Chromaticity

Chromaticity is the colour component of a light spectrum independent of its intensity. Two lights with the same chromaticity differ only in brightness — they’re “the same colour” in the conventional sense.

The standard representation is the CIE 1931 xy chromaticity diagram: a horseshoe-shaped 2D plot where every visible colour at any intensity maps to a single point. White (D65 daylight) sits near (0.31, 0.33). Pure red, green, and blue primaries sit at three corners; mixing produces every colour inside the triangle formed by those primaries.

A colour space is defined largely by its primary chromaticity points. sRGB picks its three primaries at one set of (x, y) coordinates; Display P3 picks them further out, especially the red, expanding the achievable gamut. The chromaticity diagram shows visually which colours each space can represent — the larger the triangle, the wider the gamut.

Practical use: when picking a colour space for an asset (web design, print, video), the chromaticity diagram is the chart that shows whether the space covers the colours you need. For print, CMYK’s achievable gamut is a sub-region of sRGB’s, which is why some sRGB colours simply can’t be reproduced on a printer regardless of technique.

Worked example

sRGB’s three primaries sit at chromaticity coordinates: red (0.640, 0.330), green (0.300, 0.600), blue (0.150, 0.060), with the D65 white point at (0.3127, 0.3290). Plot those four points on the CIE 1931 diagram and the triangle formed by the three primaries covers roughly 35% of the visible-colour horseshoe. Display P3 keeps the same blue, moves green slightly to (0.265, 0.690) and red out to (0.680, 0.320), expanding the triangle by about 25% — most of the gain is in saturated reds and greens used by professional photography and HDR video. Rec. 2020 (UHDTV) pushes the primaries to spectral-locus colours: red (0.708, 0.292), green (0.170, 0.797), blue (0.131, 0.046), covering ~76% of the visible gamut but at the cost of requiring laser primaries no consumer display currently delivers — most Rec. 2020 mastering targets only 90% Rec. 2020 coverage in practice.

When and why it matters

Chromaticity matters whenever a designer’s “saturated red” refuses to print or display correctly. The colour exists in human perception, has a chromaticity coordinate, but falls outside the destination gamut — so the renderer clips it to the nearest in-gamut chromaticity, shifting the hue. Print designers handle this via soft-proofing in Photoshop, which simulates how a CMYK printer’s smaller gamut will distort each source colour. Photographers exporting for the web should match colour spaces (export from Adobe RGB → sRGB, not vice versa) so the browser does not have to guess. Video editors mastering for both Rec. 709 (SDR) and Rec. 2020 (HDR) deliveries need to know which clips contain colours that exist in Rec. 2020 but not Rec. 709, so they can prepare alternate trims. Reference: ICC Specifications — colour management documentation.

The CIE 1931 diagram’s known flaws and the modern alternatives: the 1931 horseshoe is perceptually non-uniform — equal distances in xy do not correspond to equal perceived colour differences, especially in the green region which dominates the diagram’s area despite humans being relatively insensitive to green-vs-green variations. CIE 1976 UCS (u′v′) and the more recent CIECAM02 chromaticity coordinates compress the green and stretch the blue to produce a closer approximation to perceptual uniformity. Modern colour-science tools default to u′v′ for visualisation, though xy remains the canonical form in specifications because every legacy display standard is defined in it.

White point — the third coordinate that matters: a chromaticity diagram is two-dimensional, but a colour space needs a defined “white” to convert chromaticity to RGB values. sRGB and Rec. 709 use the D65 white point (6504 K, midday daylight in Europe); cinema DCI-P3 uses a slightly different white closer to 6300 K to match projector tungsten; print uses D50 (5000 K) because warmer light better simulates indoor viewing. Mismatched white points are the most common reason a calibrated monitor and a printer still disagree on “white” — both are in-gamut but pointing at different reference whites. Related: colour temperature, ICC profile. Reference: CIE 015:2018 Colorimetry, 4th edition.