Color-rendering index

From Wikipedia, the free encyclopedia

The color rendering index (CRI) (sometimes called Color Rendition Index), is a measure of the ability of a light source to reproduce the colors of various objects being lit by the source. It is a method devised by the International Commission on Illumination (CIE). The best possible rendition of colors is specified by a CRI of one hundred, while the very poorest rendition is specified by a CRI of zero. For a source like a low-pressure sodium vapor lamp, which is monochromatic, the CRI is nearly zero, but for a source like an incandescent light bulb, which emits essentially black body radiation, it is nearly one hundred. The CRI is measured by comparing the color rendering of the test source to that of a "perfect" source which is generally a black body radiator, except for sources with color temperatures above 5000K, in which case a simulated daylight (e.g. D65) is used. For example, a standard "cool white" fluorescent lamp will have a CRI near 62. Newer "triphosphor" fluorescents often claim a CRI of 80 to 85.

CRI is a quantitatively measurable index, not a subjective one. A reference source, such as black body radiation, is defined as having a CRI of 100 (this is why incandescent lamps have that rating, as they are, in effect, blackbody radiators), and the test source with the same color temperature is compared against this. Both sources are used to illuminate eight standard samples. The perceived colors under the reference and test illumination (measured in the CIE 1931 color space) are compared using a standard formula, and averaged over the number of samples taken (usually eight) to get the final CRI. Because eight samples are usually used, manufacturers use the prefix "octo-" on their high-CRI lamps.

The standard formula consists of measuring the color indices of eight sample colors on the 1964 W*U*V* uniform color space (which is now obsolete). The indices of the samples are first measured while being illuminated by the reference source, yielding indices $[W i 0, U i 0, V i 0]$ where the index i specifies the particular sample color. The indices of the samples are then measured under the test source yielding indices $[W i, U i, V i]$ . The distances $Δ E i$ between the measured colors is then calculated:

$\Delta E_i=\sqrt{(U_i-U_{i0})^2+(V_i-V_{i0})^2+(W_i-W_{i0})^2}\,$

The color rendering index $R i$ is calculated for each of the eight samples:

$R_i=100-4.6\Delta E_i\,$

which gives the color rendering index with respect to each sample. The general color rendering index $R a$ is then the average of these eight separate indices.

$R_a=\frac{1}{8}\sum_{i=1}^8 R_i$

Although an objective measure, the CRI has come under a fair bit of criticism in recent years as it does not always correlate well with the subjective color-rendering quality for real scenes, particularly for modern (e.g. fluorescent) lightsources with spikey emission spectra, or white LEDs. It is understood that the CIE is looking at developing newer color-rendering performance metrics.