The Standard Observer

Since color matching is meant for humans, it is natural to define color in terms of an average, or Standard Observer . Our first step is to build an instrument which contains three colored lamp sources, a place for the observer, intensity detectors, and a monochromator. One design is shown in 6.7.15., given on the next page. 

There are two (2) sources of light to be compared. One is from a set of three lamps whose emission is modified by means of suitable filters to give a red beam, a green beam and a blue beam. These are mixed at the screen to form a single spot (although we have not illustrated it in that way, so as to be more discernible). The other source comes from a monochromator so as to obtain a monochromatic beam of light. 

There are controls to adjust the individual beams of red, green and blue light, as well as that of the monochromatic beam. In this way, the mixed beams of light can be directly compared to the monochromatic spot. In the back of the screen are detectors to measure the energy intensity of the beams of light being compared. 

We need about 5000 observers to obtain a satisfactory average, both for the dark-adapted and the light-adapted hiunan eye. Note that we can compare any color In terms of red -i- green -i- blue to a monochromatic color. 

There are three (3) things that we need to accomplish  

---

The GIE Standard Observer. The CIE standard observer is a set of curves giving the tristimulus responses of an imaginary observer representing an average population for three primary colors arbitrarily chosen for convenience. The 1931 CIE standard observer was deterrnined for 2° foveal vision, while the later 1964 CIE supplementary standard observer appHes to a 10° vision a subscript 10 is usually used for the latter. The curves for both are given in Eigure 7 and the differences between the two observers can be seen in Table 2. The standard observers were defined in such a way that of the three primary responses x(X),jy(X), and X), the value ofjy(X) corresponds to the spectral photopic luminous efficiency, ie, to the perceived overall lightness of an object. 

CIE used the 1931 CIE standard observer to estabUsh a color representation system in which the hue and saturation could be represented on a two-dimensional diagram. Three tristimulus values X, Y, and Z are first obtained, based on the standard observer, so that the hue and saturation of two... 

Colorimeters. Also known as tristimulus colorimeters, these are instniments that do not measure spectral data but typically use four broad-band filters to approximate the jy, and the two peaks of the x color-matching functions of the standard observer curves of Figure 7. They may have lower accuracy and be less expensive, but they can serve adequately for most industrial color control functions. Examples of colorimeters are the BYK-Gardner Co. XL-835 the Hunter Lab D25 series the Minolta CA, CL, CS, CT, and CR series (the last of these is portable with an interface) and the portable X-Rite 918. 

Once this fact was rccdized, it was understood that an average of what each person saw would have to be made if a standard system weis to be formed and promulgated. This led to the concept of the "Standard Observer". Thus, the research required to define and measure color took a completely different path from the original methods such as the Munsell Color Tree. 

In this case, the relative response of the observers are summed into a response called "THE STANDARD OBSERVER" and is normalized for easier usage. Now, let us examine the effects of colors (chroma) as perceived by the human eye. Keep in mind that each person perceives "color" somewhat differently from other persons. 

The difficulty in setting up the initial system for color comparisons cannot be underestimated. The problem was enormous. Questions as to the suitability of various lamp sources, the nature of the filters to be used, and the exact nature of the primary colors to be defined occupied many years before the first attempts to specify color in terms of the standard observer were started. As we said previously, the Sun is a black-body radiator having a spectral temperature of about 10,000 °K (as viewed directly from space). Scattering and reflection... 

The result is finalized response curves for the Standard Observer, also called "Tristimulus Response curves", and is shown as follows ... 

However, these analogues are actually hypothetical. The reason for this is that it is nearly impossible to obtain optical measurement components, such as the source and the detector, whose response to light across the visible spectrum is flat (or nearly so). However, this is not an impossible task and we find that an excellent match can be obtained to the transmission functions of 7.8.21., i.e.-those of the Standard Observer. This is typical for commercially available instruments. Now, we have an instrument, called a Colorimeter, capable of measuring reflective color. 

The CIELAB system (1976) strictly standardizes the light source and the observer. CIE recommends three standard sources, A is an incandescent lamp, and B and C are lamps provided with different two-cell Davis-Gibson liquid hlters that simulate noon daylight and average daylight, respectively. Since the main object of the system is to obtain colorimetric results for normal tri-chromats (people with normal color vision), the standard observer must represent the human population with normal... 

The development of the observer response functions is the foundation for color measured by an instrument. The Standard Observer established a recognized method for converting... 

The normalization projects all colors onto the plane at X + Y + Z = 1. Therefore, the third coordinate is redundant and we have to only specify the coordinates x and y. The third coordinate is then given by = 1 — x — y. Since the weights as defined by the standard observer are all positive, the coordinates x, y. and will also be positive. Owing to the constraint x + y + = 1, they will all be smaller than one. Instead of specifying a color by the triplet X, Y, and Z, we can also specify the color by the chromaticity coordinates x and y as well as the absolute luminance Y. The absolute coordinates X and Z are then given as (Poynton 2003)... 

The standard observation that thermal and photochemical electrocyclic and cycloaddition reactions always take place by opposite stereochemical pathways [96] is explained directly by the unit of angular momentum carried by a photon. Photochemical activation disturbs the balance between angular-momentum vectors and dictates a different molecular conformation. 

The distribution coefficients (x, y, and 2) have been calculated for the standard observer. They arc the amounts of the three imaginary reference colours. A, Y, and Z, weighted to conform with the equi-energy spectrum... 

In a tristimulus colorimeter, three or four filters duplicate the response of the standard observer. The filters, which correspond to the three primary colours in the spectrum (red, green and blue), can be combined to match most colours. The more sophisticated instruments carry a fourth filter to simulate the blue part of the CIE [[X bar]] function shown in Fig. 5.1. The other essential parts of a tristimulus colorimeter are a white light source, an array of photometers and, nowadays, a computer or an interface to one, as shown in Fig. 5.2a. The computer can collect responses as well as carry out data transformations between CIE and other colour scale systems or between different standard white light sources or white diffusers. Hunter and Harold (1987) give a good summary of the transformation formulae employed. Data from a colorimeter is given as a three-point output, commonly CIELAB, HunterLab or 7, x, y. 

Once we have done this, we now have our three primary colors in the form of standard lamps, and can proceed to determine Items 1,2 3, given in 6.7.16. To do this, we vary the wavelength of the monochromatic light, and determine relative amounts of red, green and blue light required to match the monochromatic color. This is done, as stated before, for about 5000 observers. The result is finalized response curves for the Standard Observer, also called "Tristimulus Response curves". 

We finally arrive at the result we want, since we can now set up Tristimulus Filters" to use in defining colors. We can now define "y as our standard luminosity curve for the human eye (photopic vision). Note that X, the red tristimulus value, has a certain amount of blue in it in order to duplicate the response of the red preceptor in the retina. Note that these colors are the result of the "Standard Observer" measurements that we started in the first place. 

Let us now summarize the results we have achieved. We have measured the luminosity response of the human eye, in terms of photopic and scotopic behavior. We also defined a "black-body" and its wavelength emission, stipulating its absolute temperature. We then defined Standard Sources. We next designed a Color Comparator and then determined the transmission characteristics of three (3) filters required to duplicate the response of the three color preceptors of the human eye. These we called the tristimulus response of the Standard Observer. 

Finally, we obtained Chromaticity Coordinates which we could use to plot various hue values. If we have a Color Comparator, then we can measure any hue and compare it to any other. But, if we do not, we either build or buy one, or resort to an instrumental method. The Color Comparator uses a human observer to specify color. If we wish to use an instrumental method, then we must correct each component of the instrument to the chroma response of the Standard Observer. While the Color Comparator given above in 6.7.15. was perfectly satisfactory for setting up a system of chromaticity coordinates, it was difficult and awkward to use. What was really required was an instriunental method of color measurement. 

You might question why we need a light source to measure an emitter. The light source acts as a standard for comparison, using the reflectcmce from a standardized reflectance material so as to be able to use the STANDARD OBSERVER response we have already developed for color matching. 

Fig. 3). For this reason, always radiometric units must be used in photosynthesis instead of photometric units. The radiometric analogue of the lumen (luminous flux) is the watt (radiant flux). At each wavelength both are coupled according to the standard observer curve at 555 nm 1 W of radiant flux corresponds to a luminous flux of 683 lumens by definition. As a comparison, at 650 nm, 1 W corresponds to 73 lumens only. 

The tristimulus colorimeter is the simplest means for instrumental color specification. A light source, filters, and a photodetector are combined such that together they yield a direct evaluation of the tristimulus values X, Y, and Z. A spectrophotometer may also be used for color specification. That is, a spectral reflectance or transmittance curve (R) can also be converted into X, Y, Z data. The curve R versus wavelength 2) is integrated over the visible range, with the spectral emission of an illuminant (E), and the standard observer functions (x,y,z). The areas under the resulting curves yield the values of X, Y, and Z that is ... 

Luminous lu-m3-n9s [ME, fr. L luninosus, fr. lumin-, lumen] (15c) adj. (1) Adjective used to imply dependence on the spectral response characteristic of the Standard Observer defined in the CIE System. Thus, the luminous reflectance or the luminous transmittance is described by the Y-tristimulus value in the CIE System. The adjective is applied to many measures of light, such as... 

Tristimulus computation data n. Products of relative spectral-energy distribution of an illuminant multiplied by each of the three color matching functions in the CIE system designated as EcY, EcT e.g., for Illuminant C and the color mixture data for the standard observer at designated wavelengths. 

These difficulties usually involve the assignment of different values for the color difference between two samples when judged visually and by the instrument in question. This is best illustrated in the near-white region, where even the basic C.I.E. system gives incorrect values owing to incomplete evaluation of the 1931 standard observer. This was first noted by Jacobsen (1948), though Wald (1945) had already indicated that the eye was more sensitive to the blue than is shown by the standard observer. These findings were corroborated by Judd (1949), who modified the characteristics of the standard observer to account for differences observed visually in titanium paints but not indicated by reference to the C.I.E. 

---