Tutorial: Color Rendering and Its Applications in Lighting

ABSTRACT

This tutorial explains how the human perception of color rendering arises, in terms of the underlying phenomena of light and vision, and using those concepts it presents a clear explanation of the CIE Color Rendering Index. The strengths and weaknesses of the CIE Color Rendering Index are reviewed and some common misunderstandings about color rendering are addressed. It is suitable for self-study, with learning outcomes stated at the beginning and a conceptual summary provided at the end.

KEYWORDS:

• color rendering
• light source
• spectral power distribution

Notes

1. Definitions of terms that appear in italics are provided in Appendix A.

2. It is worth clarifying what can be confusing abbreviations. The CIE method for color rendering is known as the Color Rendering Index, abbreviated as CRI. The CIE method employs 14 test color samples. Each test color has its own CIE Special Color Rendering Index (note the use of the word “Special”), abbreviated as R_i, where i is an integer ranging from 1 to 14. The first eight values of R_i are averaged to compute the CIE General Color Rendering Index (note the word “General”), which is abbreviated as R_a. One way to remember this is that the subscript i denotes a selected integer and a denotes average. Frequently, the abbreviations R_a and CRI are used interchangeably, and context is usually enough to infer intent, but to be strict they are not the same. Any value of R_i is also a CRI. Don’t despair if this is confusing now—read it again and at the end of the tutorial it will make sense.

3. In quantitative work—for example, when using a spectral illuminance meter for measurement—the intensity is often first recorded in the radiometric measure of power per unit area per unit wavelength, which technically is W m⁻³ and more often is expressed as W m⁻² nm⁻¹ because wavelengths are usually expressed in nanometers. In other cases, the intensity may be first measured by a spectral irradiance meter and recorded in the radiometric measure of power per unit area per unit solid angle, which technically is W sr⁻¹ m⁻³ or W sr⁻¹ m⁻² nm⁻¹. These details will not matter to most users, especially because the SPD is most often expressed in relative terms by dividing by the peak radiometric value, which eliminates the need for units in quantifying the SPD.

4. In some cases the radiometric measure radiance is used (in units of W m⁻² sr⁻¹), in which case the output of the calculation is in units of luminance, lm m⁻² sr⁻¹ (also known as candela per m², with the symbol cd m⁻²).