A set of primary colors is, most tangibly, a set of real colorants or colored lights that can be combined in varying amounts to produce a gamut of colors. This is the essential method used in applications that are intended to elicit the perception of diverse sets of color, e.g. electronic displays, color printing, and paintings. Perceptions associated with a given combination of primary colors are predicted by applying the appropriate mixing model (additive, subtractive, additive averaging etc.) that embodies the underlying physics of how light interacts with the media and ultimately the retina.
Primary colors can also be conceptual, either as additive mathematical elements of a color space or as irreducible phenomenological categories in domains such as psychology and philosophy. Color-space primaries are precisely defined and empirically rooted in psychophysical color matching experiments which are foundational for understanding color vision. Primaries of some color spaces are complete (that is, all visible colors are described in terms of their weighted sums with nonnegative weights) but necessarily imaginary (that is, there is no plausible way that those primary colors could be represented physically, or perceived). Describing primary colors from a phenomenological perspective is difficult to do succinctly, but phenomenological accounts, such as the psychological primaries, have led to practically useful insights.
All sets of real and color-space primaries are arbitrary, in the sense that there is no one set of primaries that can be considered the canonical set. Primary pigments or light sources selected for a given application on the basis of subjective preferences as well as practical factors such as cost, stability, availability etc. Color-space primaries can be subjected to meaningful one-to-one transformations so that the transformed space is still complete and each color is specified with a unique sum.
Elementary art education materials, dictionaries, and electronic search engines often define primary colors effectively as conceptual colors (generally red, yellow, and blue; or red, green, and blue) that can be used to mix "all" other colors and often go further and suggest that these conceptual colors correspond to specific hues and precise wavelengths. Such sources do not present a coherent, consistent definition of primary colors since real primaries cannot be complete.
Additive mixing of light
The perception elicited by multiple light sources co-stimulating the same area of the retina is additive, i.e., predicted via summing the spectral power distributions or tristimulus values of the individual light sources (assuming a color matching context). For example, a purple spotlight on a dark background could be matched with coincident blue and red spotlights that are both dimmer than the purple spotlight. If the intensity of the purple spotlight was doubled it could be matched by doubling the intensities of both the red and blue spotlights that matched the original purple. The principles of additive color mixing are embodied in Grassmann's laws.
Additive mixing of coincident spot lights was applied in the experiments used to derive the CIE 1931 colorspace. The original monochromatic primaries of the (arbitrary) wavelengths of 435.8 nm (violet), 546.1 nm (green), and 700 nm (red) were used in this application due to the convenience they afforded to the experimental work.
Red, green, and blue light are popular primaries for additive color mixing since primary lights with those hues provide a large triangular chromaticity gamut. Small red, green, and blue elements in electronic displays mix additively from an appropriate viewing distance to synthesize compelling colored images.
The exact colors chosen for additive primaries are a technological compromise between the available phosphors (including considerations such as cost and power usage) and the need for large chromaticity gamut. The ITU-R BT.709-5/sRGB primaries are typical.
It is important to note that additive mixing provides very poor predictions of color perception outside the color matching context. Well known demonstrations such as The dress and other examples show how the additive mixing model alone is not sufficient for predicting perceived color in many instances of real images. In general, we cannot completely predict all possible perceived colors from combinations of primary lights in the context of real world images and viewing conditions. The cited examples suggest just how remarkably poor such predictions can be.
Subtractive mixing of ink layers
The subtractive color mixing model predicts the spectral power distributions of light filtered through overlaid partially absorbing materials on a reflecting or transparent surface. Each layer partially absorbs some wavelengths of light from the illumination spectrum while letting others pass through multiplicatively, resulting in a colored appearance. Overlapping layers of ink in printing mix subtractively over reflecting white paper in this way to generate photorealistic color images. The typical number of inks in such a printing process ranges from 3 to 6 (e.g., CMYK process, Pantone hexachrome). In general, using fewer inks as primaries results in more economical printing but using more may result in better color reproduction.
Cyan, magenta, and yellow are good subtractive primaries in that the spectral power distributions of light reflected from idealized inks can be combined for the largest chromaticity gamuts. An additional key ink (shorthand for the key printing plate that impressed the artistic detail of an image, usually black) is also usually used since it is difficult to mix a dark enough black ink using the other three inks. Before the color names cyan and magenta were in common use, these primaries were often known as blue and red, respectively, and their exact color has changed over time with access to new pigments and technologies.
Mixing paints in limited palettes
The color of light (i.e., the spectral power distribution) reflected from illuminated surfaces coated in paint mixes, slurries of pigment particles, is not well approximated by a subtractive or additive mixing model. Color predictions that incorporate light scattering effects of pigment particles and paint layer thickness require approaches based on the Kubelka–Munk equations. Even such approaches cannot predict the color of paint mixtures precisely since small variances in particle size distribution, impurity concentrations etc. can be difficult to measure but impart perceptible effects on the way light is reflected from the paint. Artists typically rely on mixing experience and "recipes" to mix desired colors from a small initial set of primaries and do not use mathematical modelling.
There are hundreds of commercially available pigments for visual artists to use and mix (in various media such as oil, watercolor, acrylic, gouache, and pastel). A common approach is to use just a limited palette of primary pigments (often between four and eight) that can be physically mixed to any color that the artist desires in the final work. There is no specific set of pigments that are primary colors, the choice of pigments depends entirely on the artist's subjective preference of subject and style of art as well as material considerations like lightfastness and mixing heuristics. Contemporary classical realists have often advocated that a limited palette of white, red, yellow, and black pigment (often described as the "Zorn palette") is sufficient for compelling work.
A chromaticity diagram can illustrate the gamut of different choices of primaries, for example showing which colors are lost (and gained) if you use RGB for subtractive color mixing (instead of CMY).
A contemporary description of the color vision system provides an understanding of primary colors that is consistent with modern color science. The human eye normally contains only three types of color photoreceptors, known as long-wavelength (L), medium-wavelength (M), and short-wavelength (S) cone cells. These photoreceptor types respond to different degrees across visible electromagnetic spectrum. The S cone response is generally assumed to be negligible at long wavelengths greater than about 560 nm while the L and M cones respond across the entire visible spectrum. The LMS primaries are imaginary since there is no visible wavelength that stimulates only one type of cone (i.e., humans cannot normally see a color that corresponds to pure L, M or S stimulation). The LMS primaries are complete since every visible color can be mapped to a triplet specifying the coordinates in LMS color space.
The L, M and S response curves (cone fundamentals) were deduced from color matching functions obtained from controlled color matching experiments (e.g., CIE 1931) where observers matched the color of a surface illuminated by monochromatic light with mixtures of three monochromatic primary lights illuminating a juxtaposed surface. Practical applications generally use a canonical linear transformation of LMS space known as CIEXYZ. The X, Y, and Z primaries are typically more useful since luminance (Y) is specified separately from a color's chromaticity. Any color space primaries which can be mapped to physiologically relevant LMS primaries by a linear transformation are necessarily either imaginary or incomplete or both. The color-matching context is always three dimensional (since LMS space is three dimensional) but more general color appearance models like CIECAM02 describe color in six dimensions and can be used to predict how colors appear under different viewing conditions.
Thus for trichromats like humans, we use three (or more) primaries for most general purposes. Two primaries would be unable to produce even some of the most common among the named colors. Adding a reasonable choice of third primary can drastically increase the available gamut, while adding a fourth or fifth may increase the gamut but typically not by as much.
Most placental mammals other than primates have only two types of color photoreceptor and are therefore dichromats, so it is possible that certain combinations of just two primaries might cover some significant gamut relative to the range of their color perception. Meanwhile, birds and marsupials have four color photoreceptors in their eyes, and hence are tetrachromats. There is one scholarly report of a functional human tetrachromat.
The presence of photoreceptor cell types in an organism's eyes do not directly imply that they are being used to functionally perceive color. Measuring functional spectral discrimination in non-human animals is challenging due to the difficulty in performing psychophysical experiments on creatures with limited behavioral repertoires who cannot respond using language. Limitations in the discriminative ability of shrimp having twelve distinct color photoreceptors have demonstrated that having more cell types in itself need not always correlate with better functional color vision.
The opponent process is a color theory that states that the human visual system interprets information about color by processing signals from cones and rods in an antagonistic manner. The theory states that every color can be described as a mix along the three axes of red vs. green, blue vs. yellow and white vs. black. The six colors from the pairs might be called "psychological primary colors", because any other color could be described in terms of some combination of these pairs. Although there is a great deal of evidence for opponency in the form of neural mechanisms, there is currently no clear mapping of the psychological primaries to neural substrates.
The three axes of the psychological primaries were applied by Richard S. Hunter as the primaries for the colorspace ultimately known as CIELAB. The Natural Color System is also directly inspired by the psychological primaries.
There are numerous competing primary colour systems throughout history. Scholars and scientists engaged in debate over which hues best describe the primary color sensations of the eye. Thomas Young proposed red, green, and violet as the three primary colors, while James Clerk Maxwell favoured changing violet to blue. Hermann von Helmholtz proposed "a slightly purplish red, a vegetation-green, slightly yellowish, and an ultramarine-blue" as a trio. In modern understanding, human cone cells do not correspond precisely to a specific set of primary colors, as each cone type responds to a relatively broad range of wavelengths.
- Beran, Ondrej (2014). "The Essence (?) of Color, According to Wittgenstein". From the ALWS archives: A selection of papers from the International Wittgenstein Symposia in Kirchberg am Wechsel.
- Bruce MacEvoy. "Do 'Primary' Colors Exist?" (imaginary or imperfect primaries section Archived 2008-07-17 at the Wayback Machine.). Handprint. Accessed 10 August 2007.
- Goldstein, E. Bruce; Brockmole, James (2018). Sensation and Perception. Cengage Learning. p. 206. ISBN 978-1-305-88832-6.
- "Color". www.nga.gov. Retrieved 10 December 2017.
- "primary color | Definition of primary color in US English by Oxford Dictionaries". Oxford Dictionaries | English. Retrieved 10 December 2017.
- "Definition of PRIMARY COLOR". www.merriam-webster.com. Retrieved 10 December 2017.
- "Wolfram|Alpha – Primary colors". www.wolframalpha.com. Retrieved 10 December 2017.
- Westland, Stephen (2016). Handbook of Visual Display Technology | Janglin Chen | Springer (PDF). Springer International Publishing. p. 162. Retrieved 12 December 2017.
- Reinhard, Erik; Khan, Arif; Akyuz, Ahmet; Johnson, Garrett (2008). Color imaging : fundamentals and applications. Wellesley, Mass: A.K. Peters. pp. 364–365. ISBN 978-1-56881-344-8. Retrieved 31 December 2017.
- Fairman, Hugh S.; Brill, Michael H.; Hemmendinger, Henry (February 1997). "How the CIE 1931 color-matching functions were derived from Wright-Guild data". Color Research & Application. 22 (1): 11–23. doi:10.1002/(SICI)1520-6378(199702)22:1<11::AID-COL4>3.0.CO;2-7.
- Fairchild, Mark. "Why Is Color - Short Answers - Q: Why are red, blue, and green considered the primary colors?". Color Curiosity Shop. Retrieved 4 September 2018.
- Thomas D. Rossing & Christopher J. Chiaverina (1999). Light science: physics and the visual arts. Birkhäuser. p. 178. ISBN 978-0-387-98827-6.
- Kircher, Madison Malone. "This Baffling Picture of Strawberries Actually Doesn't Contain Any Red Pixels". Ney York Magazine. Retrieved 21 February 2018.
- MacEvoy, Bruce. "subtractive color mixing". Handprint. Retrieved 7 January 2018.
- Frank S. Henry (1917). Printing for School and Shop: A Textbook for Printers' Apprentices, Continuation Classes, and for General use in Schools. John Wiley & Sons.
- Ervin Sidney Ferry (1921). General Physics and Its Application to Industry and Everyday Life. John Wiley & Sons.
- Nyholm, Arvid (1914). "Anders Zorn: The Artist and the Man". Fine Arts Journal. 31 (4): 469. doi:10.2307/25587278.
- Kubelka, Paul; Munk, Franz (1931). "An article on optics of paint layers" (PDF). Z. Tech. Phys. 12: 593–601.
- MacEvoy, Bruce. "Mixing Green". Handprint. Retrieved 24 October 2017.
- Bruce, MacEvoy. "The Artists' "Primaries"". Handprint. Retrieved 24 October 2017.
- Gurney. "The Zorn Palette". Gurney Journey. Retrieved 27 September 2016.
- Steven Westland, "subtractive mixing – why not RGB?", October 4, 2009 http://colourware.org/2009/10/04/subtractive-mixing-why-not-rgb/ Archived 2017-09-19 at the Wayback Machine.
- Stockman, Andrew; Sharpe, Lindsay, T. (2006). "Physiologically-based colour matching functions" (PDF). Proceedings of the ISCC/CIE Expert Symposium '06: 75 Years of the CIE Standard Colorimetric Observer: 13–20.
- Best, Janet (2017). Colour Design: Theories and Applications. p. 9. ISBN 978-0-08-101889-7.
- Jordan, G.; Deeb, S. S.; Bosten, J. M.; Mollon, J. D. (20 July 2010). "The dimensionality of color vision in carriers of anomalous trichromacy". Journal of Vision. 10 (8): 12–12. doi:10.1167/10.8.12.
- Morrison, Jessica (23 January 2014). "Mantis shrimp's super colour vision debunked". Nature. doi:10.1038/nature.2014.14578.
- Conway, Bevil R. (12 May 2009). "Color Vision, Cones, and Color-Coding in the Cortex". The Neuroscientist. 15 (3): 274–290. doi:10.1177/1073858408331369.
- Cohen, Jonathan; editors, Mohan Matthen, (2010). Color ontology and color science (New ed.). Cambridge, Massachusetts: MIT Press. pp. 159–162. ISBN 978-0-262-51375-3.
- Maffi, ed. by C.L. Hardin [and] Luisa (1997). Color categories in thought and language (1. publ. ed.). Cambridge: Cambridge University Press. pp. 163–192. ISBN 978-0-521-49800-5.
- Edward Albert Sharpey-Schäfer (1900). Text-book of physiology. 2. Y. J. Pentland. p. 1107.
- Alfred Daniell (1904). A text book of the principles of physics. Macmillan and Co. p. 575.