 1931 Commision Internationale d'Elairage (CIE) chromaticity system Trichromatic color matches using three colors can be illustrated on Newton's color circle (figure 23). Newton's color circle consists of the following components: a circle representing the spectral colors (although not shown here, mixtures of blue and red (purple locus are not spectral colors and hence a straight line between R and B is more appropriate) a triangle whose vertices represent the three primary colors used to make color matches (R, G and B) the centre of the circle representing white (W) Figure 23. Newton's color circle Newton's color circle provides a qualitative description of color matches and can be used to explain why two colors may not be sufficient to make color matches and also the use of 'negative' colors. For example, if 500 nm is required to be matched (spectral color located on the circle), blue and green will be required. However, blue and green primaries alone will produce a desaturated 500 nm. Therefore, red must be added to the spectral color to desaturate it and make the match. When the third primary is added to desaturate the color mixture, negative tristimulus values result. Figure 24. The use of Newton's color circle to illustrate matching a spectral color of 500 nm using the three primates In order to deal with the 'negative' colors, the CIE devised the XYZ system that uses unreal (imaginary) primaries to describe color space. The 1931 CIE chromaticity system chose three imaginary primaries (reference stimuli) X, Y and Z, so that all spectral loci lying inside this triangle will be positive. The alychne are locus of colors with no luminosity and this was chosen to lie along the X to Z on the XYZ chromaticity system. All luminosity is expressed in Y. The reference loci of Y was chosen to just enclose the domain of real colors. Equi-energy white was chosen to have equal chromaticity co-ordinates, that is, 0.33, 0.33 (figure 25 and 26). Chromaticity co-ordinates represents the relative contribution of the three primaries, the sum of the co-ordinates equals 1.0. Therefore, z can be calculated, by knowing the co-ordinates x and y since x + y + z = 1. Figure 25. 1931 CIE chromaticity diagram. The triangle represents the three primaries used in the RGB system (From Le Grand, Y. Light, Color and Vision, 2nd ed. London: Chapman and Hall, 1968) Figure 26. 1931 CIE chromaticity diagram with approximate color representation (From Benjamin, W. J. (Ed), Borish's Clinical Refraction. Philadelphia: W. B. Saunders Company, 1998) Dominant wavelength, complementary wavelength and excitation purity can be easily located for a sample. The dominant wavelength represents the principle wavelength of the color. The complementary wavelength is the wavelength that produces white when mixed in appropriate portions with the dominant wavelength. Spectral complementaries can be found when a line joined by a lines which passes through the achromatic point shown as C (figure 27). The dominant wavelength of A is given by the spectral wavelength at DA and the complementary by the wavelength at CA (figure 27). Point C identifies the location of the white point and B identifies another wavelength, that when mixed at the appropriate proportions, will produce white. Figure 27. Complementary and dominant spectral wavelengths of color A. Color B is also complementary to color A, since an appropriate mix of the two wavelengths will produce white The achromatic point varies depending upon the standard illuminant that is used (figure 28). A shift in the x and y co-ordinates occurs as color temperature increases. For standard illuminant C, there is no complimentary wavelengths for green (between wavelengths 492 nm to 567 nm) However, white light can be formed with a suitably chosen purple light (figure 28). Figure 28. Variation position of the achromatic point according to color temperature. (From Benjamin, W. J. (Ed), Borish's Clinical Refraction. Philadelphia: W. B. Saunders Company, 1998)