Is CRI metric relevant anymore?
CRI or Color Rendering Index was the first industry standard, developed in the 1930s and updated in the 1970s, for measuring how color appeared under different light sources. The CRI, back then & till the conventional light sources were predominantly used, was an important parameter , but it now faces severe limitations. These limitations are becoming more and more apparent as this standard ages and new lighting technologies pushes the limits of the possibilities.
The CIE established a scale of 8 CIE standard color samples for the CRI calculation method. In rare cases, there are an additional 7 other colors that can be used, but commonly only 8 are used to measure CRI value. The test involves comparing the eight color samples under the test light source and then comparing it to a reference light source, usually the sun. The average differences are then subtracted by 100 to get the CRI value.
Normally, CRI is compared with only 8.
The higher the number, the light source is “said” to be more closer to the reference and hence better in color rendition; but it has the following limitations:
- This method uses only 8 color samples of the many colors which are present naturally and that too pastels which are not saturated
- Light sources can only be compared when they have equal CCT (Color Temperature)
- Being an average value only, certain colors (Reds most Commonly) might not have a good individual value in the distribution but still can have a higher CRI value
IES (Illuminating Engineering Society) has developed the new and improved method called TM-30-15 for evaluating light source rendition. The TM-30-15 method uses a Fidelity Index, Gamut index and Color Vector Graphic to evaluate the light source color rendition. The difference between the Fidelity Index of CRI and TM-30-15 method is that the latter uses 99 color evaluation samples (CES), instead of just 8.
CRI R9 and why it is important
Red is an important color for many applications like photography, retail, textiles, medicine and reproduction of human skin tones. Many objects that do not appear red actually are a combination of colors, including red. Skin tones, for example, are very much influenced by the redness of the blood that flows right beneath our skin. Therefore, a light that lacks red will make a person look pale, or even green.
Spectral Test results of a reputed manufacturer which claim Ra=85 but have low R9 value (below 10). A light source can actually perform quite well with the first 8 test color samples, scoring quite well for R1-R8.
For the general CRI Ra metric, this means that an LED with poor red color rendering can still get away with with an 85-90 CRI (Ra) rating.
A light source with high CRI, say 90 or even above, can have low values of R9 or saturated red. Under such a source certain objects will not reproduce reds properly defeating the whole purpose of a “high CRI” light source.
CRI R9 is one of the test color samples (TCS) used in the calculation of extended CRI. Many manufacturers will only report general CRI, however, which does not include the CRI R9 score. CS9 spectrum is almost entirely composed of red light (wavelengths longer than 600 nm).
CRI R9 is a very important metric because many light sources will be lacking in red content, but this fact will be hidden due to the averaging out of CRI calculations which do not include R9.
Conclusion
A generic CRI value which most of the manufacturers give in their technical data sheets is an incomplete metric. Specifiers and customers should at least demand the CRI R9 values from them which it critical for most of the lighting applications.
The latest TM-30-15 remedies the limitations of the older CRI scale but its adoption remains in early stages till date.