Abstract
The dependence of the General Colour Rendering Index (GCRI) on the specific choice of test colours, standard observer and colour difference formula was investigated using the spectral power distribution (spd) of 29 fluorescent lamps. The conclusion is drawn that the substitution of the CIE 1931 colorimetric observer by the CIE 1964 supplementary observer or of the 1964 CIEUVW colour difference formula by the 1978 CIELUV formula does not significantly influence the ranking of light sources with respect to their GCRI's. Substitution of CIEUVW by CIELAB may have some effect on this ranking although this effect is less than one would expect from the correspondence between CIELAB and CIEUVW colour differences. Replacement of a set of 94 test colours by the CIE set of 8 test colours or by a set of 13 saturated ones has only a minor effect on the ranking. Therefore, there is no reason to change the CIE set of test colours as long as the colour rendering properties of light sources are expressed by a single number, the GCRI. It is shown that we can gain more information on the colour rendering properties of a light source by considering the frequency distribution of the individual CRI's of a large set of test colours. In this investigation special attention was given to the three-band fluorescent lamps. The optimum properties of these lamps appear to depend only slightly on the specific choice of test colours, standard observer and colour difference formula.
Get full access to this article
View all access options for this article.
