The same object can look different under various light sources due to spectrum differences. The accuracy of color reproduction under specific lighting is called color rendering.
Since 1965, the CRI (Color Rendering Index) has been used to evaluate color rendering. It is a dimensionless value that cannot exceed 100 and is measured on the Ra scale, which includes 14 colors. Color samples are illuminated by a reference source D65 (its spectrum is similar to the midday sun in Western Europe) and the light source being measured. Originally, CRI was determined by surveying people who were shown color plates under different lights to calculate the percentage of correct matches. Modern equipment allows CRI to be measured without human input. CRI measurement for LEDs is governed by international standard CIE 177:2007.
Recommended CRI values for various applications (more precise values are set by national lighting standards):
Drawbacks of CRI
Like the Sun, incandescent lamps emit light from heated material. The difference is that the filament temperature in incandescent lamps is lower, shifting its emission towards the red part of the spectrum. Thus, CRI for incandescent lamps is considered 100. However, color differences, such as in clothing under incandescent and natural lighting, are noticeable.
Another drawback is that CRI shows only how accurately a human recognizes colors, ignoring an important factor like "vividness-dullness" of color hues. This factor is vital in many applications, especially in retail.
Standard IES TM-30-20
In 2015, a new color rendering standard, IES TM-30-15, was adopted in the US and is used in many countries. The new edition, IES TM-30-20, was released in 2020.
IES TM-30 evaluates 99 color tones. The method is designed for automated measurement processes. It uses not only color plates but also reference objects shaped like everyday items. Color rendering is characterized by two indicators: fidelity (Rf) and gamut (Rg).
Rf is similar to CRI, also not exceeding 100, with minimal numerical difference between them. "Vividness-dullness" parameters are evaluated by Rg. If the tested light source doesn't alter the color saturation compared to the reference, Rg = 100. If the colors become less saturated, then Rg < 100, and if more saturated, Rg > 100.
For CRI and Rf, the rule is simple - closer to 100 is better. The situation with Rg is more complex. For example, Rg > 100 is good for retail but not for a museum, where Rg = 100 is needed. Overall, the IES TM-30 standard allows for more precise selection of LEDs for specific applications.
