Gamut Overlap Calculator

What a gamut triangle actually shows

A color gamut is defined by three primaries — the most saturated red, green and blue a system can produce — plotted as chromaticity coordinates on a 1931 CIE xy diagram (or the perceptually rescaled 1976 u′v′ diagram). Draw a straight line between any two primaries and every color a mix of those two lights can produce sits on that line; draw the third primary and the triangle those three edges enclose is every color that system can display or capture, purely in terms of hue and saturation — luminance (how bright) is a separate axis this flat plot doesn't show. That is why "gamut area" here means chromaticity area, not color volume — more on that in the caveat below.

Coverage is directional, not symmetric

"Covers 95% of P3" and "is 95% of P3" are different claims, and mixing them up is the single most common misread of a gamut spec. Coverage of A over B is overlap-area divided by area(B): how much of B's triangle sits inside A's triangle. It says nothing about how much bigger A's triangle is overall — a wide display gamut can cover 100% of a small reference gamut while also containing color the reference never touches at all. Rec.2020 fully covers Rec.709, but a monitor spec that reads "covers 100% Rec.709" does not mean the monitor's gamut is Rec.709-sized. Always read gamut coverage as A-covers-B, not as one symmetric percentage.

Worked example: Rec.709 inside DCI-P3, in xy

Rec.709 and DCI-P3 share the identical blue primary (0.150, 0.060) — a deliberate choice in the DCI-P3 spec so P3 cleanly supersets Rec.709's blue corner. Triangle area comes from the shoelace formula: Area = half the absolute value of xR(yG − yB) + xG(yB − yR) + xB(yR − yG).

Rec.709 — R(0.640, 0.330), G(0.300, 0.600), B(0.150, 0.060): Area = 0.5 × |0.640(0.600−0.060) + 0.300(0.060−0.330) + 0.150(0.330−0.600)| = 0.5 × |0.3456 − 0.0810 − 0.0405| = 0.5 × 0.2241 = 0.11205

DCI-P3 — R(0.680, 0.320), G(0.265, 0.690), B(0.150, 0.060): Area = 0.5 × |0.680(0.690−0.060) + 0.265(0.060−0.320) + 0.150(0.320−0.690)| = 0.5 × |0.4284 − 0.0689 − 0.0555| = 0.5 × 0.3040 = 0.15200

Clipping Rec.709's triangle against P3's with Sutherland–Hodgman returns Rec.709 unchanged — every Rec.709 vertex already sits inside P3 — so overlap = 0.11205 exactly. P3 covers 100% of Rec.709; Rec.709 covers 0.11205 ÷ 0.15200 = 73.7% of P3. DCI-P3's own triangle is about 35.7% larger in xy area than Rec.709's.

Why xy and u′v′ disagree

CIE 1931 xy is not perceptually uniform — equal xy distances represent wildly different amounts of perceived color difference, and the distortion is worst in the green corner, where the diagram stretches a small perceptual range across a large plotted area. CIE 1976 u′v′ reparameterizes the same spectral locus to be closer to perceptually even, which shrinks that green bulge relative to the rest of the diagram. Because P3 and Rec.2020 both reach further into the stretched green region than Rec.709 does, their xy areas — and their coverage percentages against Rec.709 — read larger than their u′v′ areas do. The same Rec.709-vs-P3 pair above computes to Rec.709 covering 79.6% of P3 in u′v′, not 73.7%. If the real question is "how much wider does this gamut actually look," trust u′v′ over xy; xy systematically overstates green-heavy gamuts.

Imaginary primaries: why camera gamuts go outside the locus

ARRI Wide Gamut 4's blue primary sits at (0.0991, −0.0308) — a negative y coordinate, physically outside the visible spectral locus. No real light source produces that chromaticity; it is an "imaginary primary," chosen on purpose. Camera and post-production working gamuts (AWG4, ACES AP0/AP1, Sony S-Gamut among them) push primaries past the real locus so every color a sensor can actually capture can still be encoded with positive RGB values. Pull a primary back inside the real locus and some sensor-captured colors would need a negative coefficient to reconstruct, which breaks simple RGB math and clips highlights or saturation during grading. It is headroom built into the pipeline, not a claim that the camera can capture, or a monitor can display, that exact chromaticity.

Practical guidance

Display spec sheets ("110% sRGB," "95% DCI-P3") are always directional coverage of the display's real gamut over the named reference — check which space, xy or u′v′, the vendor used, because the same panel measures differently in each and vendors tend to quote whichever number is larger. Rec.2020 coverage numbers on consumer monitors are close to meaningless marketing today: almost no display comes near filling Rec.2020, and even a panel "covering" a large share of it in xy area is unlikely to hold workable luminance across that whole triangle, which is exactly where the 2-D-versus-3-D gap below starts to matter. When choosing a grading or VFX working space, wider is not automatically better — AP1 and AWG4 buy real headroom over camera-native colors, but only if every tool downstream in the pipeline actually respects that working space's transforms instead of clipping silently.

The volume caveat

This tool measures a flat chromaticity triangle: hue and saturation only. Real gamut "size" is a 3-D volume — chromaticity combined with luminance, usually evaluated in CIELAB or CIECAM02 across many lightness levels — and two gamuts can share nearly identical 2-D triangle areas while having very different achievable volumes, because of how each clips near black and near white. Treat every percentage on this page as a chromaticity-only comparison, not a complete colorimetric verdict on a display or camera.

1. Choose two gamuts to compare - Rec.709, DCI-P3, Rec.2020, Adobe RGB, ACES AP1 or ARRI Wide Gamut 4.

2. Read the overlap area and both coverage directions, since "covers 95% of P3" and "is 95% of P3" are different claims.

3. Switch between CIE 1931 xy and 1976 u'v' to see why spec sheets disagree, then check patch values in the SMPTE Color Bars Code Value Chart.