Image courtesy of Robert Hirschler
Eugene Allen was born on November 7, 1916, in Newark, NJ to Celia and Mitchell Kaplan. He attended Columbia College in New York, receiving a bachelor’s degree. He transferred to Stevens Institute of Technology where in 1944 he received a master’s degree in mathematics while working at the military research facility Picatinny Arsenal in Dover, NJ. In 1952, he obtained a Ph.D. degree in chemistry from Rutgers University while working, under the guidance of E. I. Stearns, in the Research Department of American Cyanamid Corp. in Bound Brook, NJ, then a textile colorant manufacturer using the latest technologies in colorant quality control. In 1967, he became director of the color science laboratory at Lehigh University’s Center for Surface and Coatings Research from which he retired in 1982.
In 1967, Allen received the Armin J. Bruning Award from the Federation of Societies for Coatings Technology for his efforts in regard to the introduction of color measurement and colorimetry into that business group. He was named a Fellow of the Optical Society of America and in 1983 received the Inter-Society Color Council of America’s top honor, the Godlove Award. His colleagues at Lehigh University considered him to be a Renaissance man “always reaching for wider connections between science, arts and humanities.” Allen died on January 18, 2005 [1].
In the late 1950s, Allen’s main interest was to establish a clear understanding of the mode of operation of fluorescent whitening agents and how their whitening efficiency could be objectively evaluated [2].
84.1 Digital Colorant Formulation
Allen’s primary efforts related to color resulted from the introduction in 1958 of the Davidson and Hemmendinger COMIC (colorant mixture computer), an analog computer for matching reflectance functions of dyed or painted samples with related colorants, a technology that replaced centuries-old color matching by trial and error. Familiar with the technical limitations of analog computers, Allen developed a set of basic mathematical equations for color matching, using matrix algebra and the Kubelka–Munk equations for use in a digital computer. Small digital computers, such as IBM’s 1130 with input via punched cards, began being introduced at the time. Allen’s equations were published in 1966 and became an important basis for widespread industrial use of digitally based computer colorant formulation [3, 4]. Allen proposed matching algorithms for both one- and two-constant Kubelka–Munk data, the former for dyes and the latter for pigments.
The matching procedure is relatively complex. Among the variables are the reflectance function of the standard to be matched, the reflectance functions of the colorants to be used for matching at various concentrations as well as their price, and the reflectance function for the substrate material. An initial solution is then calculated together with the resulting theoretical reflectance function of the match. The tristimulus values for the reference and the match functions are calculated and compared in regard to color difference. If the calculated match formula is not sufficient, the formula undergoes an iteration process until the tristimulus values are sufficiently close, or the colorant combination is abandoned. Of sufficiently close formulations metameric and color constancy indices as well as cost are calculated for a final decision as to which formula is to be used.
- 1.
Metamerism
An important issue in colorant formulation is color metamerism and in particular, metamerism related to small differences in the color vision properties of observers. In 1969, Allen described a mathematical methodology to quantify it and proposed a statistically derived “standard deviate observer” based on the statistical standard deviation in color-matching functions of observers with normal color vision [5]. The standard deviate observer was later incorporated into the CIE colorimetric system.