Actually, eyes are the best tools to evaluate light quality. But when we need to test it quantitatively, what should we do? CIE, NIST and other divisions have defined CRI, CQS, TLCI and TM-30-15, which one is better?
Color Rendering Index (CRI)
- Color Rendering Index (CRI) history, definition and calculation.
Almost all of the people who care about light quality are aware and pay attention to the Color Rendering Index (CRI), which is developed in the middle of 20th century for then-new fluorescent lamps. As its name, CRI measures the capability of an illuminant for rendering the original color of a physical object. Generally we use the score of Ra to evaluate the CRI level, the range of Ra is 0-100, the higher the better.
In the calculation of the CRI, the color appearance of 14 reflective samples is calculated when illuminated by a reference illuminant and the test light source. The simulated color of the 14 samples, when illuminated by a CIE Daylight illuminant of 6500K (D65), is shown as below:
Please notice that, even though there are 14 color samples, Ra only picks 8 in the top row. For the CRI calculations, the reference illuminant is a Planckian radiator (if below 5000K) or a CIE Daylight illuminant (if at or above 5000K), matched to the correlated color temperature (CCT) of the test source. The calculative process is complicated, let’s ignore it for the moment and just check the schematic below:
The General Color Rendering Index (Ra) is the average of Ri for the first eight samples.
- The limitation, irrationality of CRI
1. The 8 color samples are low saturated.
Let’s see the 8 tested colors again:
None of the eight reflective samples used in the computation of Ra are highly saturated. This hidden danger would lead to the phenomenon that a high Ra light source has a poor quality. RGB white LED is an example.
For the rest 6 color samples, the 9th one, which is much more important for rendering skin tones, stands for saturated red. This value is ignored by CRI and sometimes by people.
For some poor LED, R9 could be even minus, for most of nominal high CRI LED, R9 is still undesirable. For high end and special lighting application like museum lighting, photography lighting and motion film lighting, R9 is particularly
important. Another color is the 12th saturated blue, due to the theory of blue chip LED emitted, it is very difficult to reach high value.
2. Calculation is not rational enough.
Ra is simply the average Ri.
We can imagine that if one or several Ri are low while the rest are high, the Ra still seems good, under this situation you cannot expect a uniform and good final quality.
3. Not suitable for evaluating LED light source.
CRI is defined almost 40 years ago, it is used for judging continuous emission spectrum and at that time we do not have common LED illuminant like today. LED is an amazing invent, different matches of chip and phosphor create many possibilities, but also the opportunity for cheating. Manufactures can pick up the wavelengths which help the CRI, it is not a difficult thing.
- Distinguish good CRI and poor CRI
As we talked above, from Ra we can hardly tell the real light quality, then what can we do? Analyzing spectrum is always the most convenient way, every detail cannot be hidden in the spectrum data, below are Yuji VTC series 2700K and 6500K typical SPDs (Spectral Power Distribution), professionals can identify the features immediately.
Due to the irrationality of CRI to evaluate LED light, NIST proposes another parameter named CQS (Color Quality Scale), what improvements does CQS make and what about the effect? Let’s talk about it later.
“The Color Quality Scale”
Wendy Davis and Yoshi Ohno, National Institute of Standards and Technology
Gaithersburg, MD, 20899, USA