James A. Worthey

Opponent-colors approach to color rendering

Starting with an opponent-colors formulation of color vision, two parameters, t and d, may be defined that express an illuminants ability to realize red-green and blue-yellow contrasts of objects. For instance, calculation of t and d for daylight shows that on a gray day, color contrasts are actually reduced. By these measures, many common vapor-discharge illuminants systematically distort object colors. Because red-green contrasts contribute to border distinctness, and both types of color contrast contribute to brightness, such systematic distortions probably affect the overall clarity and brightness of what is perceived visually, Experimental data are consistent with this idea. In relation to color-constancy (retinex) experiments, it is approximately true that the visual system discounts the color of an illuminant but not its t and d.

Refinements of formulas for receptor sensitivity

When it is desired to fit a polynomial to a function of wavelength, roundoff errors may be minimized by shifting the origin of the independent variable (wavelength or frequency) to the center of the region of interest.

Visual Clarity With A Black-And-White Scene.

Visual clarity experiments are usually done with colorful test objects, and it is generally concluded that the results of such experiments are related to the color-rendering properties of the illuminants involved. Nonetheless, it has been observed that a clarity difference between illuminants may be seen, even with black-and-white objects. An experiment was performed to measure differences of perceived clarity using only black-and-white fabric and black yarn as test objects. (The word “clarity” was not used in the instructions to subjects. They were asked questions concerning “preferences” and “blackness.”) The differences measured seem to indicate a role for color in black-and-white vision, but not a pure clarity effect independent of illuminant color.