Progress in characterizing the multidimensional color quality properties of white LED light sources

摘要

With the introduction of solid state light sources, the variety in emission spectra is almost unlimited. However, the set of standardized parameters to characterize a white LED light source, such as correlated color temperature (CCT) and CIE general color rendering index (R_a), is known to be limited and insufficient for describing perceived differences between light sources. Several characterization methods have been proposed over the past decades, but their contribution to perceived color quality has not always been validated. To gain more insight in the relevant characteristics of the emission spectra for specific applications, we have conducted a perception experiment to rate the attractiveness of three sets of objects, including fresh food, packaging materials and skin tones. The objects were illuminated with seven different combinations of Red, Green, Blue, Amber and White LEDs, all with the same CCT and illumination level, but with differences in R_a and color saturation. The results show that, in general, object attractiveness does not correlate well with R_a, but shows a positive correlation with saturation increase for two out of three applications. There is no clear relation between saturation and skin tone attractiveness, partly due to differences in preference between males and females. A relative gamut area index (G_a) represents the average change in saturation and a complementary color vector graphic shows the direction and magnitude of chromatic differences for the eight CIE-1974 test-color samples. Together with the CIE general color rendering index (R_a) they provide useful information for designing and optimizing application specific emission spectra.