Practical Problems of Specifying and Measuring Colour-Rendering Properties

Abstract

The visual sensations involved in the perception of colour-rendering cannot be measured directly, but subjective experience shows some correlation, over limited ranges, with the spectral energy distribution (spectral radiant power) of the light incident on the surfaces and objects illuminated, when they are viewed under strictly defined conditions. The practical problems which arise in attempts to specify and measure these objective characteristics of the stimulus which are related to colour-rendering, are reviewed and appraised; in particular, attention is drawn to the difficulties of accurate spectro-radiometry, the maintenance of light standards, and the complex nature of the colour shift method now preferred by the USA and Germany.Suggestions are made bearing on the possible proposals to the CIE in 1963 concerning revision of the CIE 1948 provisional recommendations in relation to the measurement and specification of colour-rendering. An interim modified spectral band system is suggested, which would include the ultra-violet region of the spectrum because of its importance in relation to the colour appearance of objects and materials which are fluorescent; colour-rendering in the objective sense would be specified in terms of chromaticity and relative luminance (as in the CIE 1948 system) in the visible bands, with an appropriate weighting factor for the ultra-violet radiation. Tolerances would be based on the Crawford system which takes account of the effects in contiguous bands.Because of the importance of chromaticity and spectral energy distribution measurements in relation to colour and colour-rendering, reference is made to the accuracy currently achieved, both in the standardising laboratory and when using commercial apparatus and techniques at present available to industry. Possible future developments are also indicated.