Color Perception: Physiological

Cone Receptors

Human color vision is characterized by three different classes of cone photoreceptors, denoted long- (L), middle- (M), and short-wavelength-sensitive- (S) cones. Light is absorbed in the cones and converted by a complex photochemical reaction into an electrical signal, whose magnitude is determined by the number of photons absorbed by each cone. Figure 1 shows the most recent and precise estimates of the cone absorption spectra. Thus, although the light stimulus is determined by its intensity at an infinite number of wavelengths between approximately 400 and 700 nm, the output of the cones can be characterized by only three numbers—the absorption for each cone class. This is the principle of trichromacy. The three classes of cone therefore drastically reduce the dimensionality of color vision. Several characteristics about the human cone absorption spectra are worth noting.

Figure 1 shows that all types of cone are sensitive to a wide range of wavelengths. L- and M-cones are sensitive nearly over the whole visible spectrum. The L-cones have their peak sensitivity at a wavelength that would appear yellowish and the S cones have their peak sensitivity at a wavelength that would appear violet under neutral viewing conditions. Therefore, the common labeling of the cones as red, green, and blue is misleading because these labels do not correspond to the wavelength to which the cones are maximally responsive. Thus, there is no simple relationship between the excitation of a single class of photoreceptors and the color we perceive.

Instead, photons of different wavelengths have different likelihoods of being absorbed by the three cone classes. Once absorbed, the only remaining information is the photon count in each cone, not the wavelength of the absorbed photons, a principle termed univariance. An increase in the photon count of a photoreceptor can thus result from an increase in light intensity, a change to a more favorable wavelength for that cone class, or both. Therefore, to compute the color of an object unambiguously, the magnitudes of the output signals of the three cone classes have to be compared. This starts at the [Page 272]next stage of processing, which is performed by bipolar cells and ganglion cells in the retina.

...