Specification of Uniform Color Tolerances for Textiles

Abstract

Work on methods of specifying uniform color tolerances for various materials has been going on at the National Bureau of Standards for several years. This paper describes the progress of this work, particularly as it relates to textile materials which offer special complications of their own. A formula for the purpose of expressing the size of color differences is derived for observing conditions characteristic of the color inspection of textiles. This formula is based upon the standard observer recommended in 1931 by the International Commission on Illumination and upon a co-ordinate system (the uniform-chromaticity-scale system) intended to give , equal spacing for equally perceptible chromaticity differences. An instrument designed particularly for the measurement of small chromaticity differences is described, together with its calibration in terms of the uniform-chromaticity-scale system. The chromaticity differences between 15 representative pairs of dyed woolen textiles supplied by the duPont Technical Laboratory have been measured by means of this instrumecnt ; lightness differences for the 15 pairs were evaluated by measurements of luminous apparent reflectance with a photoelectric reflectometer, and the color-difference formula has been applied to combine the chromaticity and lightness differences to form computed values of the color difference. These computed differences are compared with those obtained at the duPont Technical Laboratory by direct visual estimate and with differences similarly obtained at the National Bureau of Standards. Results by the formula have also been compared with two other proposals (Nutting, Nickerson) for evaluation of color differences between textile samples. It is concluded that the present formula agrees with direct visual estimate for the samples tested except for certain deep reds and blues. Some of these deviations are irregular, and possible reasons for the failure of the formula to apply to these samples are discussed. A majority of them are regular, however, and suggest how the formula can be improved to make it sufficiently accurate for practical and convenient specification of color tolerances for these colors also.