Optipedia • SPIE Press books opened for your reference.
Excerpt from Color Vision and Colorimetry: Theory and Applications, Second Edition
Around 1900, many years before the first CIE convention, Albert H. Munsell, an artist, empirically prepared a set of color charts with an almost uniform color representation. In the Munsell color space, the colors are represented in a cylinder with zero-saturation colors (black, gray, and white) along the axis. The coordinate's configuration in this space is illustrated in Fig. 5.4. The lowest extreme of the cylinder axis corresponds to black, while the highest extreme corresponds to white. The position along this axis, with ten steps from zero to nine, is called the value, representing the perceived lightness, which is nonlinear with the luminance Y. To take into account this nonlinearity, the value was originally taken as the square root of the luminance (Fig. 5.5), although it was later redefined to improve it.
The chroma represents the saturation of the color. It increases in a perpendicular direction to the axis, toward the edge of the cylinder, with values from zero to eight, and with monochromatic and purple colors around the periphery.
The hue of the color is represented by the angle. The Munsell circle is divided into the following 10 angular sectors with an angle of 36 deg: yellow (Y), yellow-red (YR), red (R), red-purple (RP), purple (P), purple-blue (PB), blue (B), blue-green (BG), green (G), and green-yellow (GY). Each of these sectors is divided into 10 sub-sectors with an angle of 3.6 deg. Figure 5.6 shows a Munsell circle inside the CIE diagram. In this system, the different colors are specified by Hue Value/Chroma, as in the following example: 4 YR 7/3, which means a yellow-red color in subsection 4 (hue = 4 YR) with lightness (value) = 7 and chroma = 3. Figure 5.7 shows a circle of colors with the same Munsell value.
In this system there are many planes with colors—one for each hue. A three-dimensional representation of all colors in the Munsell color space is formed by a series of radial planes with constant hue, as illustrated in Fig. 5.8. Thus, on each radial plane all colors have the same hue, but with the chroma increasing outward from the central axis, and the value increasing with height. It is interesting to note that not all planes have the same shape. They bulge outward for blue and purple colors at low values, and for yellow colors at high values. Only black is considered for the lowest value, while only white is represented for the highest value. We will see later in this chapter that these characteristics are also present in other uniform color spaces. Figure 5.9 is a close representation of the colors for four Munsell planes with different hues.
The original Munsell System was later modified to correct some obvious errors in the location of some colors. The new color designations and coordinates are known as Munsell Renotations, and the system is called the Munsell Renotation System.
The Munsell color system is almost perfectly uniform, where each color is separated from its closest neighbor by equal perceptual distances (Fig. 5.10) in comparison to the location of the same colors in the CIE x-y diagram. We can see that these curves closely resemble the constant chroma curves in Fig. 5.3.
An important practical disadvantage with the Munsell color system is that colors are defined only for a two-degrees observer and a C illuminant. Another problem with this system is that no analytical expressions to convert the CIE system to the Munsell system, or vice versa, exist. However, look-up table programs to perform these conversions have been proposed (Newhall, 1943, and Rheinboldt and Menard, 1960). Many attempts had been made to analytically define a color system that closely resembles the Munsell System, but without success. Some of these systems will be described in the following sections.