You are here

The frequently asked questions shown below cover color fundamentals

13. What is color rendering?

Knowledge Level: *

Objects don’t have an inherent color, but rather reflect different amounts of energy over the visible spectrum. That means that the spectrum emitted by a light source changes the way objects appear. Color rendering describes this phenomenon. By using color rendering metrics, we can assign ratings to light sources using a set of standardized color samples. Although the objects in a given space won’t exactly match those of the standardized color samples, metrics can give us a good idea of performance and simplify comparisons. Color rendering metrics can characterize a number of different things, including color fidelity (see question 14), color gamut (see question 15), and color preference, among others.

14. What is color fidelity?

Knowledge Level: *

Color fidelity refers to the degree of similarity for a color or colors rendered by a test source (what you’re interested in) and a reference condition, usually a theoretical/modelled illuminant, such as Planckian radiation (see question 22) or a D Series Illuminant (see question 21). Color fidelity only describes the magnitude of the difference, not how the colors are different (i.e., is chroma [see question 17] increased or decreased).

CIE CRI and TM-30 Rf, as well as many proposed measures, characterize average fidelity across many samples. In cases where the reference illuminant is daylight or a heated mass (i.e., light created without electricity, commonly known as blackbody radiation or Planckian radiation), fidelity measures are sometimes thought of as characterizing naturalness; however, research has shown that fidelity is poorly correlated with human perceptions of color quality. Nonetheless, characterizing color fidelity helps ensure that colors are not too distorted compared to what we expect; if distortion is too great, the environment may be uncomfortable or confusing. Understanding color fidelity can also help ensure that two different sources, perhaps in the same space, make objects look very similar, although it does not have any bearing on ensuring that the color of the emitted light appears the same.

15. What is color gamut?

Knowledge Level: *

Generally, gamut refers to an area enclosed by points in a chromaticity (see question 16) diagram. However, gamut can mean different things depending on its specific use. In the printing and display industries, gamut refers to the colors that can be created based on the inks or display primaries; that is, the points represent sources. A similar situation is mixing light from multiple sources. For color rendering, gamut refers to the area enclosed by a set of sample object colors. There’s no such thing as a color that can’t be rendered, just various levels of difference between sources. A gamut area is relatively meaningless on its own, but when compared to another area (for another source, using the same samples), it can indicate relative changes in color appearance, such as average increases in chroma.

16. What is chromaticity?

Knowledge Level: *

Chromaticity is a numerical specification of color, regardless of luminance (brightness). It accounts for hue and chroma (see question 18). Because it is separate from luminance, it is best used for specifying the color of light emitted from a source, but it can also be used as a simplified quantification of object appearance. Two sources with the same chromaticity coordinates should theoretically appear to emit the same light, although the chromaticity calculations are subject to the limitations of models of human visual sensitivity called color matching functions. Two sources with the same chromaticity may have different spectral power distributions, and therefore render object colors (see question 13) differently.

17. What is correlated color temperature (CCT)?

Knowledge Level: *

Correlated color temperature (CCT) is calculated as the color temperature of the Planckian radiator nearest the chromaticity (see question 16) coordinates of the test source. It’s calculated in the CIE 1960 (uv) color space (see question 19). Although far from a perfect metric, it indicates whether the apparent color is warm or cool or something in between. For interior architectural lighting applications, sources typically range between 2700 K and 6500 K.

18. What are the dimensions of color?

Knowledge Level: *

Three dimensions are needed to describe human perception of color, although the exact terms can vary based on the color description model. For example, colors may be classified using hue, saturation, value (HSV) or hue, chroma, lightness (HCL), among others. TM-30 uses the CAM02-UCS (see question 20) color space (see question 19), which can be used to determine the hue, chroma, and lightness of a sample, among other attributes.

Hue is a more precise term for what one might call “color”; it’s the differentiation among red, yellow, green, blue, and the shades in between. Lightness (which is closely related to value or brightness), is the brightness of an area judged relative to a similarly illuminated area that appears to be white. Finally, chroma (which is closely related to saturation and colorfulness) is defined as the colorfulness of an area relative to a similarly illuminated area that appears white—it can be thought of as the intensity of the color. Importantly, some hues can reach higher chroma values than others.

19. What is a color space?

Knowledge Level: **

Color spaces are numerical representations of color, which are derived from color matching functions (CMFs). CMFs are experimentally-derived models of human color vision sensitivity. Color space representations may be either two dimensional (for light sources) or three dimensional (for objects). Two-dimensional chromaticity diagrams, such as CIE 1931 (xy) or CIE 1976 (u', v'), can be used for specifying the color appearance of light sources to ensure the color of the light matches; however, they do not consider lightness, so they are unsuitable for characterizing the difference in objects’ color appearance. Three dimensional object color spaces, such as CIE 1964 U*V*W*, CIE LAB, or CAM02-UCS, are suitable for object color specification.

Many color spaces have been developed over the years, with the goal being to improve the uniformity of the space. With a uniform color space, equal numerical distances correspond to equal perceptual differences throughout the space. That is, an equal measured distance in the color space corresponds to the same perceived difference regardless of hue, chroma, or lightness. CAM02-UCS (see question 20), which is used in TM-30, has been shown to be the most uniform color space to date.

20. What is CAM02-UCS?

Knowledge Level: **

CAM02-UCS is a color space (see question 19) based on the CIECAM-02 color appearance model. It is the most recent and most uniform color space [1]. This ensures that perceived color differences are accurately characterized. It has coordinates J', a', b', where (a', b') characterizes chromaticity and J' specifies lightness. Hue and chroma (see question 18) are related to the chromaticity (see question 16) coordinates.

  1. Luo MR, Cui G, Li C. 2006. Uniform colour spaces based on CIECAM02 color appearance model. Color Research and Application 31(4): 320–330.
21. What are the D Series illuminants?

Knowledge Level: **

The CIE D Series illuminants correspond to daylight at different color temperatures. A formula is used to calculate the spectral power distribution (SPD), given the color temperature. The formula approximates real measurements of daylight. CIE D Series illuminants serve as the reference illuminants in TM-30 when the correlated color temperature (CCT) of the test source is 5500 K or greater.

22. What is a Planckian radiator (blackbody radiation)?

Knowledge Level: **

A Planckian radiator, also called a blackbody radiator, is a calculated illuminant based on the temperature of a black body. A black body is an ideal emitter; that is, it does not reflect any light, and the emission is purely based on the temperature of the mass. The formula used in the calculation is known as Planck’s law. Incandescent lamp filaments are a close approximation of a Planckian radiator. Planckian radiation serves as the reference illuminant in TM-30 when the correlated color temperature (CCT) of the test source is 4500 K or less.

23. How does light level affect color rendering?

Knowledge Level: **

Light level has a substantial effect on the appearance of colors, with apparent chroma increasing as luminance increases. This is known as the Hunt effect. TM-30 relies upon the CIECAM-02 color appearance model, which accounts for the Hunt effect. However, the light level parameters input into the calculation are fixed; this ensures that all reported values are comparable from product to product, and it is necessary because color rendering measures characterize a source, not an installation (each situation of an object illuminated by a light source can lead to a difference object luminance, which can affect color perception). The bottom line is that a specifier may want to consider the relationship between light level and color rendering for some applications, especially when considering tradeoffs between fidelity and gamut.

24. What is a reference-based color rendering measure?

Knowledge Level: *

Almost all existing color rendering measures, including the CIE’s CRI and TM-30, rely on comparing how colors are rendered by a source in question compared to a reference condition. Usually, the reference is at the same correlated color temperature (CCT) as the test source. This is called a relative reference, and allows color rendering to be considered independent of CCT.

With a measure like CRI, where color fidelity (see question 14) is the only attribute characterized, the reference source is particularly important because it represents an ideal condition. The various measures included in TM-30 allow for greater understanding of the attributes of the test source, which makes the reference a point of comparison, but not necessarily an ideal source.

25. What is the difference between referenced-based measures of color rendering and fixed-reference metrics?

Knowledge Level: **

There are several ways to characterize light source color rendition. The familiar CRI uses a reference-based methodology, where the light source of interest is compared to a reference condition at the same CCT. Other options are to compare a light source to just one reference, regardless of CCT, or to eliminate references altogether.

The developers of TM-30 chose to use the familiar reference-based methodology, with references that change with the CCT of the test source. This approach allows color temperature to be considered independently of color rendering, which allows the tone of the environment to be chosen. For example, a 5000 K source may be too cool in appearance for the atmosphere of a home furnishings store, so it may not be appropriate to compare a 2700 K source against a 5000 K reference source. Fixed-reference measures often result in inherent differences in ratings for products of different CCTs. Whether or not these effects are perceived by humans, they can create confusion because different CCTs have different possible score ranges. One of the most important aspects of the TM-30 method is that all the measures use the same reference sources. This allows for direct evaluation of tradeoffs between the measures.

Importantly, the reference sources are not necessarily ideal sources for any application; they only serve as an anchor point for consistent comparisons between products. Unlike fidelity-only measures, TM-30 characterizes how a light source differs from a reference, which enables users to select products most suited for an application, rather than simply using a “higher number is better” approach.