| Abstract | As light-emitting diode (LED) light sources mature, lighting designers will be able to deliver white light with a variety of spectral power distributions and a variety of color rendering properties. This experiment examined the effects of three spectral power distributions (SPDs) that were matched in illuminance and chromaticity on three measures of color perception: one objective (performance on the Farnsworth-Munsell 100 hue test) and two subjective (judgments of the attractiveness of one's own skin, and preferences for the saturation of printed images). The three SPDs were a quartz-halogen (QH) lamp and two LED sources that were matched to the QH lamp in terms of both illuminance and chromaticity; the three light sources were nominally CCT = 3500 K, x = 0.40, y = 0.39 and ∼ 400 lx. LED A used three channels (red, green, blue), and had very poor color rendering (Ra = 18). LED B used four channels (red, amber, cyan, white) and had very good color rendering (R a = 96, whereas the QH had Ra = 98). Secondary hypotheses addressed the effects of age and skin and eye color on the dependent measures. As expected, LED A delivered very different color perceptions on all measures when compared to QH; LED B did not differ from QH. The results show that it is possible for LED sources to match the familiar incandescent sources. However, although it is possible to deliver what appear to be millions of colors with a three-chip (RGB) device, there is the risk of creating a very poor luminous environment. © 2013 National Research Council Canada and Wiley Periodicals, Inc. |
|---|
