A System of Photometry and Colorimetry Based Directly on Visual Response

Abstract

In order that the CIE system of photometry and colorimetry agree with perceived brightness and with perceived color more closely than at present, some redefinition of the standards of light and color is necessary. A different choice of color-matching functions, but still acceptable within the framework of the 1931 CIE color-matching data, yields excellent agreement with the many experiments on the relationship between perceived brightness and luminance, carried out since 1950. The expression for perceived brightness is simply the sum of three tristimulus values; one is the familiar quantity Y, derived from the y function and representing luminance or achromatic brightness, and the other two represent chromatic brightness and derive from two newly-chosen color-matching functions. The addition of achromatic and chromatic components of brightness is in accord with modern views.

This “perceived brightness” is a practical measure of useful light. The design of photometers is described. If desired, the signals from a Barnes colorimeter, for example, may be used. The new photometer will agree with the footcandle meter if they are measuring light of the same color. If the colors differ, the new photometer will agree closely with what the eye of the typical observer sees, whereas the footcandle meter often will not.

The newly-chosen color-matching functions are expected also to help improve the standard of color, so that even when spectral yellow light is matched in brightness and color to a mixture of spectral green and spectral red, the calculated match will agree with a match by a typical observer. The present color-matching functions fail in this case.

The addition of measured chromatic brightness, to the familiar measured achromatic brightness (lumens), is in the nature of a correction which becomes appreciable when the source colors are rather different. For example, the Cool White (4100K) lumen is about 15 per cent higher in perceived brightness than the Warm White (3000K) lumen, as experiments by Harrington, and Sanders and Wyszecki, showed fifteen years ago. The new system of photometry accurately measures such differences.