Unicolour

Unicolour is the most comprehensive .NET library for working with colour:

• Colour space conversion
• Colour mixing / colour interpolation
• Colour blending
• Colour difference / colour distance
• Colour gamut mapping
• Colour chromaticity
• Colour temperature
• Wavelength attributes
• Pigments for natural paint mixing
• ICC profiles for CMYK conversion

Written in C# with zero dependencies and fully cross-platform compatible.

Targets .NET Standard 2.0 for use in .NET 5.0+, .NET Core 2.0+ and .NET Framework 4.6.1+ applications.

See a live demo in the browser — a colour picker for any colour space, including print (CMYK · ICC profiles) and paint (RYB · pigments) — made with Unicolour.

Contents

1. 🧭 Overview
2. 🔆 Installation
3. ⚡ Quickstart
4. 🌈 Features
5. 💡 Configuration
6. ✨ Examples
7. 🔮 Datasets
8. 🥽 Experimental

🧭 Overview

A Unicolour encapsulates a single colour and its representation across 40 colour spaces. It can be used to mix and compare colours, and offers many useful features for working with colour.

Unicolour pink = new("#FF1493");
Console.WriteLine(pink.Oklab); // 0.65 +0.26 -0.01

This library was initially written for personal projects since existing libraries had complex APIs, missing features, or inaccurate conversions. The goal of this library is to be accurate, intuitive, and easy to use. Although performance is not a priority, conversions are only calculated once; when first evaluated (either on access or as part of an intermediate conversion step) the result is stored for future use.

Unicolour is extensively tested, including verification of roundtrip conversions, validation using known colour values, and 100% line coverage and branch coverage.

🔆 Installation

1. Install the package from NuGet

dotnet add package Wacton.Unicolour

2. Import the package

using Wacton.Unicolour;

3. Use the package

Unicolour colour = new(ColourSpace.Rgb255, 192, 255, 238);

⚡ Quickstart

The simplest way to get started is to make a Unicolour and use it to see how the colour is represented in a different colour space.

var cyan = new Unicolour("#00FFFF");
Console.WriteLine(cyan.Hsl); // 180.0° 100.0% 50.0%

var yellow = new Unicolour(ColourSpace.Rgb255, 255, 255, 0);
Console.WriteLine(yellow.Hex); // #FFFF00

Colours can be mixed or interpolated using any colour space.

var red = new Unicolour(ColourSpace.Rgb, 1.0, 0.0, 0.0);
var blue = new Unicolour(ColourSpace.Hsb, 240, 1.0, 1.0);

/* RGB: [1, 0, 0] ⟶ [0, 0, 1] = [0.5, 0, 0.5] */
var purple = red.Mix(blue, ColourSpace.Rgb);
Console.WriteLine(purple.Rgb); // 0.50 0.00 0.50
Console.WriteLine(purple.Hex); // #800080

/* HSL: [0, 1, 0.5] ⟶ [240, 1, 0.5] = [300, 1, 0.5] */
var magenta = red.Mix(blue, ColourSpace.Hsl);
Console.WriteLine(magenta.Rgb); // 1.00 0.00 1.00
Console.WriteLine(magenta.Hex); // #FF00FF

// #FF0000, #FF0080, #FF00FF, #8000FF, #0000FF
var palette = red.Palette(blue, ColourSpace.Hsl, 5);
Console.WriteLine(palette.Select(colour => colour.Hex));

The difference or distance between colours can be calculated using any delta E metric.

var white = new Unicolour(ColourSpace.Oklab, 1.0, 0.0, 0.0);
var black = new Unicolour(ColourSpace.Oklab, 0.0, 0.0, 0.0);
var difference = white.Difference(black, DeltaE.Ciede2000);
Console.WriteLine(difference); // 100.0000

Other useful colour information is available, such as chromaticity coordinates, temperature, and dominant wavelength.

var equalEnergy = new Unicolour(ColourSpace.Xyz, 0.5, 0.5, 0.5);
Console.WriteLine(equalEnergy.RelativeLuminance);  // 0.5
Console.WriteLine(equalEnergy.Chromaticity.Xy);    // (0.3333, 0.3333)
Console.WriteLine(equalEnergy.Chromaticity.Uv);    // (0.2105, 0.3158)
Console.WriteLine(equalEnergy.Temperature);        // 5455.5 K (Δuv -0.00442)
Console.WriteLine(equalEnergy.DominantWavelength); // 596.1

Reference white points (e.g. D65) and the RGB model (e.g. sRGB) can be configured.

🌈 Features

Convert between colour spaces

Unicolour calculates all transformations required to convert from one colour space to any other, so there is no need to manually chain multiple functions and removes the risk of rounding errors.

Unicolour colour = new(ColourSpace.Rgb255, 192, 255, 238);
var (l, c, h) = colour.Oklch;

Unicolour pink = new("ff1493");
var hex = pink.Hex; // #FF1493

Unicolour grey = new(ColourSpace.Rgb, 0.5);
var hex = grey.Hex; // #808080

%%{
  init: {
  "theme": "base",
  "themeVariables": {
    "primaryColor": "#4C566A",
    "primaryTextColor": "#ECEFF4",
    "primaryBorderColor": "#2E3440",
    "lineColor": "#8FBCBB",
    "secondaryColor": "#404046",
    "tertiaryColor": "#404046"
    }
  }
}%%

flowchart LR
  RGBLIN(Linear RGB)
  RGB(RGB)
  HSB(HSB / HSV)
  HSL(HSL)
  HWB(HWB)
  HSI(HSI)
  XYZ(XYZ)
  XYY(xyY)
  WXY(WXY)
  LAB(LAB)
  LCHAB(LCHab)
  LUV(LUV)
  LCHUV(LCHuv)
  HSLUV(HSLuv)
  HPLUV(HPLuv)
  YPBPR(YPbPr)
  YCBCR("YCbCr / YUV (digital)")
  YCGCO("YCgCo")
  YUV("YUV (PAL)")
  YIQ("YIQ (NTSC)")
  YDBDR("YDbDr (SECAM)")
  TSL(TSL)
  XYB(XYB)
  LMS([LMS])
  IPT{{IPT}}
  ICTCP{{ICtCp}}
  JZAZBZ{{JzAzBz}}
  JZCZHZ{{JzCzHz}}
  OKLAB{{Oklab}}
  OKLCH{{Oklch}}
  OKHSV{{Okhsv}}
  OKHSL{{Okhsl}}
  OKHWB{{Okhwb}}
  OKLRAB{{Oklrab}}
  OKLRCH{{Oklrch}}
  CAM02(CAM02)
  CAM02UCS(CAM02-UCS)
  CAM16(CAM16)
  CAM16UCS(CAM16-UCS)
  HCT{{HCT}}
  Munsell{{"Munsell HVC"}}
  ICC{{"ICC Profile"}}
  CMYK("CMYK")
  
  XYZ --> ICC
  ICC -.-> CMYK
  RGB -.-> CMYK
  XYZ --> XYY
  XYY --> Munsell
  XYY --> WXY
  XYZ --> RGBLIN
  RGBLIN --> RGB
  RGB --> HSB
  HSB --> HSL
  HSB --> HWB
  RGB --> HSI
  RGB --> YPBPR
  YPBPR --> YCBCR
  RGB --> YCGCO
  RGB --> YUV
  YUV --> YIQ
  YUV --> YDBDR
  RGB --> TSL
  RGBLIN --> XYB
  XYZ --> LAB
  LAB --> LCHAB
  XYZ --> LUV
  LUV --> LCHUV
  LCHUV --> HSLUV
  LCHUV --> HPLUV
  XYZ --> LMS
  XYZ --> IPT
  XYZ --> ICTCP
  XYZ --> JZAZBZ
  JZAZBZ --> JZCZHZ
  XYZ --> OKLAB
  OKLAB --> OKLCH
  OKLAB --> OKHSV
  OKLAB --> OKHSL
  OKHSV --> OKHWB
  OKLAB --> OKLRAB
  OKLRAB --> OKLRCH
  XYZ --> CAM02
  CAM02 -.-> CAM02UCS
  XYZ --> CAM16
  CAM16 -.-> CAM16UCS
  XYZ --> HCT

Unicolour pink = new(ColourSpace.Munsell, Hue.FromMunsell(6.1, "RP"), 5.5, 19.5);
Console.WriteLine(pink.Munsell); // 6.1RP 5.5/19.5

Mix colours

Two colours can be mixed by interpolating between them in any colour space, taking into account cyclic hue, interpolation distance, and alpha premultiplication. Palettes provide a range of evenly distributed mixes of two colours.

var red = new Unicolour(ColourSpace.Rgb, 1.0, 0.0, 0.0, alpha: 1.0);
var blue = new Unicolour(ColourSpace.Hsb, 240, 1.0, 1.0, alpha: 1.0);
var magenta = red.Mix(blue, ColourSpace.Hsl, 0.5, HueSpan.Decreasing);
var green = red.Mix(blue, ColourSpace.Hsl, 0.5, HueSpan.Increasing);
var palette = red.Palette(blue, ColourSpace.Hsl, 10, HueSpan.Longer);

The hue of colours created from a single achromatic value are ignored.

var yellow = new Unicolour(ColourSpace.Hsb, 60, 1, 1);

// hue moves 50% from 60° to 240° = 150°
var blackWithBlueHue = new Unicolour(ColourSpace.Hsb, 240, 0, 0);
var darkGreen = yellow.Mix(blackWithBlueHue, ColourSpace.Hsb);

// hue stays at 60°
var blackWithNoHue = new Unicolour(ColourSpace.Hsb, grey: 0);
var darkYellow = yellow.Mix(blackWithNoHue, ColourSpace.Hsb);

Blend colours

Two colours can be blended as though they are layered elements. Compositing is performed using the source-over operator.

var blue = new Unicolour(ColourSpace.Rgb, 240, 1.0, 1.0, alpha: 0.5);
var red = new Unicolour(ColourSpace.Rgb, 1.0, 0.0, 0.0);
var purple = blue.Blend(red, BlendMode.Normal);
var pink = blue.Blend(red, BlendMode.Screen);

Compare colours

Two methods of comparing colours are available: contrast and difference. Difference is calculated according to a specific delta E (ΔE) metric.

var red = new Unicolour(ColourSpace.Rgb, 1.0, 0.0, 0.0);
var blue = new Unicolour(ColourSpace.Hsb, 240, 1.0, 1.0);
var contrast = red.Contrast(blue);
var difference = red.Difference(blue, DeltaE.Cie76);

Map colour into gamut

Colours that cannot be displayed with the configured RGB model can be mapped to the closest in-gamut RGB colour. Mapping to Pointer's gamut will return the closest empirically real surface colour of the same lightness and hue. Mapping to MacAdam limits will return the closest theoretically real surface colour of the same wavelength and luminance.

var veryRed = new Unicolour(ColourSpace.Rgb, 1.25, -0.39, -0.14);

var isDisplayable = veryRed.IsInRgbGamut;
var displayRed = veryRed.MapToRgbGamut();

var isEmpiricalSurface = veryRed.IsInPointerGamut;
var empiricalRed = veryRed.MapToPointerGamut();

var isTheoreticalSurface = veryRed.IsInMacAdamLimits;
var theoreticalRed = veryRed.MapToMacAdamLimits();

Simulate colour vision deficiency

Colour vision deficiency (CVD) or colour blindness can be simulated, conveying how a particular colour might be perceived. Anomalous trichromacy, where cones are defective instead of missing, can be adjusted using the severity parameter.

var colour = new Unicolour(ColourSpace.Rgb255, 192, 255, 238);
var missingRed = colour.Simulate(Cvd.Protanopia);
var defectiveRed = colour.Simulate(Cvd.Protanomaly, 0.5);

Access colourimetric components

Notable colourimetric components are conveniently accessible, and can be used to create a colour.

var grey = new Unicolour(ColourSpace.RgbLinear, 0.5, 0.5, 0.5);
var chromaticity = grey.Chromaticity;
var luminance = grey.RelativeLuminance;

var white = new Unicolour(chromaticity, luminance: 1.0);

Derive temperature metrics

Correlated colour temperature (CCT) and delta UV (∆_uv) can be derived from a colour, and can be used to create a colour. CCT from 500 K to 1,000,000,000 K is supported but only CCT from 1,000 K to 20,000 K is guaranteed to have high accuracy.

var chromaticity = new Chromaticity(0.3457, 0.3585);
var d50 = new Unicolour(chromaticity);
var (cct, duv) = d50.Temperature;

var temperature = new Temperature(6504, 0.0032);
var d65 = new Unicolour(temperature);
var (x, y) = d65.Chromaticity;

Get wavelength attributes

The dominant wavelength and excitation purity of a colour can be ascertained using the spectral locus. They can be used to create a colour alongside the WXY colour space. Wavelengths from 360 nm to 700 nm are supported.

var chromaticity = new Chromaticity(0.1, 0.8);
var hyperGreen = new Unicolour(chromaticity);
var dominantWavelength = hyperGreen.DominantWavelength;
var excitationPurity = hyperGreen.ExcitationPurity;

var laserRed = new Unicolour(ColourSpace.Wxy, 670, 1.0, 0.5);

Detect imaginary colours

Colours that lie outside the spectral locus — the horseshoe-shaped curve of the CIE xy chromaticity diagram — cannot be produced by the eye. These imaginary colours are mathematically possible and can be detected.

var chromaticity = new Chromaticity(0.05, 0.05);
var impossibleBlue = new Unicolour(chromaticity);
var isImaginary = impossibleBlue.IsImaginary;

Interpret spectral power distributions

A colour can be created from a spectral power distribution (SPD). Wavelengths should be provided in either 1 nm or 5 nm intervals, and omitted wavelengths are assumed to have zero spectral power.

/* [575 nm] -> 0.5 · [580 nm] -> 1.0 · [585 nm] -> 0.5 */
var spd = new Spd(start: 575, interval: 5, coefficients: [0.5, 1.0, 0.5]);
var intenseYellow = new Unicolour(spd);

Model pigment and paint colours

Pigments can be combined using the Kubelka-Munk theory. The result is a colour that reflects natural paint mixing. Pigment measurements are required, either coefficients for absorption k and scattering s (two-constant) or a reflectance curve r (single-constant). Saunderson correction can be applied when using k and s and assumes measurements were taken in SPEX mode.

/* populate k and s with measurement data */
var phthaloBlue = new Pigment(startWavelength: 380, wavelengthInterval: 10, k: [], s: []);
var hansaYellow = new Pigment(startWavelength: 380, wavelengthInterval: 10, k: [], s: []);
var green = new Unicolour(pigments: [phthaloBlue, hansaYellow], weights: [0.5, 0.5]);

Use ICC profiles for CMYK conversion

Device-dependent colour prints of 4 (e.g. FOGRA39 CMYK) or more (e.g. FOGRA55 CMYKOGV) are supported through ICC profiles. If no ICC profile is provided, or if the profile is incompatible, naive conversion for uncalibrated CMYK is used instead.

using Wacton.Unicolour.Icc;

var fogra39 = new IccConfiguration("./Fogra39.icc", Intent.RelativeColorimetric);
var config = new Configuration(iccConfig: fogra39);

var navyRgb = new Unicolour(config, ColourSpace.Rgb255, 0, 0, 128);
Console.WriteLine(navyRgb.Icc); // 1.0000 0.8977 0.0001 0.2867 CMYK

var navyCmyk = new Unicolour(config, new Channels(1.0, 1.0, 0.0, 0.5));
Console.WriteLine(navyCmyk.Rgb.Byte255); // 46 37 87

The following tables summarise which ICC profiles are compatible with Unicolour:

A wider variety of ICC profiles will be supported in future releases. If a problem is encountered using an ICC profile that meets the above criteria, please raise an issue.

Handle invalid values

It is possible for invalid or unreasonable values to be used in calculations, either because conversion formulas have limitations or because a user passes them as arguments. Although these values don't make sense to use, they should propagate safely and avoid triggering exceptions.

var bad1 = new Unicolour(ColourSpace.Oklab, double.NegativeInfinity, double.NaN, double.Epsilon);
var bad2 = new Unicolour(ColourSpace.Cam16, double.NaN, double.MaxValue, double.MinValue);
var bad3 = bad1.Mix(bad2, ColourSpace.Hct, amount: double.PositiveInfinity);

Sensible defaults, highly configurable

Unicolour uses sRGB as the default RGB model and standard illuminant D65 (2° observer) as the default white point of all colour spaces, ensuring consistency and a suitable starting point for simple applications. These can be overridden using the Configuration parameter, and common configurations have been predefined.

var defaultConfig = new Configuration(RgbConfiguration.StandardRgb, XyzConfiguration.D65);
var colour = new Unicolour(defaultConfig, ColourSpace.Rgb255, 192, 255, 238);

Zero dependencies, quality controlled

Each line of artisan code is exquisitely handcrafted in small-batch programming sessions. No dependencies are used, so there is no risk of reliance on deprecated, obsolete, or unmaintained packages. Every line of code is tested, and any defect is Unicolour's responsibility.

💡 Configuration

The Configuration parameter can be used to define the context of the colour.

Example configuration with predefined

• Rec. 2020 RGB
• Illuminant D50 (2° observer) XYZ

Configuration config = new(RgbConfiguration.Rec2020, XyzConfiguration.D50);
Unicolour colour = new(config, ColourSpace.Rgb255, 204, 64, 132);

Example configuration with manually defined

• Wide-gamut RGB
• Illuminant C (10° observer) XYZ, using Von Kries method for white point adaptation

var rgbConfig = new RgbConfiguration(
    chromaticityR: new(0.7347, 0.2653),
    chromaticityG: new(0.1152, 0.8264),
    chromaticityB: new(0.1566, 0.0177),
    whitePoint: Illuminant.D50.GetWhitePoint(Observer.Degree2),
    fromLinear: value => Math.Pow(value, 1 / 2.19921875),
    toLinear: value => Math.Pow(value, 2.19921875)
);

var xyzConfig = new XyzConfiguration(Illuminant.C, Observer.Degree10, ChromaticAdaptation.VonKries);

var config = new Configuration(rgbConfig, xyzConfig);
var colour = new Unicolour(config, ColourSpace.Rgb255, 202, 97, 143);

A Configuration is composed of sub-configurations. Each sub-configuration is optional and will fall back to a sensible default if not provided.

RgbConfiguration

Defines the RGB colour space parameters, often used to specify a wider gamut than standard RGB (sRGB).

mindmap
  root(RGB)
    ("R 0.64 0.33<br> G 0.30 0.60<br>B 0.15 0.06")
      ("D65")
        ("sRGB")
        ("Rec. 709")
        ("xvYCC")
    ("R 0.68 0.32<br>G 0.27 0.69<br>B 0.15 0.06")
      ("D65")
        ("Display P3")
    ("R 0.71 0.29<br>G 0.17 0.80<br>B 0.13 0.05")
      ("D65")
        ("Rec. 2020")
        ("Rec. 2100 PQ")
        ("Rec. 2100 HLG")
    ("R 0.64 0.33<br>G 0.21 0.71<br>B 0.15 0.06")
      ("D65")
        ("A98 RGB")
    ("R 0.73 0.27<br>G 0.16 0.84<br>B 0.03 0.00")
      ("D50")
        ("ProPhoto RGB")
    ("R 0.73 0.27<br>G 0.00 1.00<br>B 0.00 -0.08")
      ("W 0.32168 0.33767")
        ("ACES 2065-1")
    ("R 0.71 0.29<br>G 0.17 0.83<br>B 0.13 0.04")
      ("“D60”")
        ("ACEScg")
        ("ACEScct")
        ("ACEScc")
    ("R 0.64 0.33<br>G 0.29 0.60<br>B 0.15 0.06")
      ("D65")
        ("Rec. 601 (625-line)")
        ("PAL (Rec. 470)")
        ("PAL 625 (Rec. 1700)")
        ("SECAM (Rec. 470)")
        ("SECAM 625 (Rec. 1700)")
    ("R 0.67 0.33<br>G 0.21 0.71<br>B 0.14 0.08")
      ("C")
        ("PAL-M (Rec. 470)")
        ("NTSC (Rec. 470)")
    ("R 0.63 0.34<br>G 0.31 0.60<br>B 0.16 0.07")
      ("C")
        ("PAL 525 (Rec. 1700)")
      ("D65")
        ("Rec. 601 (525-line)")
        ("NTSC (SMPTE-C)")
        ("NTSC 525 (Rec. 1700)")

• Parameters
  • Red, green, and blue chromaticity coordinates
  • Reference white point
  • Companding functions to and from linear values

XyzConfiguration

Defines the XYZ white point (which is also inherited by colour spaces that do not need a specific configuration), the observer to use when colour matching functions (CMFs) are required, and the chromatic adaptation matrix to use for any white point adaptation (the Bradford method will be used if unspecified).

• Parameters
  • Reference white point or illuminant
  • Observer
  • Chromatic adaptation matrix

YbrConfiguration

Defines the constants, scaling, and offsets required to convert to YPbPr and YCbCr.

• Parameters
  • Luma constants for component video separation
  • Mapping ranges for digital encoding

CamConfiguration

Defines the viewing conditions for CAM02 and CAM16, which take into account the surrounding environment to determine how a colour is perceived.

The predefined sRGB configuration refers to an ambient illumination of 64 lux under a grey world assumption.

• Parameters
  • Reference white point
  • Adapting luminance
  • Background luminance

DynamicRange

Defines luminance values used when evaluating perceptual quantizer (PQ) transfer functions (IC_TC_P · J_za_zb_z · J_zC_zh_z · Rec. 2100 PQ RGB) and hybrid log-gamma (HLG) transfer functions (Rec. 2100 HLG RGB).

The predefined HDR configuration has a white luminance of 203 cd/m² at 75% HLG, and a minimum luminance of 0 cd/m² (no black lift).

• Parameters
  • White luminance
  • Maximum luminance
  • Minimum luminance
  • HLG % white level

IccConfiguration

Defines the ICC profile and rendering intent, typically used for accurate CMYK conversion.

Unicolour does not embed or distribute ICC profiles. Some commonly used profiles can be found in the ICC profile registry.

• Parameters
  • ICC profile (.icc file)
  • Rendering intent

White points

Most colour spaces are impacted by the reference white point. Unicolour applies different reference white points to different sets of colour spaces, as shown in the table below. When a conversion to or from XYZ space involves a change in white point, a chromatic adaptation transform (CAT) is performed. The default chromatic adaptation is the Bradford method but this can be customised.

Convert between configurations

A Unicolour can be converted to a different configuration, in turn enabling conversions between different RGB models, XYZ white points, CAM viewing conditions, ICC profiles, etc.

/* pure sRGB green */
var srgbConfig = new Configuration(RgbConfiguration.StandardRgb);
var srgbColour = new Unicolour(srgbConfig, ColourSpace.Rgb, 0, 1, 0);
Console.WriteLine(srgbColour.Rgb); // 0.00 1.00 0.00

/* ⟶ Display P3 */
var displayP3Config = new Configuration(RgbConfiguration.DisplayP3);
var displayP3Colour = srgbColour.ConvertToConfiguration(displayP3Config);
Console.WriteLine(displayP3Colour.Rgb); // 0.46 0.99 0.30

/* ⟶ Rec. 2020 */
var rec2020Config = new Configuration(RgbConfiguration.Rec2020);
var rec2020Colour = displayP3Colour.ConvertToConfiguration(rec2020Config);
Console.WriteLine(rec2020Colour.Rgb); // 0.57 0.96 0.27

✨ Examples

This repository contains projects showing how Unicolour can be used to create:

1. Gradient images
2. Heatmaps of luminance
3. Diagrams of colour data
4. A colourful console application
5. A colour picker web application
6. 3D visualisations of colour spaces in Unity

Gradients

Example code to create gradient images using 📷 SixLabors.ImageSharp can be seen in the Example.Gradients project.

Gradient of deep pink to aquamarine through Oklch colour space using Mix()	Gradient of deep pink to aquamarine through Oklch colour space using Palette()

Gradient of purple to orange generated in every colour space	Gradient of black to green generated in every colour space

Visualisation of temperature from 1,000 K to 13,000 K

Colour spectrum rendered with different colour vision deficiencies

Demonstration of interpolating from red to transparent to blue, with and without premultiplied alpha

Perceptually uniform colourmaps from Unicolour.Datasets	Perceptually uniform colour palettes from Unicolour.Datasets

Mixes of two-constant pigments to titanium white from Unicolour.Datasets	Palettes of two-constant pigments to titanium white from Unicolour.Datasets

Colour wheel based on RYB pigments from Unicolour.Experimental (see also: my grandmother's patented Colour Joy kleurwiel)	Colour wheel based on hue from Unicolour.Experimental (see also: my grandmother's patented Colour Joy kleurwiel)

Colour harmonies based on RYB pigments from Unicolour.Experimental	Colour harmonies based on hue from Unicolour.Experimental

Heatmaps

Example code to create heatmaps of luminance using 📷 SixLabors.ImageSharp with images from 🚀 NASA can be seen in the Example.Heatmaps project.

Heatmap of the ☀️ sun using perceptually uniform colourmaps from Unicolour.Datasets

Heatmap of the 🌕 moon using perceptually uniform colourmaps from Unicolour.Datasets

Diagrams

Example code to create diagrams of colour data using 📈 ScottPlot can be seen in the Example.Diagrams project.

CIE xy chromaticity diagram with sRGB gamut	CIE 1960 colour space with sRGB gamut

CIE xy chromaticity diagram with Planckian or blackbody locus	CIE 1960 colour space with Planckian or blackbody locus

CIE xy chromaticity diagram with MacAdam limits	CIE 1960 colour space with MacAdam limits

CIE xy chromaticity diagram with spectral locus plotted at 1 nm intervals	CIE 1960 colour space with spectral locus plotted at 1 nm intervals

Console

Example code to create a colourful console application using ⌨️ Spectre.Console can be seen in the Example.Console project.

Console application displaying colour information from a hex value

Web

Example code to create a client-side colour picker web application using 🕸️ Blazor can be seen in the Example.Web project.

See the live demo!

Web application for picking colours in any colour space

Unity

Example code to create 3D visualisations of colour spaces using 🎮 Unity can be seen in the Example.Unity project.

Try it out online in Unity Play!

3D visualisation of colour spaces in Unity

3D movement through colour spaces in Unity

🔮 Datasets

Some colour datasets have been compiled for convenience in the Unicolour.Datasets project.

Commonly used sets of colours:

• CSS specification named colours
• ISCC-NBS colour centroids
• xkcd colour survey results
• Macbeth ColorChecker colour rendition chart
• Nord theme colours

Perceptually uniform colourmaps / palettes:

• Viridis, Plasma, Inferno & Magma (sequential)
• Cividis (sequential)
• Mako, Rocket, Crest & Flare (sequential)
• Vlag & Icefire (diverging)
• Twilight & Twilight Shifted (cyclic)
• Turbo (rainbow)
• Cubehelix (sequential)

Colour data used in academic literature:

• MacAdam limits of surface colours
• Hung-Berns constant hue loci data
• Ebner-Fairchild constant perceived-hue data

Known pigments:

• Artist Paint Spectral Database two-constant pigment data

Example usage:

1. Install the package from NuGet

dotnet add package Wacton.Unicolour.Datasets

2. Import the package

using Wacton.Unicolour.Datasets;

3. Reference the predefined Unicolour

var pink = Css.DeepPink;
var green = Xkcd.NastyGreen;
var mapped = Colourmaps.Viridis.Map(0.5);
var palette = Colourmaps.Turbo.Palette(10);

🥽 Experimental

There are additional useful features that are considered too ambiguous, indeterminate, or opinionated to be included as part of the core Unicolour library. These have been assembled in the Unicolour.Experimental project.

1. Install the package from NuGet

dotnet add package Wacton.Unicolour.Experimental

2. Import the package

using Wacton.Unicolour.Experimental;

Generate colour harmonies

Traditionally, harmonious colour combinations are based on their relationship within a colour wheel. A colour wheel can be generated using pigments or a hued colour space.

/* populate pigment k and s with measurement data */
var quinaRed = new Pigment(380, 1, k: [], s: []);
var bismuthYellow = new Pigment(380, 1, k: [], s: []);
var ceruleanBlue = new Pigment(380, 1, k: [], s: []);
var titaniumWhite = new Pigment(380, 1, k: [], s: []);
var boneBlack = new Pigment(380, 1, k: [], s: []);

var colourWheel = usePigments 
    ? ColourWheel.From(quinaRed, bismuthYellow, ceruleanBlue, titaniumWhite, boneBlack)
    : ColourWheel.From(ColourSpace.Oklch, reference: new Unicolour("ff0000"));

var orange = colourWheel.Pure(hue: 60);
var lightGreen = colourWheel.Tint(hue: 180, weight: 1);
var darkPurple = colourWheel.Shade(hue: 300, weight: 1);
var greyRed = colourWheel.Tone(hue: 0, weight: 1);

var orangePalette = colourWheel.Harmony(hue: 60, Harmony.Analogous);

Approximate pigments

A reflectance curve can be generated for any colour, approximating a single-constant pigment. This enables Kubelka-Munk pigment mixing without taking reflectance measurements. Note that, similar to metamerism, there are infinitely many reflectance curves that can generate a single colour; this will find just one.

var redPigment = PigmentGenerator.From(new Unicolour("#FF0000"));
var bluePigment = PigmentGenerator.From(new Unicolour("#0000FF"));
var magenta = new Unicolour([redPigment, bluePigment], [0.5, 0.5]);

Emulate Spectral.js

Spectral.js uses artificial reflectance curves to perform single-constant pigment mixing. However, input concentrations are adjusted according to luminance and a custom weighting curve that the author found to give aesthetically pleasing results. This behaviour has been replicated here except 1) reflectance curves are more accurately generated at a performance cost and 2) it has been extended to be able to mix more than two colours.

var blue = new Unicolour("#0000FF");
var yellow = new Unicolour("#FFFF00");
var green = SpectralJs.Mix([blue, yellow], [0.5, 0.5]);
var palette = SpectralJs.Palette(blue, yellow, 9);

Wacton.Unicolour is licensed under the MIT License, copyright © 2022-2026 William Acton.