Why can red and green light look yellow?

Their combined stimulation of human color vision can produce a yellow appearance, even though the spectrum differs from a single yellow wavelength.

Do equal RGB numbers always mean the same white?

No. The color space, device and viewing conditions matter; equal values are interpreted within those conventions.

Is printing additive?

Ordinary ink printing mainly relies on selective absorption of reflected light, usually discussed as subtractive color mixing.

AI Glossary / Computer vision and multimodal AI / Beginner

Additive color

Definition

Additive color is the production of color by combining light. Red, green and blue are commonly used as additive primaries: varying their contributions allows a display or lighting system to produce many perceived colors.

What it means

The starting point is little or no light. Adding contributions makes the combined light different in brightness and color. With suitable primary lights and proportions, red and green can appear yellow, green and blue cyan, and red and blue magenta. A balanced combination can appear white.

This is why a screen can show a yellow icon without containing a separate yellow light source for every pixel. Small colored subpixels contribute light that the viewer perceives together at the normal viewing distance.

The concept differs from mixing pigments. Paints and inks absorb parts of the incident light, so their mixtures behave differently. A recipe that works for colored lights should not be assumed to work for paint on a wall.

How it works

At a physical level, light from different sources contributes to the light reaching the eye. Human color perception depends on the pattern of responses across cone types, which is why different spectral mixtures can produce similar color appearances.

An RGB color model describes contributions using three channels. The exact primaries, white point and transfer function belong to a particular color space or device; three bare numbers are not a complete physical specification. The range of reproducible colors is its gamut.

Numerical addition also needs care. Many image files store nonlinear encoded values rather than values directly proportional to light. For light-based calculations, a workflow usually converts to an appropriate linear representation, performs the operation, and then encodes the result for storage or display. Simply adding encoded channel numbers is not generally the same as adding physical light.

Three practical examples

A display creates the appearance of a yellow navigation button from red and green channel contributions.
A theater technician combines colored spotlights on a neutral surface to create a changing lighting effect.
A graphics renderer adds contributions from several simulated light sources before applying its output color and tone processing.

Think of it this way

Think of several lamps illuminating the same patch of a wall. Each supplies some light, and the viewer sees their combined contribution. This is more useful than imagining pots of paint being mixed. The analogy has limits: the wall's reflectance, ambient lighting and the lamps' spectra affect the result, while digital RGB values require a defined color space to describe their intended appearance.

Worked example

Consider idealized linear RGB contributions. One light contributes (0.4, 0, 0), and another contributes (0, 0.3, 0). Their sum is (0.4, 0.3, 0), a red-and-green mixture with no blue contribution. Adding a third contribution of (0, 0, 0.2) produces (0.4, 0.3, 0.2).

These numbers describe relative contributions within a specified linear RGB system. They do not mean that every display will show the same appearance when given those three unlabeled values.

If two contributions produce a channel value of 1.3 while an output format allows only values through 1, the workflow needs a range-management decision. Clipping to 1 discards the excess distinction. A rendering system may retain a wider internal range and apply tone mapping later.

Illustrative case study

Fictional example: Brightstage Learning builds an interactive color lesson. Its first prototype lets children overlap digital red, green and blue lights on a dark background. The team notices that the results look inconsistent when the software mixes stored display values without considering their encoding.

The developer revises the calculation to use the intended linear-light representation, then converts the result for display. The lesson also includes a separate pigment demonstration so that children do not expect paint mixtures to follow identical rules.

The team checks the lesson on its target screens and labels the examples as display simulations. This fictional case illustrates implementation choices, not a claim that all screens reproduce colors identically.

Common misconceptions and mistakes

Red, green and blue are a useful set of additive primaries, not three indivisible ingredients of all physical light. A white appearance does not require a perfectly continuous spectrum or equal physical power at exactly three wavelengths.

RGB channel values also do not determine appearance without their color-space context. More light can exceed the output range, and clipping may alter color relationships. Finally, additive mixing should not be confused with alpha compositing: transparency blending uses weights and coverage conventions, even though its color calculations can involve addition.

Business use

Understanding additive color helps teams manage websites, video, digital signage and rendered product imagery. It clarifies why a screen preview may differ from print and why color profiles matter when files move between applications.

For brand assets, define the intended color space and review actual output conditions rather than treating an RGB triplet as a universal guarantee. For image-processing systems, document whether calculations occur in linear or encoded values. Such details can prevent unexpected dark blends, clipped highlights and inconsistent visual results across an otherwise shared workflow.

Check your understanding

Why is adding two encoded image channel values not necessarily equivalent to adding their light contributions?

Answer: The stored values may be nonlinear encodings. Light-based addition requires an appropriate linear representation and a defined color space.

Related terms

Color space
Linear light
Gamma encoding
Tone mapping
RGB
gamut
white point
subtractive colour
alpha compositing

Learn next

Compare overlapping colored lights with overlapping transparent inks or filters. Then inspect the color profile attached to a digital image. Learn why linear-light calculations and display encoding are separate steps, and test a simple blend in both representations to see how the choice changes the result.

Sources

Exploratorium, additive colour demonstrations and coloured shadows: https://annex.exploratorium.edu/xref/exhibits/colored_shadows.html Exploratorium, colour mixing activities: https://annex.exploratorium.edu/wsw/projects/color_mixing_lab/ W3C CSS Color specification, colour spaces and linear-light representations: https://www.w3.org/TR/css-color-4/ The numerical lighting example and learning-project case are original illustrations.

Previous: Add list Next: Additive effect