Colour Vision

Foundations of Light

Wavelength and Intensity

Two physical properties of light are especially important for vision:

Property	Definition	Perceptual effect
Wavelength	Distance between peaks of a light wave, measured in nanometres (nm)	Mainly relates to colour/hue
Intensity	Amplitude or strength of the light wave	Mainly relates to brightness

• Human vision is sensitive to wavelengths of approximately 400–700 nm.
• Short wavelengths are perceived as more blue/violet.
• Medium wavelengths are perceived as more green.
• Long wavelengths are perceived as more red.
• Wavelength and intensity are independent properties. A red light can be bright or dim without changing its basic hue.

Newton’s Prism Experiments

Newton passed white sunlight through a prism and showed that:

• white light can be separated into a spectrum of colours
• each part of the spectrum corresponds to a different wavelength
• the separated colours can be recombined to form white light

This showed that the prism does not create colour; it separates the wavelengths already present in white light.

Physical Light vs Perceived Colour

Important distinction:

• Wavelength is an objective, physical property of light.
• Colour is a perceptual experience produced by the visual system.

So the physical stimulus is light, but the experience of colour is constructed by the brain.

Colour Mixing

There are two main types of colour mixing.

Additive Colour Mixing

Additive mixing occurs when different sources of light are combined.

Primary colours of light:

• red
• green
• blue

Key combinations:

• red + green → yellow
• blue + green → turquoise/cyan
• red + blue → purple
• red + green + blue → white

Adding more light increases the total amount of light reaching the eye, so the mixture generally becomes brighter.

Examples:

• computer screens
• televisions
• projectors
• stage lighting

Subtractive Colour Mixing

Subtractive mixing occurs when pigments, paints or inks are combined.

Pigments absorb particular wavelengths and reflect others.

As more pigments are added:

• more wavelengths are absorbed
• less light is reflected
• the colour becomes darker

For example:

blue paint + yellow paint → green

The pigments absorb different wavelengths, leaving mainly green wavelengths reflected back to the eye.

Additive vs Subtractive Mixing

	Additive	Subtractive
Involves	Light	Pigments
Primary colours	RGB	CMY
More mixing causes	More light	More absorption
Overall result	Brighter	Darker
All primaries	White	Very dark / near black
Example	Screen	Paint

Photoreceptors and the Retina

The retina contains photoreceptors that detect incoming light and convert it into neural signals.

There are two main photoreceptor types:

Rods

• highly sensitive to light
• useful in dim conditions
• do not support colour vision
• provide relatively poor spatial detail
• concentrated more heavily in the peripheral retina

Cones

• work best in bright light
• responsible for colour vision
• provide fine visual detail
• concentrated in the fovea
• exist in three different types

Rods vs Cones

Feature	Rods	Cones
Light level	Dim	Bright
Colour	No	Yes
Detail	Low	High
Number of types	1	3
Main retinal location	Peripheral retina	Fovea

Scotopic, Photopic and Mesopic Vision

Scotopic Vision

Vision under low-light conditions.

• mainly uses rods
• high sensitivity
• poor colour vision
• poor fine detail

Photopic Vision

Vision under bright-light conditions.

• mainly uses cones
• strong colour discrimination
• high visual detail

Mesopic Vision

Intermediate lighting conditions, such as dusk.

• both rods and cones are active
• provides some colour and some detail

The Three Cone Types

Humans normally have three types of cones.

Cone type	Most sensitive to	Approximate peak
S cone	Short wavelengths	~420 nm
M cone	Medium wavelengths	~530 nm
L cone	Long wavelengths	~560 nm

The cone sensitivity curves overlap.

This is important because colour perception does not depend on one cone responding to one specific colour.

Instead, the brain compares the relative activity across all three cone types.

Trichromatic Theory

The Young–Helmholtz trichromatic theory proposes that colour vision is based on three types of photoreceptors.

Thomas Young first proposed that a small number of receptors could account for many colours. Helmholtz later developed this into the idea of three receptor types.

Modern physiology later confirmed the existence of S, M and L cones.

Population Coding

Colour is encoded through the pattern of activity across the three cone populations.

For example, yellow may produce:

• strong L activity
• strong M activity
• weaker S activity

The brain interprets this overall pattern as yellow.

Therefore:

Colour depends on the relative response across multiple cones, rather than the activity of one cone alone.

Principle of Univariance

The principle of univariance states that a single photoreceptor cannot distinguish between:

• a change in wavelength
• a change in intensity

For example, strong firing from one cone could be caused by:

• its preferred wavelength at moderate intensity
• another wavelength presented at greater intensity

Therefore, a single photoreceptor cannot uniquely identify colour.

The visual system solves this problem by comparing the activity of multiple cone types.

Metamerism

Metamers are physically different light stimuli that appear to be the same colour.

Example:

• pure yellow light
• a mixture of red and green light

These can look identical if they produce a similar pattern of activity across S, M and L cones.

Metamerism demonstrates that the visual system responds to cone activation patterns, rather than directly identifying every wavelength present in the stimulus.

Colour Matching Experiments

In colour matching experiments, people can normally match a test light using mixtures of three primary lights.

This provided strong evidence for trichromatic theory.

Opponent Process Theory

Trichromatic theory explains colour coding at the cone level, but it does not fully explain how colour is experienced later in the visual system.

Hering proposed that colour is organised into opponent pairs:

• red ↔ green
• blue ↔ yellow
• black ↔ white

Colours at opposite ends of the same channel are treated as mutually exclusive.

For example:

• reddish-yellow is possible
• bluish-green is possible
• reddish-green is not normally perceived
• bluish-yellow is not normally perceived

Dual Process Theory

Hurvich and Jameson proposed that trichromatic theory and opponent process theory describe different stages of the same colour vision system.

Stage 1: Trichromatic Coding

At the photoreceptor level:

• S cones
• M cones
• L cones

encode colour through relative activation.

Stage 2: Opponent Processing

The cone signals are then combined into opponent channels:

• red–green
• blue–yellow
• black–white

This combined explanation is called the dual process theory.

Opponent Colour Channels

The main opponent channels are approximately:

Red–Green Channel

Compares:

L cone activity vs M cone activity

Blue–Yellow Channel

Compares:

S cone activity vs combined L + M activity

Black–White / Luminance Channel

Uses combined:

L + M activity

These comparisons allow the visual system to represent colour differences more efficiently.

Opponent Processing Through the Visual Pathway

Opponent processing occurs at several stages.

Retinal Ganglion Cells

Opponent processing begins in the retina.

Ganglion cells receive cone inputs and compare them to generate colour-opponent signals.

For example:

• red–green
• blue–yellow

Lateral Geniculate Nucleus

The LGN continues processing colour information through different parallel pathways.

Pathway	Main role
Parvocellular	Red–green colour + fine detail
Koniocellular	Blue–yellow colour
Magnocellular	Luminance, motion and rapid brightness changes

Parallel Processing

These pathways allow different visual features to be processed at the same time.

For example:

• colour
• detail
• movement
• brightness

are processed partly separately before being integrated later.

Colour Processing in V1

Colour signals travel from the LGN to the primary visual cortex (V1).

V1 contains several important colour-processing mechanisms.

Single-Opponent Cells

• respond to differences in colour
• useful for signalling the presence of colour within a region

Double-Opponent Cells

• respond to both colour contrast and spatial contrast
• compare colour information across neighbouring regions
• useful for detecting colour edges and boundaries

For example, they can respond strongly where a red region meets a green region.

Double-opponent cells are particularly important for detecting colour boundaries and supporting stable colour perception.

Blobs

Blobs are clusters of colour-selective neurons in V1.

They are specialised for colour processing.

Colour Constancy

Colour constancy is the tendency to perceive an object’s colour as relatively stable even when the lighting changes.

For example, a red object usually continues to look red:

• outside in sunlight
• inside under artificial lighting
• on a cloudy day

even though the wavelengths reaching the eye are different.

Why Colour Constancy Is Necessary

The light reaching the eye depends on both:

illumination + properties of the surface

Therefore, the visual system has to estimate whether a change in the incoming light is caused by:

• the object itself
• the lighting conditions

Land’s Mondrian Experiments

Land’s experiments demonstrated the importance of context in colour perception.

Isolated Colour Patch

When a coloured patch is viewed alone:

• changes in illumination cause larger changes in perceived colour
• there is little contextual information available

Patch Within a Complex Scene

When the same patch is surrounded by other coloured patches:

• perceived colour remains more stable
• the visual system can compare the patch with surrounding areas

This suggests that colour constancy depends strongly on relative comparisons across the visual scene.

V4 and Colour Constancy

Higher visual areas, particularly V4, are strongly involved in colour constancy.

V4:

• integrates information from across the visual scene
• helps estimate the colour of the illumination
• compares colours across different regions
• helps maintain stable colour perception despite changing light

V4 responses are more stable across changes in illumination than responses in earlier visual areas.

Limits of Colour Constancy

Colour constancy is effective, but not perfect.

When lighting information is ambiguous, the brain may make different assumptions about the illumination.

This can produce perceptual illusions.

The Dress Illusion

People viewing the same photograph may perceive the dress as:

• blue/black
• white/gold

The visual system is making different assumptions about the lighting in the image and compensating accordingly.

This shows that colour perception depends partly on interpretation and context, not just the wavelengths reaching the eye.

Individual Differences in Colour Vision

People do not all have identical colour vision.

Differences can arise from:

• genetics
• differences in cone photopigments
• differences in cone sensitivity
• differences in neural processing
• brain injury or disease

Colour Vision Deficiencies

Anomalous Trichromacy

All three cone types are present, but one cone type has an altered sensitivity.

Examples:

• protanomaly — altered L cone system
• deuteranomaly — altered M cone system
• tritanomaly — altered S cone system

This usually causes reduced colour discrimination rather than complete colour loss.

Dichromacy

One cone type is absent.

Condition	Missing cone
Protanopia	L
Deuteranopia	M
Tritanopia	S

This reduces the range of colours that can be discriminated.

Monochromacy

Only one cone type functions, or in rare cases vision relies mainly on rods.

This causes extremely limited colour vision.

Genetics of Colour Vision Deficiencies

The genes responsible for the L and M cone photopigments are located on the X chromosome.

Therefore, many red–green colour deficiencies are X-linked.

They are more common in males because males typically have only one X chromosome.

If the relevant gene on that chromosome is altered, there is no second X chromosome carrying another copy.

Blue–yellow deficiencies involve different genetic mechanisms.

Cerebral Achromatopsia

Colour vision problems do not always originate in the retina.

Cerebral achromatopsia occurs when cortical colour-processing areas are damaged, particularly areas such as V4/V8.

A person may have:

• healthy eyes
• normal cones

but still lose the ability to experience colour because the cortical processing system has been damaged.

Retinal vs Cortical Colour Loss

Retinal colour deficiency	Cerebral achromatopsia
Problem with cones/photopigments	Problem with cortical processing
Often inherited	Usually acquired
Often present from birth	May occur after brain injury
Certain colours confused	Colour perception may be severely lost

Whole Colour Vision Pathway

Light

↓
wavelength + intensity

↓

Retina

S, M and L cones

↓

Trichromatic coding

relative cone activation
population coding
principle of univariance
metamerism

↓

Retinal ganglion cells

opponent comparison begins

↓

Opponent channels

red–green
blue–yellow
black–white

↓

LGN

P → red/green + detail
K → blue/yellow
M → luminance + motion

↓

V1

single-opponent cells
double-opponent cells
blobs
colour boundaries

↓

Higher visual areas, especially V4

context + illumination estimation

↓

Colour constancy

stable colour perception despite changing lighting

↓

Conscious colour experience