Encyclopedia of Color Science and Technology

2016 Edition
| Editors: Ming Ronnier Luo

Color Vision, Opponent Theory

Reference work entry
DOI: https://doi.org/10.1007/978-1-4419-8071-7_92



Opponency in human color vision refers to the idea that our perceptual color mechanisms are arranged in an opponent fashion. One mechanism, the red-green mechanism, signals colors ranging from red to green; the other one, the yellow-blue mechanism, signals colors ranging from yellow to blue. This opponency is often referred to as hue opponency, as opposed to cone opponency.

Behavioral Evidence for Color-Opponent Processing

Hering [1] was the first to notice that some pairs of colors, namely, red and green and yellow and blue, cannot be perceived at the same time. He named these pairs of colors “Gegenfarben” [opponent colors] since they are mutually exclusive colors; in Hering’s original figure (Fig. 1), this mutual exclusivity is conveyed by the lack of overlap between red and green and between yellow and blue. The idea is that these opponent colors constitute the end points of the two chromatic mechanisms. For example, the putative yellow-blue mechanism signals colors from yellow to blue; if the stimulus contains neither yellow nor blue, then this stimulus elicits no response in this mechanism (Fig. 2) and this mechanism is at an equilibrium. The colors for which the yellow-blue mechanism is at an equilibrium are called “unique red” or “unique green,” depending on which side of the neutral point they lie. Colors that silence the red-green opponent mechanisms are called “unique yellow” or “unique blue.” Hurvich and Jameson [2] were the first to use a hue cancellation procedure to find the equilibria points (null responses) of these opponent mechanisms.
Color Vision, Opponent Theory, Fig. 1

Hering [1], page 42 ibid. The opponent colors, red versus green and yellow versus blue, are mutually exclusive. This idea is conveyed in Hering’s original figure by showing that there is no overlap between red and green and between yellow and blue

Color Vision, Opponent Theory, Fig. 2

The response of a putative yellow-blue mechanism is shown. When this mechanism is at equilibrium, i.e., produces zero output in response to a stimulus, then this stimulus is – by definition – perceived as “neither blue nor yellow.” Stimuli that are perceived as “neither blue nor yellow” are defined as “unique red” or “unique green,” depending on which side of the neutral gray point they are located

Linearity and Constancy of the Color-Opponent Mechanisms

Krantz and colleagues [3, 4] tested the linearity of these color-opponent mechanisms and concluded that the red-green opponent mechanism is approximately linear but the yellow-blue mechanism exhibited significant deviations from linearity and additivity. These basic results has been replicated by numerous studies, either using a hue cancellation procedure [5, 6] or a modified hue selection task [7], as shown in Fig. 3. Here the task of the observer was to chose that patch of light that appears “neither yellow nor blue” (to obtain unique red and green) or “neither red nor green” (to obtain unique yellow and blue). Figure 4 shows the unique hue settings for 185 color-normal observers using the hue selection task [8, 9], plotted in an approximately uniform u′,v′ chromaticity diagram. All experiments were conducted under three different viewing conditions: either under dark conditions (only source of illumination was the gray monitor background), under adaptation to D65 (daylight simulator), or under adaptation to typical office light (CWF). The symbols and the solid lines (first principal component) denote the unique hues under the dark viewing condition; dashed and dotted lines are the first principal components under adaptation to D65 or CWF.
Color Vision, Opponent Theory, Fig. 3

Modified hue selection task to obtain unique red [7]. The task of the observer is to identify the patch which appears “neither yellow nor blue.” Stimuli that are perceived as “neither blue nor blue” are defined as “unique red” or “unique green,” depending on which side of the neutral gray point they are located

Color Vision, Opponent Theory, Fig. 4

Unique hue settings for 185 color-normal observers are plotted in an approximately uniform u′,v′ diagram. A hue selection task was used to obtain the unique hues [7]. Symbols are proportional to the number of data points and are only shown for the dark viewing condition for clarity. For each unique hue, a total of 1,665 data points are shown (185 observers × 9 saturation/lightness combinations). The solid, dashed, and dotted lines indicate the first principal component for the dark, D65 (daylight simulator), and CWF (office light) viewing conditions

Figure 4 demonstrates the two major features of the color-opponent mechanisms: (1) Consistent with Krantz and colleagues, the red-green opponent mechanism was found to be approximately linear, that is, unique yellow and blue are colinear [3]; the yellow-blue opponent mechanism on the other hand is not a single linear mechanism [4], that is, unique red and green do not lie on a line through the neutral gray background. Therefore, one needs to either postulate a highly nonlinear yellow-blue mechanism or, which is more likely, two separate unipolar mechanisms, one signaling yellow and the other one signaling blue. (2) The red-green opponent mechanism is fairly color constant in comparison to the yellow-blue opponent mechanism; the unique yellow and blue settings are not affected by the changes in the ambient illumination (solid, dashed, and dotted lines are coincident), while large shifts in unique hue settings are observed for the yellow-blue equilibria. Unique green settings, in particular, shift toward yellow when viewed under typical office light (CWF).

Physiological Basis for Opponent Hue Processing: Hue Opponency Versus Cone Opponency

The physiological basis of these color-opponent (hue-opponent) mechanisms is still unclear. We know that the early cone-opponent mechanisms that originate in the retina and are inherited by the lateral geniculate nucleus [10] are not the neural substrate of the behaviorally measured hue opponency. This is easily seen by replotting the unique hue settings in a cone-opponent space. Figure 5 shows the data in an isoluminant Boynton-MacLeod diagram (BML); the L/(L + M) axis denotes the L-M cone-opponent mechanism; the S/(L + M) denotes the S−(L + M) cone-opponent mechanism. Apart from scaling, the BML diagram and the DKL space (Derrington-Krauskopf-Lennie space) are identical for isoluminant stimuli. The BML diagram is not intended be a uniform chromaticity diagram, but the basic features of the color-opponent mechanisms can also be seen here. (1) It is clear that the color-opponent mechanisms (unique hue lines) are not aligned with the cone-opponent axes (L/(L + M); S/(L + M)). The red-green equilibria (line connecting unique yellow and blue) are not parallel to the S/(L + M) cone axis; the yellow-blue equilibria (line connecting unique red and green) are not aligned with the L/(L−M) axis. In reference to the gray background (“x” in the middle of Fig. 5), unique blue requires, in addition to the S-cone input, also an M-cone input; unique yellow requires a negative S-cone input and a positive L-cone input. Similarly, unique green requires a negative S-cone and a positive M-cone input. Unique red is the only hue that is at least approximately aligned with the L/(L + M) axis; but even here a systematic negative S-cone input is required. This confirms that the early cone-opponent mechanisms (depicted by the axes in Fig. 5) do not constitute the neural mechanisms underlying the observed hue opponency (lines connecting the unique hues). Further recombinations of the early opponent mechanisms must occur between the LGN and the primary visual cortex or within the visual cortical areas.
Color Vision, Opponent Theory, Fig. 5

Unique hue settings for 185 color-normal observers are plotted in a cone-opponent diagram. For details, see Fig. 4


Firstly, the color-opponent mechanisms obtained using behavioral measures such as hue cancellation are not aligned with the cone-opponent mechanisms that have been identified in the retina and the lateral geniculate nucleus. It may therefore be more appropriate to refer to these mechanisms as hue-opponent and cone-opponent mechanisms, respectively. Secondly, the red-green opponent mechanism (yielding unique yellow and blue) is an approximately linear mechanism. In contrast, the yellow-blue opponent mechanism (yielding unique red and green) cannot be modeled as a single linear opponent mechanism since unique red and green do not lie on a line through the neutral gray origin. The most parsimonious explanation is to postulate two separate yellow-blue mechanisms; when at equilibrium, one of them signals red, the other one green. Thirdly, while the red-green opponent mechanism is almost completely color constant (unique yellow and blue settings are invariant under changes in ambient illumination), the equilibria point of the yellow-blue mechanism change under changes in ambient illumination: unique green in particular undergoes a major shift toward yellow when viewed under CWF in comparison to simulated daylight (D65). We speculate that this failure of constancy for unique green might have the same neural origin as the relatively large interobserver variability found in unique green settings [11, 12].



  1. 1.
    Hering, E.: Grundzüge der Lehre vom Lichtsinn. Julius Springer, Berlin (1920)CrossRefGoogle Scholar
  2. 2.
    Jameson, D., Hurvich, L.: Some quantitative aspects of an opponent-colors theory. I. Chromatic responses and spectral saturation. J. Opt. Soc. Am. 45, 546–552 (1955)ADSCrossRefGoogle Scholar
  3. 3.
    Larimer, J., Krantz, D., Cicerone, C.: Opponent-process additivity. I: red/green equilibria. Vision Res. 14, 1127–1140 (1974)CrossRefGoogle Scholar
  4. 4.
    Larimer, J., Krantz, D., Cicerone, C.: Opponent-process additivity. II: yellow/blue equilibria and nonlinear models. Vision Res. 15, 723–731 (1975)CrossRefGoogle Scholar
  5. 5.
    Webster, M.A., Miyahara, E., Malkoc, G., Raker, V.E.: Variations in normal color vision. II. Unique hues. J. Opt. Soc. Am. A 17, 1545–1555 (2000)ADSCrossRefGoogle Scholar
  6. 6.
    Werner, J.S., Wooten, B.R.: Opponent chromatic mechanisms: relation to photopigments and hue naming. J. Opt. Soc. Am. 69, 422–434 (1979)ADSCrossRefGoogle Scholar
  7. 7.
    Wuerger, S.M., Atkinson, P., Cropper, S.J.: The cone inputs to the unique-hue mechanisms. Vision Res. 45, 3210–23 (2005)CrossRefGoogle Scholar
  8. 8.
    Wuerger, S.: Colour constancy across the life span: evidence for compensatory mechanisms. PLoS One 8, e63921 (2013)ADSCrossRefGoogle Scholar
  9. 9.
    Xiao, K., Fu, C., Mylonas, D., Karatzas, D., Wuerger, S.: Unique hue data for colour appearance models. Part II: chromatic adaptation transform. Color. Res. Appl. 38, 22–29 (2013)CrossRefGoogle Scholar
  10. 10.
    Derrington, A.M., Krauskopf, J., Lennie, P.: Chromatic mechanisms in lateral geniculate nucleus of macaque. J. Physiol. 357, 241–265 (1984)CrossRefGoogle Scholar
  11. 11.
    Kuehni, R.G.: Unique hues and their stimuli – state of the art. Color. Res. Appl. 39, 279–287 (2014)CrossRefGoogle Scholar
  12. 12.
    Mollon, J.D., Jordan, G.: On the nature of unique hues. In: Murray, I., Carden, D., Dickinson, C. (eds.) John Daltons colour vision legacy, pp. 381–392. Taylor and Francis, London (1997)Google Scholar

Copyright information

© Springer Science+Business Media New York 2016

Authors and Affiliations

  1. 1.Department of Psychological SciencesUniversity of LiverpoolLiverpoolUK
  2. 2.Department of Psychological SciencesInstitute of Psychology, Health and Society, University of LiverpoolLiverpoolUK