One triangle. Different definitions.

P3 originated in digital cinema but now appears in several colour definitions. A shared primary triangle does not make a theatrical projector, web artwork and an HDR master the same format.

Establish the target first: conventional SDR cinema, a Display P3 file, or monitoring with P3 primaries. This chapter treats the conventional cinema path separately from HDR extensions. A P3 label cannot replace a delivery specification.

P3 variants share R (0.680, 0.320), G (0.265, 0.690), B (0.150, 0.060). Blue matches sRGB/Rec.709; red and green differ.

Primaries define linear basis directions. The white point determines their scaling so equal RGB yields the intended white. Changing DCI white to D65 leaves the triangle unchanged but changes the RGB ↔ XYZ matrix. This is numerically testable without relying on a swatch’s subjective appearance.

Conventional D-Cinema uses reference white near (0.314, 0.351) and an SDR screen-white reference of 48 cd/m². D65 has different coordinates: (0.3127, 0.3290). This describes white chromaticity, not just a preset colour temperature.

Moving an image between whites requires an explicit adaptation decision. The experiment offers linear Bradford adaptation to D65 for preview. That is a selected mathematical model, not a simulation of visual adaptation in a theatre.

A 2.6 power response belongs to its corresponding theatrical context. Display P3 instead uses the piecewise sRGB transfer. Selecting “P3” without specifying the transfer leaves a major part of interpretation undefined.

The lab keeps input codes fixed while changing the definition. Grey code 0.5 decodes to about 0.214 with sRGB and 0.165 with a 2.6 power. Linear values differ before white-point handling even begins. This demonstrates interpretation, not colour-preserving conversion.

In the conventional digital-cinema chain, DCDM imagery is described as nonlinear X′Y′Z′ with 12-bit components. P3 relates to reference reproduction, not a claim that every DCP simply contains P3 RGB.

Creating a theatrical package involves more than selecting an export space: conversion, encoding, packaging, format checks and viewing are all involved. This chapter explains RGB versus X′Y′Z′; it does not replace a DCP acceptance procedure.

P3-D65 specifies P3 primaries and D65 white, not a complete transfer function or image state. One context may pair it with a power function; another may use PQ or a different explicitly defined system.

The lab deliberately labels its “P3-D65 / 2.6” option in full. It is one selected comparison profile, not a definition of every possible P3-D65 signal. This avoids inferring gamma automatically from a gamut name.

Display P3 combines P3 primaries, D65 and the sRGB transfer. It is a defined display space, distinct from a DCI-white/2.6 profile.

A real conversion decodes source values, transforms coordinates with white adaptation when needed, then encodes for the destination. Copying old RGB values into another definition assigns a different meaning; it does not preserve the original colour. This connects the chapter to “A tag is not a transform”.

On the web, Display P3 has an explicit CSS notation and can be used for colour-managed imagery. What matters is format support, interpretation and actual viewing-path capability, not a device brand.

Prepare a correct source, then test wide and conventional viewing paths. Accepting a P3 file does not mean reproducing all its chromaticities. Do not alter artwork to compensate with artificial saturation after its profile was lost.

An HDR pipeline must distinguish signal space, mastering limiting gamut and the display’s actual colour volume. An image can use BT.2020 signal primaries while its actual colours remain within P3-D65; there is no contradiction.

PQ or HLG, target luminance and output rendering cannot be inferred from “P3”. These properties must agree across file and viewing paths. Neither a P3 profile nor a larger diagram triangle creates highlight detail missing from the source.

A coverage percentage describes one chosen gamut comparison. It does not verify white accuracy, tone response, uniformity or reproduction across luminance levels. Area on xy and reproducible colour volume are different measures.

Our diagrams do not produce a “monitor quality percentage”: the browser experiment cannot establish it. Evaluating a working display requires measurements and verification in its intended operating mode, not a single advertised percentage.

Define the destination before configuring the project: Display P3 artwork, a specific P3 monitoring profile, conventional cinema or HDR output. Then verify gamut, white, transfer and source image state separately.

Timeline working space need not match output space. Apply the required conversion once at a known point. Keep a test with a neutral ramp, saturated colours and an output description so the recipient does not select profiles by appearance.