Color instruments and software packages all contain references to CIE standard illuminants and observers. What do these terms mean and how are they used in practical color evaluations? The CIE (from the French Commission Internationale de L’Eclairage), or International Commission on Illumination, is the worldwide authority on how we see and measure color. BYK-Gardner uses many of these standards in its instruments and available software packages. This paper will give some easy to understand explanations of these standards and how they should be used.
The first and easiest standard to explain and understand is the Standard Observer. In 1931 there was evidence that most of the cones located in the human eye were located in a fairly small area at the back of the eye, defined by a 2 degree angle of incidence on the fovea. A small sampling of human observers set about matching a set of colors, viewed through a pinhole type device, using a set of red, green and blue lights. The angle of viewing was set at the 2 degree angle that would activate the known color receptors in the eye. Hence the 2 degree standard observer was born.
However, in 1964, this standard observer was modified to include a 10 degree angle of foveal stimulation. Cones were found to cover a much larger area at the back of the eye than previously believed. This 10 degree viewing angle is now thought to provide the best average spectral response in human observers. However, there are certain industries and conditions that still specify the 2 degree observer. Generally color evaluation where the samples being tested are normally at some great distance away (road signs, for example) or are the size of a dime at arm’s length (printing applications) are best evaluated using the 2 degree observer.
So that explains the 2 and 10 degree angles of illumination. But what about the letter codes found before the “/” sign in the illuminant and observer combinations? These all refer to the mathematically defined spectral power distributions of a number of artificial illuminants, or light sources.
The first of these, as defined by the CIE, were the A, B and C illuminants defined in 1931. “A” illuminant was defined to represent average incandescent light sources, or tungsten filament bulbs. “B” and “C” illuminants were then defined, using liquid filters, to represent direct sunlight at noon with a color temperature of 4874K (B) and average day light at 6774K (C). B is not used anymore and is rarely even referred to in specifications. C illuminant is still referred to quite often, even though it is not an official CIE illuminant. It was loosely defined by the CIE, but was never given the Standard Illuminant status.
In the years following these originally defined illuminants a number of other illuminants have been defined. Most are theoretical, mathematically described illuminants that cannot be found anywhere in real life. But they are useful in that they provide standards for comparison and relate well to light in our environment and how it changes the appearance of physical samples.
Following the A, B and C set of illuminants is the set of Daylight illuminants. None of these illuminants is found naturally, but the spectral power distribution used to characterize these various illuminants is derived from three vectors. The first is a mean of all SPD factors used to reconstitute an SPD with only one fixed vector. The second vector corresponds to the yellow-blue shift brought about by changes in the correlated color temperature caused by the presence or absence of direct sunlight or clouds. The third vector corresponds to the pink-green variation caused by the presence of water in the form of vapor or haze.
These D illuminants are named according to the color temperature that they are meant to emulate. For example, D50 is used as an emulator for Horizon Light with a color temperature of 5000° Kelvin. D55 is used for Mid-morning light with a color temperature of 5500° Kelvin. D65 is used for Noon Daylight with a temperature of 6500° Kelvin. And finally D75 corresponds to North sky Daylight at 7500° Kelvin.
All four of these D illuminants have their own industry proponents and are used in the definitions of standards. For example, the paint industry will often specify D65/10° for the illuminant/observer. The printing industry will usually specify D55/2°. It is very important to know what the standard illuminant and observer angle are when comparing instrument measurements and deltas. Using the wrong combination can and will lead to misleading / incorrect measurements and approval of out-of-spec materials.
Nowhere are these difference more pronounced and supplier specific than in the fluorescent light industry.
There are at least 8 different fluorescent illuminants, meant to match the spectral output of various manufacturer’s bulbs. However, manufacturers match each other’s bulbs and even though the color temperature may be equivalent, the set of phosphors used to create that temperature will likely be different between manufacturers and result in very different spectral peaks and valleys when evaluated with a spectral radiometer. The various fluorescent types are listed here:
F4 is called a Warm White Fluorescent (WWF) and is used to closely match the warmth of an incandescent bulb in the home environment.
F6 is a Fluorescent bulb with a temperature of 4150° Kelvin. The F6 tube is often referred to as a Cool White Fluorescent bulb.
F7 bulbs are Daylight Fluorescent (DLF) with a color temperature of 6500° Kelvin.
F8 bulbs have a color temperature of 5000° K, much like the D50 illuminant described above.
F10 bulbs also have a color temperature of 5000°K. However, they differ enough from the F8 illuminant to warrant their own spectral signature.
F11 bulbs are also called T84 or TL84. These are energy conserving fluorescents with very narrow bands of spectral illumination.
So the big question after reading about all these illuminants is “what does it all mean to me, and why is it so important that I know what illuminant(s) to use?” L*a*b* color values are derived from a combination of both the spectral curve of the sample and the illuminant spectral values. A sample can look completely different under two different light sources. However, a single sample will always look the same to the human eye – the eye has super-computer intelligence when it comes to compensating for various external light sources. However, when evaluating two different samples which may have slightly different pigmentation, the illuminant will play a huge role in how the two samples appear in relation to each other. Consider an orange sample evaluated under daylight at around 5000° K. A second sample with slightly different pigmentation may look identical under true daylight at around 5000° K. However, under an F10 fluorescent bulb, which is also a 5000° K illuminant, much of the light that would reflect off of the sample in the 550-600 nm range and over 625 nm is missing. The L*a*b* values will be very different and metamerism is highly likely.
The samples may look identical under true daylight. They may look identical under cool white fluorescent bulbs. They may look identical under incandescent bulbs. But then you walk into an area with energy saving F10 bulbs and they look like completely different colors. The mathematical values used in the firmware of the spectro-guides and in most color evaluation software, such as BYK-Gardner’s auto-QC, help to determine if the colors will be a match under the various lighting conditions to which the sample may be subjected.
Know the illuminant requirement and use it. It’s the only way to guarantee good color matches that don’t vary when you change the surrounding light.