Numerical Modeling of Color

EXTENDED ABSTRA. NUMERICAL MODEliNG OF COLOR S. Eskinazi, P.O. Box 91765, Tucson, AZ 85752, U.S.A. Accepted byRogerF. Malina. Although the ocular perception we call color is totally physical in its emanation and structure, it continues to remain for many a personal experience-an intimate sensation and, therefore, far removed from an immutable and universal order. There is good reason for this: the imperfect human eye sees in a color what the brain says it sees. Since the concept ofvisual literacy is almost non-existent in our culture, the perception of color, and forms to some extent, remains totally arbitrary and often shamefully chaotic. According to J.J. Gibson, the meaning of the term color is one of the worst muddles in the history of science. The old masters spoke of color systems based on psychological primaries, a concept used by Leonardo da Vinci. Each artist or school had its preferred system. The Impressionists benefitted from the scientific foundations of color, which were initiated by great physicists such as Isaac Newton, James Maxwell and Ludwig Helmholtz, who performed many physical experiments with color and began the establishment of a physical theory of color. When the new field of color dynamics was born, artists began to speak of hue, saturation, intensity, color scales, chromatic and achromatic scales, etc. Suddenly color appeared to be on the way of acquiring quantitative dimensions and, with those dimensions, new expressions arose, some pleasing and harmonious, some shocking, and yet others leaving the viewer totally indifferent . A century ago, the Bostonian artist and scientist A.H. Munsell was deeply concerned about the irrational understanding of color and devised a semiquantitative system that remained in existence for many years. Because his system lacked an absolute scientific basis , it kept changing as new dyes were discovered. What began as a cylindrical model similar to the one shown in Fig. 1 constantly evolved as an unwieldly tree with different sized branches. Ultimately the system lost its popularity. There is no need for color to remain subjective. A rational definition of color in no way threatens creativity. On the contrary, it helps convey exactly a color message and its relationship to other colors. No single word or parameter can properly or fully describe it. It is not enough to say "red" or "pinkish red" for others to understand precisely what is meant. The artist is not the exclusive user of color. There are many applications, such as photography, television, color Fig. 1. Cylindrical color model. The cylinder is divided into 12 parts, so that an equal mixture of red and green results in yellow ; green and blue results in cyan; and blue and red results in magenta. Hues between two colors are the average value of mixing the two adjacent colors. Saturation increases from the center of the cylinder to the outside. Luminosity increases from one end of the cylinder to the other. computers and textiles, that stand to benefit from a unified system. Although there have been attempts at rationalizing color in each of these fields, success has remained for the most part specialized and non-interdisciplinary. A viable universal model must possess the following qualities: First, it must describe fully the phenomenon it is meant to model. Second, its validity must be widely encompassing. And third, it must remain integral and independent of ancillary factors that appear to have a bearing on the phenomenon but are part of separate considerations. Designing a model with the goal of encompassing every possible purpose and application is nearly impossible, or at least it would be so complex that its use would become prohibitive to many users . For that simple reason the intransigent attributes of dyes and the human retina must be kept out of consideration . The principal object of rationalizing color was not necessarily to improve coloristic techniques but to lay down a rational framework of the structure and dynamics of color-the relationship that binds all colors in a single framework and eventually enhances a universal coloristic appreciation. e 19961SAST LEONARDO, Vol. 29, No. l,pp. 75-76, 1996 75 A minimum of three independent attributes (normally the color primaries) are necessary to define a color. Anyone of...