A new measure of colour discrimination for LEDs and other light sources

We show conclusively that, contrary to common belief, larger gamut area is not associated with better colour discrimination. We propose a new measure of colour discrimination, R d , which quantifies the number of cap transpositions in the Farnsworth-Munsell 100 Hue Test (FM-100) that are introduced by a source spectrum. R d was developed from 480 FM-100 tests (24 light sources × 20 participants per source), where the light source spectra systematically varied in IES R f (from 65 to 95), R g (from 80 to 120), and gamut shape. R d was back tested on more than 200 common light sources to verify accordance with experience and anecdote and is a strong predictor of colour discrimination. We offer a direct way to interpret the R d score. Light sources with superior colour discrimination have an R d score of zero, meaning they introduce no transpositions, thus preserving FM-100 cap order. Light sources with average colour discrimination have an R d score between 4 and 12; sources with poor colour discrimination have an R d score of 16 or greater.