Color vision

Human color vision begins when cone photoreceptors absorb different portions of the visible spectrum. Most people with typical vision have short-, medium-, and long-wavelength-sensitive cones, often called S, M, and L cones. Each cone type is broadly tuned rather than responsive to only one wavelength, so the visual system infers color by comparing their relative activity.¹ Signals are then transformed by retinal opponent channels, including red–green and blue–yellow comparisons, before reaching visual areas of the brain. This arrangement separates chromatic information from brightness information and helps explain why mixtures of lights can match the appearance of a single spectral light.

Color is therefore a perceptual construction, not a simple label attached to a wavelength. A surface appears colored because it selectively reflects, absorbs, or transmits parts of the illumination; the same surface can produce different receptor signals under different light sources. Color constancy allows objects to retain a relatively stable apparent color across such changes, although it can fail in ambiguous scenes.

Color vision varies widely among species because eyes evolved under different ecological pressures. Many mammals are dichromatic, while numerous primates are trichromatic; birds, reptiles, amphibians, fish, and many insects may possess additional cone classes or sensitivity extending into ultraviolet wavelengths.² More receptor types do not automatically produce a proportionally larger perceptual color vocabulary: neural wiring, behavior, and the animal's environment determine how signals are used. Some species also have oil droplets or other optical filters that sharpen spectral distinctions.

Machine color vision measures reflected or emitted radiation with cameras containing several channels. A conventional digital camera uses three broad channels modeled loosely on human cone responses, whereas multispectral and hyperspectral instruments record more bands. Such systems support industrial inspection, remote sensing, medical imaging, and robotic navigation, but interpretation depends on calibration, illumination, sensor noise, and the chosen color space. A camera's numerical color classification is not evidence that it experiences color as an animal does.

Color-vision deficiency results when cone photopigments or their neural pathways are absent, altered, or functioning atypically. The common inherited forms affect red–green discrimination and arise mainly from variation in genes encoding the L- and M-cone photopigments; blue–yellow deficiency is rarer, and complete loss of cone function is uncommon.³ The familiar term “color blindness” can therefore be misleading: many affected people see a broad range of colors but confuse particular hues or perceive them with reduced differences.

Deficiency may also be acquired through retinal disease, optic-nerve damage, neurological injury, aging, or medication. Clinical tests use standardized colored targets or computer displays, but performance depends on lighting and test design. Everyday adaptations include redundant cues such as position, brightness, texture, and labels. Some digital accessibility tools alter contrast or color mappings; they cannot restore missing photoreceptor signals, but they can make distinctions more usable.

Color vision has several important edge cases that are easy to overlook. Two lights with different spectral compositions can appear identical to a human observer if they produce the same cone responses; such pairs are called metamers. Conversely, a color difference can become visible only under a particular illuminant, a principle used in detecting optical brighteners, inks, and counterfeit materials. The apparent color of a stimulus also changes with surrounding colors through simultaneous contrast and other contextual effects.¹

Color signals are not confined to the central fovea: peripheral vision generally has fewer cones and poorer chromatic resolution, while rods dominate in dim light and provide little conventional color discrimination. Some people possess four potentially distinct cone photopigments, a condition called tetrachromacy, yet extra receptors do not necessarily yield a consciously expanded color range. In research and engineering, these distinctions separate spectral measurement, discrimination ability, and subjective color experience.

Cone photoreceptor

A retinal cell specialized for vision in brighter conditions and for comparing wavelengths of light.

Opponent process

Neural coding in which color signals are represented as opposing channel differences, such as red versus green.

Metamer

One of two stimuli with different spectral compositions that appear identical under specified viewing conditions.

Tetrachromacy

Color vision based on four potentially distinct classes of cone photoreceptors rather than the usual three in humans.

Color constancy

The tendency to perceive an object's color as relatively stable when illumination changes.