`	`hsvToRgb`
CMY	`rgbToCmy`	`cmyToRgb`
CMYK	`cmyToCmyk`	`cmykToCmy`
IPT	`xyzToIpt`	`iptToXyz`

Convenience shortcuts: sRgbToLab, sRgbToLchAb, sRgbToHsl, sRgbToHsv, labToSRgb, lchAbToSRgb.

Delta E reference

Illuminants supported

Observer	Illuminants
2°	A, B, C, D50, D55, D65, D75, E, F2, F7, F11
10°	D50, D55, D65, D75

Running tests

# Regenerate fixtures from Python colormath (one-time)
pip3.11 install colormath numpy
python3.11 tool/generate_fixtures.py

# Run all Dart tests
dart test

License

BSD-3-Clause (matches Python colormath).