Guide to Mechanical / electrical equipment for buildings: PART III: ILLUMINATION: Lighting Fundamentals

Architecture is the masterly, correct and magnificent play of masses brought together in light. Our eyes are made to see forms in light; light and shade reveal these forms; cubes, cones, spheres, cylinders, or pyramids are the great primary forms which light reveals to advantage.

- LE CORBUSIER

1. INTRODUCTORY REMARKS

FOR MANY YEARS, AN UNWARRANTED division existed in the field of lighting design, dividing it into two disciplines: architectural lighting and utilitarian lighting. The former found expression in design that took little cognizance of visual task needs and displayed an inordinate penchant for incandescent wall washers, architectural lighting elements, and form-giving shadows. The latter saw all spaces in terms of illuminance levels and cavity ratios, and per formed its design function with footcandles (lux) and dollars as the ruling considerations. That both of these trends have generally been eliminated is due largely to the efforts of thoughtful architects, engineers, and lighting designers, assisted in part by the energy consciousness that followed the 1973 oil embargo. The last event spurred research into satisfying vision needs within a framework of minimal energy use. That research, and its resulting energy codes and continuing development of higher efficiency sources, are today motivated by environmental considerations.

Another positive factor in the rationalization of lighting design has been the work of the Illuminating Engineering Society of North America (IESNA). Its activities in research, standardization, and publication have done much to place lighting design on a stable scientific basis while taking full cognizance of its essential artistic aspects. It is precisely this combination of science and art that makes lighting design an architectural-type discipline. For each project, a responsible lighting designer will consider quantitatively:

1. Daylight-its introduction and integration with electric light

2. The interrelationship between the energy aspects of electric and daylighting, heating, and cooling

3. The effect of lighting on interior space arrangement, and vice versa

4. The characteristics, means of generation, and utilization techniques of electric lighting

5. Visual needs of specific occupants and of specific tasks

6. The effects of brightness patterns on visual acuity and qualitatively:

7. The location, interrelationship, and psycho logical effects of light and shadow-that is, brightness patterns

8. The use of color, both of light and of surfaces, and the effect of the illuminant source on object color and sometimes the reverse

9. The artistic effects possible with patterns of light and shadow, including the changes inherent in daylighting, and so on

10. Physiological and psychological effects of the lighting design, particularly in spaces occupied for extended periods

The list is almost endless, because so much of the information we receive from our senses comes via our eyes, and what we see is a direct consequence of scene lighting.

As a result of the need to consider these and other interrelated factors, many of which are mutually incompatible, the lighting designer is faced with many difficult decisions. The purpose of Sections 11 to 16 is then twofold: to provide the background that will help lighting designers make these decisions correctly and to make them proficient in the use of lighting as a design tool.

PHYSICS OF LIGHT

2. LIGHT AS RADIANT ENERGY

The IESNA defines light as visually evaluated radiant energy or, more simply, as a form of energy that permits us to see. If light is considered as a wave, similar to a radio wave or an alternating-current wave, it has a frequency and a wavelength. FIG. 1 shows the position of light in the wave spectrum and its relation to other wave phenomena of various frequencies.

FIG. 1 Electromagnetic spectrum.

From FIG. 1 we see that even the longest wavelength light (red) has a much higher frequency than radio waves and radar, and that light constitutes only a very small part of the wave energy spectrum. Color is determined by wavelength: starting with the longest wavelengths (red), on through the spectrum of orange, yellow, green, blue, indigo, and violet to the shortest visible wavelengths (highest frequency). Bordering the visible spectrum are infra red at the low-frequency (long-wavelength) end and ultraviolet at the high-frequency end. Both are invisible to human beings but not to some animals.

When a light source produces energy over the entire visible spectrum in approximately equal quantities, the combination appears white, whereas a source producing energy over only a small section of the spectrum produces its characteristic colored light. Examples are the blue-green clear mercury lamp and the yellow sodium lamp.

For our purposes, all light will be considered white unless specifically noted otherwise. This position is scientifically tenable because light sources with large differences in chromatic content all appear white after a short accommodation period, and all standard commercial sources permit colors to be easily and correctly identified. It is only when sources differing widely in chromaticity are viewed side by side that the variation in whiteness can be noticed by the effect on colored objects and on neutral surfaces.

3. TRANSMITTANCE AND REFLECTANCE

Lighting design is possible because light is predictable; that is, it obeys certain laws and exhibits certain fixed characteristics. Although some of these are so well known as to appear self-evident, a review is in order.

The luminous transmittance of a material such as a luminaire lens or diffuser is a measure of its capability to transmit incident light. By definition, this quantity, known variously as transmittance, transmission factor, and coefficient of transmission, is the ratio of the total transmitted light to the total incident light. In the case of incident light containing several spectral components passing through a material that displays selective absorption, this factor becomes an average of the individual transmittances for the various components and must be used cautiously. A piece of frosted glass and a piece of red glass may both have a 70% transmission factor, but obviously they affect the incident light differently. In general, then, transmission factors should be used only when referring to materials displaying nonselective absorption-that is, those that transmit the various component colors equally.

Clear glass, for instance, displays a transmittance between 80% and 90%, frosted glass between 70% and 85%, and solid opal glass between 15% and 40%. The remainder is absorbed and reflected. See Table 14.7 for typical transmission factors.

Similarly, the ratio of reflected to incident light is variously called reflectance, reflectance factor, and reflectance coefficient. Thus, if half the amount of light incident on a surface is bounced back, the reflectance is 50% (or 0.50). The remainder is absorbed, transmitted, or both. The amount of absorption and reflection depends on the type of material and the angle of light incidence, because light impinging on a surface at grazing angles tends to be reflected rather than absorbed or transmitted (FIG. 2). An example of almost perfect reflection from an opaque surface would be that from a well-silvered mirror, whereas almost complete absorption takes place on an object covered with lamp black or matte-finish black paint. The effect of the material's surface finish on reflection is shown in FIG. 3. See Table 14.9 for typical reflectance values. Reflectance measurement is discussed in Section 11.11.

The reflection that occurs on a smooth surface such as polished glass or stone is called specular reflection, as in FIG. 3a. If the surface is rough, multiple reflections take place on the many small surface projections, and the light is diffused, as in FIG. 3b. Reflectance is a measure of total light reflected; it may be specular or diffuse, or a combi nation of both, as shown in FIG. 3c.

Diffuse transmission takes place through any translucent material such as frosted glass, white glass, milky Plexiglas, tissue paper, and so on. This diffusing principle is widely employed in lighting fixtures (luminaries) to spread the light generated by the source within the fixture. Diffuse and non-diffuse transmission are illustrated in FIG. 4a and FIG. 4b.

FIG. 2 Relation between angle of incidence and percentage of reflectance. This effect is important when considering the penetration of sunlight into interior spaces and, conversely, the exterior glare produced by reflection of the sun from building windows.

FIG. 3 Reflection characteristics. (a) In specular reflection, angle of incidence equals angle of reflection (a = ß). Because 80% of light is reflected, reflectance is 80%; 20% of light is absorbed. (b) In diffuse reflection, incident light is spread in all directions by multiple reflections on the unpolished surface. Such surfaces appear equally bright from all viewing angles. (c) Most materials exhibit a combination of specular and diffuse reflection. Such a surface mirrors the source while producing a bright background.

FIG. 4 Transmission characteristics. (a) In non-diffuse transmission, the light is refracted (bent) but emerges in the same beam as it enters. Clear materials such as glass, water, and certain plastics exhibit this type of transmission. In the instance illustrated, the transmittance is 85% (the remaining 15% is reflected and absorbed). The source of light is clearly visible through the transmitting medium. (b) With diffuse transmission, the source of light is not visible and, in the case of multiple sources, the diffusing surface exhibits generally uniform brightness if the spacing between the light sources does not exceed approximately 1½ times their distance from the material.

4. TERMINOLOGY AND DEFINITIONS

Before beginning any discussion of lighting studies, techniques, and effects, it is important to have a basic understanding of the physical concepts and terminology involved and their interrelations. The System International (SI) system of units is used as the basic system by the IESNA and in this guide, whereas the lighting industry uses both the SI and Inch-Pound (I-P) systems. As in other Sections, we frequently use dual units, with the second unit enclosed in parentheses.

5. LUMINOUS INTENSITY

The SI unit of luminous intensity is the candela (candlepower), abbreviated cd (cp), and normally repre sented by the letter I. It is analogous to pressure in a hydraulic system and voltage in an electric system, and represents the force that generates the light that we see. An ordinary wax candle has a luminous intensity horizontally of approximately 1 candela, hence the name. The candela and candlepower have the same magnitude. Luminous intensity is a characteristic of the source only; it is independent of the visual sense.

FIG. 5 A source of 1-cd intensity produces 4p (12.57) lumens of light flux. Thus, each square foot (square meter) of spherical surface surrounding such a source receives 1 lumen of light flux. This quantity of light flux produces an illuminance of 1 fc (lux) on the spherical surface.

6. LUMINOUS FLUX

The unit of luminous flux, in both SI and I-P units, is the lumen (lm). If we take a 1-cd (candlepower) source that radiates light equally in all directions and surround it with a transparent sphere of 1 m (ft) radius (FIG. 5), then by definition the amount of luminous energy (flux) emanating from 1 m^2 (ft 2) of surface on the sphere is 1 lm. Because there is 4p m^2 (ft^2) surface area in such a sphere, it follows that a source of 1 candela (candlepower) intensity produces 4p, or 12.57, lm. The lumen, as luminous flux, or quantity of light, is analogous to flow in hydraulic systems and current in electric systems and is normally represented by the Greek letter f.

In physical terms, the lumen is a unit of power, like the watt. However, unlike the watt, which is a radiometric unit directly convertible to other power units such as Btu/h, the lumen is a measure of photometric power. This means light power as perceived by the human eye and therefore as a function of human physiology. Put another way, lumens, or luminous flux, is the time rate of flow of perceived luminous energy. Because the visual response of the eye is frequency-dependent, the apprehended light power is therefore also frequency-dependent, varying with the spectral content of the impinging light and the spectral sensitivity of the eye. FIG. 6a shows the spectral content of the visible energy produced by a 500-W incandescent lamp.

Measured radiometrically, it amounts to 45 W.

However, when passed through a selective filter (FIG. 6b), which is effectively what happens when the light enters the eye, the resultant "understood" light power appears as in FIG. 6c, and therefore can no longer be measured in watts. Instead, we use a unit of eye-perceived, or photometric, power called the lumen. If the spectral content curve in FIG. 6a were differently shaped, even if the total radiometrically measured power were the same, the resultant perceived power in FIG. 6c would be different.

Refer to FIG. 6b. A correlation can be made between photometric and radiometric power at the point of maximum response of the eye, which occurs at 555 nanometers (nm) wavelength- 1 nm is 10-9 m. One watt of monochromatic light at that wavelength produces 683 lm. However, because common light sources such as incandescent, fluorescent, mercury, and so on, are not monochromatic but produce light in many parts of the spectrum (see Figs. 11.47 and 11.48 and TABLE 12), no single conversion factor between watts and lumens exists. Each source has its own luminous "efficiency" (technically efficacy, in lumens/watt), determined by its spectrum. For the 500-W lamp used as an illustration in FIG. 6, its luminous efficacy is 10,000 lm/500 W, or 20 lumens per watt (lm/W or lpw).

FIG. 6 Graphical demonstration of the method by which the unit of light flux is defined. (a) The spectrum of the light produced by a 500-W incandescent lamp. It amounts to approximately 45 W measured radiometrically. When filtered by the human eye, whose spectral sensitivity curve is given in (b), this light power is perceived as shown in (c). The new light power curve is expressed in lumens and indicates the quantity of light as perceived by the eye.

7. ILLUMINANCE

One lumen of luminous flux, uniformly incident on 1 m^2 (ft^2) of area, produces an illuminance of 1 lux (lx) (footcandle [fc]). Illuminance is normally represented by the letter E. Restated, illuminance is the density of luminous power, expressed in terms of lumens per unit area. If we consider a lightbulb as analogous to a sprinkler head, then the rate of water flow would be the lumens, and the amount of water per unit time per m^2 (ft^2) of floor area would be the lux (footcandles). Thus, the SI unit, lux, is smaller than the corresponding I-P unit, footcandles, by the ratio of square meters to square feet. That is,

10.764 lux = 1 fc or multiply footcandles by 10.764 to obtain lux.

These relationships are shown in FIG. 5.

Restating mathematically yields lux lumens square meter area

(eqn. 1)

(eqn. 2)

As an approximation (with 8% error)

(eqn. 3)

8. LUMINANCE, EXITANCE, AND BRIGHTNESS

TABLE 1 Lighting Units-Conversion Factors

An object is perceived because light coming from it enters the eye. The impression received is one of object brightness. This brightness sensation, how ever, is subjective and depends not only upon the object luminance (L), but also upon the state of adaptation of the eye (see Sections 11.18 and 11.19). For this reason, the physiological sensation is generally referred to in the literature as subjective or apparent brightness, or simply brightness, whereas the measurable, reproducible state of object luminosity is its luminance (formerly photometric brightness). Luminance is normally defined in terms of intensity; it is the luminous intensity per unit of apparent (projected) area of a primary (emitting) or secondary (reflecting) light source. Thus, its units are candela per area. Specifically, the SI unit of luminance is candela per square meter (cd/m^2), some times referred to as the nit. Another unit, formerly in common use in the I-P system, is the foot-lambert.

Conversion factors for SI and I-P units (plus other, obsolete units for the convenience of readers using older sources) are given in TABLE 1. Other luminance terms such as stilb, apostilb, blondel, milli-lambert, and candela per square in. are best avoided. In this guide, the term luminance is used except where it is specifically intended to refer to the physiological sensation involved, in which case the terms brightness, subjective brightness, or apparent brightness will be used. Luminance has no readily conceivable mechanical or electrical analogy.

A word of caution is appropriate at this point. Although definitions and terminology are established for the specific purpose of accurate information exchange, the lighting literature is replete with articles, comments, and rebuttals that exist only because of the looseness of definitions and terminology. Some authors insist on applying brightness only to self-luminous surfaces, using lightness as the equivalent term for objects deriving their luminance from reflection. Thus, the sun has brightness; the moon, lightness. In this guide, we use the term brightness for the subjective reaction to either source type. Other sources point out that the luminance-brightness relation breaks down when light other than white light is used. Although this is demonstrable, it is of real interest only in theatrical lighting, where colored light is frequently used. For our purposes we assume white light, and as pointed out previously, the color accommodation characteristic of our eyes recognizes as white (colorless) light of a large chromatic range. Within that range, and for a very large range of intensities, object color is readily recognizable, and the fixed luminance brightness ratio is maintained. Contrasting word usages such as dim and dark, light and bright, clear and muddy, shallow and deep, and so on, as applied to lighting, are best left to experienced lighting designers because the terms are almost entirely subjective and therefore unhelpful to novice designers.

Another concept that the lighting designer will encounter is known as luminous exitance, or simply as exitance, which, as the name implies, describes the total luminous flux density leaving (exiting) a surface, irrespective of directivity or viewer position. For instance, if a 1 m^2 surface emits 1 lumen, its luminous exitance is 1 lumen per square meter (1 lm/m^2) or 0.093 lm/ft^2. A surface that is a perfect diffuser, whether by emitting light diffusely or reflecting light diffusely, is known as a Lambertian surface. It is fairly simple to demonstrate mathematically that the luminance of such a surface equals 1/p times its exitance. The importance of this relationship is its usefulness as an approximation.

Although very few surfaces are truly Lambertian, many are approximately so, and this relationship can be used as an engineering-accuracy approximation in many such cases.

The concept of exitance is important in detailed photometric calculations such as those involved in determining coefficients of utilization, surface luminance coefficients, and in detailed point illuminance calculations. All of these are beyond the scope of this guide because they are not usually performed by the lighting designer. Use of these derived coefficients is demonstrated in the applicable sections.

Detailed point calculations are today almost universally performed by computer, and the necessary mathematics is built into the computer pro gram. Readers interested in further background on luminous exitance are referred to Murdoch (1985).

Because object luminance is that which is visually perceived and is a prime factor in visibility (and glare), it is important that the reader be able to perform basic luminance calculations.

Although the eye does not differentiate between primary sources that generate and emit light and secondary sources that derive their luminance from reflection or transmission, the differentiation is important in calculation procedures. See FIG. 7 for a graphic representation of the basic relationships.

FIG. 7 Luminance may be either reflected or transmitted. In the former case, it is calculated as the product of the incident lumens and the reflectance; in the latter case, as the transmitted intensity divided by the projected area.

9. ILLUMINANCE MEASUREMENT

Field measurements of illuminance levels are most commonly made with a portable illuminance meter, two examples of which are illustrated in Figs. 11.8 and 11.9. These devices contain a photoelectric material connected to a microammeter via electronic control circuitry and are calibrated in lux, footcandles, or both.

FIG. 8 Electronic, digital, color-corrected and cosine corrected light (illuminance) meter from Minolta.

FIG. 9 Digital meter (from Li-Cor) has a variety of sensors that measure illuminance, solar irradiance, or photosynthetic radiation. Due to its small size, the illuminance sensor can be easily used for architectural model measurements.

As explained in Section 11.6 and as shown in FIG. 6, the human eye is not equally sensitive to the various wavelengths (colors). Maxi mum sensitivity at high illuminance levels is in the yellow-green area (wavelength of 555 nm), whereas sensitivity at the red and blue ends of the spectrum is quite low. This effect is so pronounced that 10 units of blue energy are required to produce the same visual effect as 1 unit of yellow-green.

Therefore, if a meter is to be useful, its inherent response, which is quite different from that of the human eye, must be corrected to correspond to the eye. For this reason, meters are "color corrected." The cells (meters) must also be corrected for light incident at oblique angles that does not reach the cell due to reflection from the surface glass and shielding of the light-sensitive cell by the meter housing. This correction is known as cosine correction. A good meter must therefore be color and cosine corrected (and will plainly so indicate).

Modern photometers may have considerable electronic circuitry, which provides such functions as automatic ranging, integration for flickering or time-varying sources, and connection facilities for data storage and transmission. For determining average room illuminance when using a conventional non-integrating meter, a number of readings should be taken and an average computed. Where no definite height is specified, readings are taken at 30 in. (750 mm) above the floor, a level known as the working plane because it is approximately nor mal desk height. The meter must always be held with the cell parallel to the plane of the test. Thus, to measure wall illuminance, the meter must be held with the cell parallel to the wall. If electric lighting readings are desired and the test is being conducted during daylight hours, readings should be taken with and without the electrical illumination, and the results subtracted. Detailed instructions for con ducting field surveys are contained in the IESNA publication How to Make a Lighting Survey. Briefly, a survey of an existing indoor lighting installation should establish:

1. Type, rating, and age of sources.

2. Type, design, and model of luminaires.

3. Maintenance schedule.

It should also measure:

1. Mounting height of luminaires.

2. Spacing and pattern of luminaires.

3. Reflectances of walls, floor, ceiling, and major items of furniture and equipment.

4. Illuminance levels throughout the area plus levels at all working plane elevations. In addition, vertical plane illuminance at walls and other major vertical surfaces should be measured.

The significance of vertical surface luminance is discussed in detail in subsequent Sections.

10. LUMINANCE MEASUREMENT

FIG. 10 Direct-reading, narrow-angle, spot-type luminance meter has an acceptance angle of 1º, a range of 0.001 to 299,000 cd/ m^2 (0.001 to 87000 fL), a variable response speed to permit measurement of flickering sources, and a comparison mode that permits direct luminance comparison of two sources. Results are displayed digitally. (Minolta Corp.)

FIG. 11 When the cell of a direct-reading illuminance meter is held in contact with a luminous source, the surface luminance can be read directly or simply calculated.

In terms of appreciation of the visual scene, including particularly considerations of glare, the measurement of luminance is more important and meaningful than that of illuminance (lux [fc]). This is so because it is luminance-or, more accurately, subjective brightness and brightness contrasts caused by photometric luminance-that we see, not illuminance. Light, as such, is invisible. That lux measurements are still more widely taken than luminance measurements and utilized as a gauge of the adequacy of a lighting installation is due to two factors:

1. Lux meters are cheaper and simpler to use than luminance meters.

2. Design recommendations for lighting levels are given in terms of illuminance.

Lux measurements can therefore be used as a rapid, simple method of determining whether a particular lighting installation meets these design requirements (assuming that material reflectances and reflectance ratios are correctly chosen).

Luminance meters are available in a number of configurations, one of which is shown in FIG. 10.

An approximation of the luminance of a reflecting or luminous source can be obtained using an illuminance (footcandle) meter of the type shown in FIG. 8. For diffuse reflecting surfaces, the cell of the meter is placed against the surface and then slowly retracted 2 to 4 in. (50 to 100 mm) until a constant reading is obtained. The luminance, in footlamberts, is then approximately 1.25 times the reading in footcandles, the 1.25 factor compensating for wide-angle losses.

For a diffuse luminous source, the cell of an illuminance meter is placed directly against the surface (FIG. 11); the source luminance in foot lamberts is equal to the reading on the meter in foot candles because footlamberts = lumens per area = footcandles. When using a meter calibrated in lux, the readings must be divided by p to obtain the dif fuse source luminance in cd/m^2.

11. REFLECTANCE MEASUREMENTS

It is often desirable to know the reflectance of a given surface because luminance can then be readily computed (see FIG. 7). Two methods of measuring diffuse (nonspecular) reflectance are shown in FIG. 12: the known-sample method and the light-ratio method. If a sample of known reflectance is available, this method should be used because it yields more accurate results than the ratio method.

The sample should be no smaller than 8 in. × 8 in.

(200 mm × 200 mm).

It is a good idea for an inexperienced lighting designer to determine the reflectances, illuminance, and luminance levels of spaces and surfaces familiar to him or her, such as an office desk, adjoining wall, and the like-even to the extent of marking these figures on the respective surfaces in order to develop an appreciation of and a memory for these parameters.

This enables the designer to visualize the result of a lighting design and should be of considerable assistance. See Table 14.9 for typical reflectance values.

12. INVERSE SQUARE LAW

FIG. 12 Two simple methods of measuring the diffuse reflectance of a surface.

FIG. 13 Relationship among candelas, lumens, and lux defined with reference to a standard light source of 1 mean spherical cp (1 cd) located at the center of a sphere with a 1-m radius.

FIG. 14 Demonstration of inverse square law properties using a solid angle of unit size. Note that the surfaces are necessarily spherical because points on a planar surface are not equidistant from the source.

We have already seen that, by definition, a point source of 1-cd intensity produces an illumination of 1 lux on the inside surface of a surrounding sphere of 1-m radius (r). Because the surface area of this sphere is 4p m^2, a 1-cd source produces 4p lm of luminous flux. Now, assume a sphere of 2-m radius surrounding this same source (FIG. 13). Because the same amount of flux is spread over a larger area, the illumination on the larger sphere is inversely proportional to the ratio of the sphere areas; that is,

(eqn. 4)

(eqn. 5)

In other words, the illumination is inversely proportional to the square of the distance from the source.

In general terms,

(eqn. 6)

where distance is expressed in meters (feet).

(This holds true for surfaces normal to a source. For other situations, see Section 15.)

This relationship can also readily be derived by using any solid angle and the area it intercepts, as in FIG. 14. A glance at this figure shows clearly that the area intercepted is proportional to the square of the distance from the source; therefore, the illumination is inversely proportional, as stated previously.

13. LUMINOUS INTENSITY: CANDELA MEASUREMENTS

Luminous intensity (candela [candlepower]) cannot be measured directly but must be computed from its illumination effects. The simplest way of doing this is to use the inverse square relationship developed in the preceding section. Measure the illuminance produced on a plane at right angles to the source at a known distance and apply Equation 11.6. For accurate measurement, the distance should be at least 5 and preferably 10 times the maximum dimension of the source because, for anything other than a point source, the equation is an approximation. The candela (candlepower) thus calculated is the luminous intensity in the direction being viewed. Because luminous intensity is not uniform in all directions for anything except an ideal point source, and because a single intensity figure for a source is desirable for calculation purposes, the average of a number of intensity figures taken from several directions is used. This average figure is called the mean spherical candlepower (mscp) and represents an equivalent point source that produces 4p lm for every candela.

Thus, a 10-cd lamp exhibits an average intensity of 610 cd in all directions and produces 40p lm.

14. INTENSITY DISTRIBUTION CURVES

FIG. 15 Typical luminous intensity (cd) distribution curve for a general diffuse-type luminaire. Because the unit is symmetrical about its vertical axis, only one curve need be shown. Further more, only the right side of this curve need be shown, due to symmetry.

FIG. 16 (a) Due to the asymmetry of a fluorescent luminaire, intensity distribution curves in (at least) three planes are required. (b) Photometric distribution for this fixture is symmetrical in each individual plane; therefore, only one side of a curve is required. By convention, the right side is used.

FIG. 17 Luminous intensity distribution curves plotted in rectangular coordinates. Note that candela values near the cutoff angles are easily read, which is not the case in polar plots.

If the luminous intensity figures calculated in the preceding section are plotted on polar coordinate axes, the resultant figure is called a candlepower distribution curve (CDC) for the particular source involved. The procedure for making this curve is straightforward. A photo cell is rotated around the source in a single plane, illuminance measured, and intensity (cd) calculated. Alternatively, the photo cell can be fixed and the source rotated. If the source's distribution is symmetrical, as shown in FIG. 15, then only a single set of values is required, and the resultant plot is valid in all vertical planes through the source. Thus, for incandescent lamps, downlights, open circular reflectors, and the like, only a single CDC is required. For a non-symmetric source such as a fluorescent luminaire, CDC curves in several planes are required to define the fixture's distribution characteristic.

Normally, manufacturers will provide longitudinal and crosswise curves, plus a diagonal (45º plane) curve on request. This is illustrated in FIG. 16, where the three planes and typical resultant curves are shown.

Most CDC plots are made on polar coordinates because such a plot clearly shows directions and magnitudes. Nevertheless, polar plots tend to crowd near the nadir, and accurate magnitude readings at the cutoff angle are difficult to make. For this reason, it is occasionally desirable to obtain a plot on rectangular coordinates. One such plot is shown in FIG. 17. The usefulness of intensity distribution curves will become clear in our subsequent discussions on lighting fixture diffusers, point-by point calculations, and direct and reflected glare. It should be noted that the area of the CDC curve is not a measure of the lumen output.

LIGHT AND SIGHT

15. THE EYE

FIG. 18 The human eye (a) and the camera (b) operate on similar optic principles. The cornea acts as an outer refracting lens that introduces light into the iris. The iris and pupil control the f-stop, or opening of the eye, and correspond roughly to a range of f2.1 to f11. The lens, which acts as a perfectly smooth automatic zoom lens, can focus from about 2 in. (50 mm) to infinity.

Because any discussion of light and lighting techniques is irrelevant to our purposes unless ultimately related to vision, we turn to a cursory examination of the human eye before proceeding further with discussions of lighting.

Light impinging on the eye enters through the pupil, the size of which is controlled by the iris, thereby controlling the amount of light entering the eye. The lens focuses the image on the retina, from which the optic nerve conveys the visual message by electric impulse to the brain. FIG. 18 shows the structure of the eye and the parallel structure of a camera.

Light is focused on the retina, which contains in all some 150 million light-sensitive cells of two types: rod and cone cells. The central portion of the eye, near the fovea, is an area of pin head size containing about 100,000 cone cells, which accounts for the extreme precision of foveal (center-focus) vision. The cones are responsible for the ability to discriminate detail and also give us our sensation of color and detect luminances in the range 3 to 1,000,000 cd/m^2. Proceeding outward from the fovea, a second type of cell is encountered called a rod cell. Rods can detect luminances from 1/1000 cd/m^2 to approximately 120 cd/m^2 and are extremely light-sensitive, giving a response to light 1/10,000 as bright as that required by cone cells. However, rod cells lack color sensitivity, thus accounting for the fact that in dim light (rod vision), we have no color perception and all colors appear as varying shades of gray. Rod cells also lack detail discrimination, making "night vision" quite coarse. Finally, rod cells are slower-acting than cone cells and therefore have a low degree of flicker fusion; stated conversely, they are highly motion-sensitive. Because these cells occur at the outer portions of the retina, their motion sensitivity results in our being best able to detect movement when looking out of the "corner of the eye." Looking at a 60 Hz fluorescent tube directly and then obliquely demonstrates this effect.

FIG. 19 is a sketch illustrating the angles involved in the field of vision. Of particular interest is the extreme narrowness of the cone of central (foveal) vision, in which acute perception of detail takes place. This area is so small that the eye must refocus on each dot in a colon (:) if you wish to examine each individually. Surrounding this central area is a cone of binocular vision of 30º half-angle, called the near field or surround, in which area most of the coarser sight information is gathered. Beyond this cone we have far-field and peripheral, primarily horizontal, monocular vision. It is the far-field and peripheral areas that largely give us our subjective, ambience-type reactions.

16. FACTORS IN VISUAL ACUITY

FIG. 19 The fields of vision of a normal pair of human eyes (a) and the subtended angles (b). The rectangles A and B superimposed on the field of vision in (a) represent a large magazine and a small book, respectively.

The three components of any seeing task are the object or task itself, the lighting conditions, and the observer. The following list gives the variables affecting each of these three components. Based upon the results of many investigations, they can be categorized as of primary or secondary importance.

I. The Task

Primary Factors

a. Size

b. Luminance (brightness)

c. Contrast, including color contrast

d. Exposure time-needed or given

Secondary Factors

e. Type of object-required mental activity; familiarity with the object (in reading, familiarity is so important as to become the primary factor)

f. Degree of accuracy required

g. Task-moving or stationary

h. Peripheral patterns

II. The Lighting Condition

Primary Factors

a. Illumination level (illuminance)

b. Disability glare

c. Discomfort glare

Secondary Factors

d. Luminance ratios

e. Brightness patterns

f. Chromaticity

III. The Observer

Primary Factors

a. Condition of the eyes (both health and age)

b. Adaptation level

c. Fatigue level

Secondary Factors

d. Subjective impressions; psychological reactions

Although in the following discussions these factors are considered individually, many are interrelated.

Thus luminance (Ib) and adaptation (IIIb) result from the presence of illumination (IIa); subjective impressions (IIId) are dependent on brightness patterns (IIe) and chromaticity (IIf); fatigue (IIIc) results from a combination of many of the factors, and so on.

In the literature, it is common to find reference to the quantity and quality of the lighting environment. In terms of these factors, the quantity of light has reference to item IIa and the quality to items IIb through IIf.

The basic visual tasks are the perception of low contrast, fine detail, and brightness gradient.

Assuming a good lighting environment-that is, low glare, acceptable luminance ratios, and white light plus a normal pair of un-fatigued eyes--visual acuity is primarily dependent on items Ia to Id, the interrelated effects of which have been determined by a large number of field tests.

Remember that the seeing task under discussion involves foveal vision (i.e., focusing and concentrating on small-area detail). This is a vastly different task than normal reading, where the eye rapidly scans familiar images without focusing on details, and the brain immediately understands even when much of the information is missing, as in poor reading copy. The task discussed next is quite different and could be compared to studying mathematical equations or reading an unfamiliar language or even proofreading spelling. All of these tasks require detailed examination of each symbol individually.

TABLE 2 Typical Luminance Values

TABLE 3 Preferred and Permissible Luminances

17. SIZE OF THE VISUAL OBJECT

Visual acuity is generally proportional to the physical size of the object being viewed, given fixed brightness, contrast, and exposure time. Because the actual parameter is not physical size but sub tended visual angle, visual ability can be increased by bringing the object nearer the eye (FIG. 20). It is assumed that we are dealing with a pair of young eyes, because at ages above 40, the accommodation ability of the eye becomes limited and bringing the object closer blurs the focus.

18. SUBJECTIVE BRIGHTNESS

FIG. 20 Relationship between object size and visibility is demonstrated by comparison of subtended angles a and b.

The sensation of vision, as explained previously, is caused by light entering the eye. This light may be thought of as a group of convergent rays, each ray coming from a different point in space and therefore carrying different visual information. The composite of these rays comprises the entire visual picture that the eye sees and the brain comprehends. The individual rays differ from each other in intensity and chromaticity, depending on the part of the viewed object from which they were reflected. The intensity of these cones of light determines and describes the perceived brightness of the object being viewed.

The human eye detects luminance over an astonishing range of more than 100 million to 1, the lower levels being accomplished after an adjustment period called adaptation time. This period varies from 2 minutes for cone vision to up to 40 minutes for rod vision for dark adaptation but is much faster for both types for light adaptation (going from dark to light). The effects of adaptation on apparent (photometric) brightness are discussed in the following section. Tables 2 and 3 list some measured luminances of everyday visual tasks.

An interesting characteristic of light-level adaptation is a shift in the sensitivity curve of the eye (FIG. 6b). Whereas for the light-adapted eye (photopic vision) maximum sensitivity occurs at 555 nm in the yellow-green region, the dark adapted eye (scotopic vision) peaks at 520 nm in the blue-green region. This means that as the light dims, the warm colors-yellow, orange, red- become grayed, and the blues and violets stand out.

This phenomenon can be important in the lighting design of restaurants, where light levels generally vary inversely with restaurant quality. Very few foods are blue or violet.

Returning then to the primary consideration of visual acuity as affected by luminance, we can state that, in general, visual performance increases with object luminance. However, a great deal depends on the background against which an object is viewed and the consequent contrast in brightness between the object being viewed and its surroundings.

19. CONTRAST AND ADAPTATION

The discussion that follows assumes full-spectrum white light and ignores the effects of chromaticity, which is considered separately. Many researchers in the area of visibility have concluded that contrast is the single most important factor in visual acuity.

This is self-evident when we realize that in fact, the eye sees only contrast. This can readily be demonstrated by viewing a large, evenly lighted, mono chromatic, diffuse-finish surface (preferably white) that encompasses the entire visual field. The eye is unable to focus on such a surface because it sees no contrast, but only a single luminance. Therefore, the eye itself attempts to provide the missing contrast by seeing an internally reflected view of the retina in yellow.

To properly evaluate the effect of contrast (luminance ratio) on visibility, we must first deter mine the nature of the visual task or, more simply, exactly what it is that we are trying to see. As stated before, the basic visual tasks are detail discrimination and detection of low contrast. Examples of the former are drafting, industrial product inspection, or something as simple as discriminating between the numbers 3, 6, and 8, which are similarly shaped. Detection of low contrast includes reading faint copy, sewing a black fabric with black thread, and the like.

Contrast is a dimensionless ratio, defined as (eqn. 7) where LT and LB are the luminance of the task and background, respectively, in any units. Thus, C varies from 0 for no contrast to 1.0 for maximum contrast. In most situations, the illumination on the task and background is the same. Therefore, because luminance is the product of illuminance (lux) and reflectance, contrast can also be expressed as

(eqn. 8)

where RT and RB are the reflectances of the task and the background, respectively. From this equation, we can conclude that contrast is generally independent of illumination (ignoring specularity). Thus, for the black-on-white lettering that you are now reading,

[...] which accounts for the excellent legibility. (We are assuming no specularity.) Reflectance figures are taken from TABLE 9.

It is obvious that high contrast is the critical factor in visual appreciation of outline, silhouette, and size, which are the factors involved in the task of reading. Thus, black-on-white print can be read with ease even in moonlight, which is at best 0.1 lux illuminance, because the contrast is so high (94%). An important conclusion can then be drawn about a reading task: with high contrast (clear, legible print), visibility is essentially independent of illumination above a certain minimum. Indeed, high illuminance values can be detrimental because they generally go hand-in-hand with high luminance sources, and these in turn can cause veiling reflections.

Now refer to FIG. 21. Note that as the contrast between the letters in the word "performance" and the background diminishes, the individual letters become harder to read. The end letters of the word require an illuminance of up to 1000 lux, and that suffices only because we expect the letter e at the end. Were it an unknown sign, illuminance of a magnitude of 10,000 lux or more would be needed.

These latter letters are an example of the second type of visual task mentioned previously-low contrast poor copy requiring surface detail study.

The e at the end of "performance" in FIG. 21 is printed with the same density as the C in "Contrast." It exists, and with enough lighting, the negative effect of lack of contrast can be overcome.

This is not so with copy from a used-up printer cartridge or a washed-out photocopy. There, the data simply do not exist, and increased lighting only makes this fact more evident.

FIG. 21 High contrast is helpful when the seeing task involves detection of silhouette detail.

High background luminance makes an object look darker, and therefore assists in outline detail discrimination, which is precisely the visual task involved in reading. For this reason, black-on white is desirable for reading. (This is a special case of lateral adaptation, which is discussed in Section 11.36.) Conversely, high background luminance makes surface examination more difficult. A simple experiment demonstrates this effect. Stand near a window and hold your hand in front of you with the floor as background. The skin surface detail is perfectly clear-in rough proportion to its luminance. Now hold your hand up against the window with the daytime sky as background. The hand out line is clear, but the skin surface appears dark-the brighter the sky, the darker the skin surface. The reason for this is that the eye automatically adapts to the average brightness of the entire scene.

It is well known that when using an automatic exposure control camera to photograph a dark object on a light background (such as a person in a snow scene) it is necessary to manually increase the camera aperture in order to obtain additional light to photograph the detail of the darker object. (In doing this, we overexpose the rest of the scene.) Because we cannot easily control the aperture of our eyes, we must compensate for the detrimental effect of high back ground luminance in another way-for example, by increasing the surface luminance of the visual task. Indeed, this method is frequently employed (see Section 11.31b). Limited visual compensation can be made by squinting; this reduces the field of vision and the overall scene brightness.

For maximum visual acuity, the luminance of a surface-type task should be the same as, or slightly higher than, that of the background, but ratios of 3:1 are acceptable in most circumstances.

Another way of understanding this is to consider the adaptation characteristic of the human eye (FIG. 22). As stated, the eye adapts to the brightness level of the overall scene and sees each object in the scene in the framework of that adaptation level. Thus, at an adaptation level of 1 fL (3.4 cd/m^2), a measured luminance ratio of 1:10 (horizontal scale on FIG. 22) appears to be only approximately 1:4 (vertical scale); that is, the apparent ratio is smaller than the actual one. Put another way, the low level of eye adaptation causes the eye to diminish the difference between high brightness. This effect becomes smaller as the adaptation level rises, until at an adaptation level of 1000 fL (3400 cd/m^2) (daylight conditions), the apparent and actual ratios correspond; that is, smaller ratios are recognizable. Because visual acuity is, by definition, the ability to distinguish between different levels of luminance, we have in effect demonstrated that visual acuity increases with increased adaptation level.

The second important conclusion that can be drawn is that at high adaptation levels, apparent brightness is lower than actual brightness, and vice versa. Thus, a shadowed object near a window looks darker than it actually is; contrary to first expectation, it must be better lighted than a similar object further inside the room for equal visibility. That this effect (high-level adaptation) is primarily important in daylight situations is also apparent from the curves. At a 100-fL (340 cd/ m^2) adaptation level, which is approximately that of a brightly lighted interior space, apparent and actual luminance levels coincide. The reverse effect, resulting from low adaptation levels, can be very important in design situations where low lighting levels are found, such as theaters, lecture halls, restaurants, and storage spaces. Sources of light that would be entirely acceptable at a higher adaptation level can easily become an annoying glare at a low level. Good examples are a theater usher's flashlight or the blinding glare of an oncoming car's headlights. Reduction of contrast due to veiling reflections and the measurement of contrast and contrast reduction are discussed in Section 11.29.

FIG. 22 The effect of an eye's adaptation level on perceived (subjective) brightness is clearly shown.

The foregoing discussion of contrast deliberately avoided any discussion of object colors and the effect of color contrast on visual acuity for several reasons:

• Office-type tasks (paperwork) are most often black-on-white tasks.

• Most work tasks involving colored objects deal with unsaturated colors, where the pronounced effects of color contrast are minimal.

• The effect of object color on visual acuity is very complex because it involves the color characteristic not only of the object, but also of the back ground and the surround plus the chromaticity of the illuminant.

That being said, we must, however, at least mention a number of important object color phenomena that bear on visual acuity.

• The subjective brightness of a colored (hetero chromatic) object is greater than that of an achromatic object for the same photometric luminance. This effect varies with hue and saturation. It is more pronounced with saturated colors than with those of low chroma and more in the blue-purple-red area than the yellow- green area.

• Colored objects on a dark-to-black background appear light and de-saturated. Conversely, colors on a light-to-white background appear darker and more saturated.

• Adjacent complementary colors produce a pale to-white border between them. The effect also appears when the task and background colors are complementary, and is most pronounced with saturated (high-chroma) colors.

These remarks deal with object color only. As stated at the beginning of this section, we are assuming white, full-spectrum light, so that neither object color nor visual acuity is affected. Effects on visual acuity resulting from unbalanced illuminants, such as those from high-pressure sodium lamps and, to a lesser extent, from halide and fluorescent sources, are the subject of intensive and ongoing research.

Some of the results are adduced in the discussion of illuminant chromaticity at the end of this Section.

These effects are most pronounced, and therefore most important, in relation to elderly persons and persons with visual defects. We must again, there fore, emphasize that unless otherwise specifically stated, our discussions of vision, light, and lighting assume full-spectrum white light and young, healthy eyes.

20. EXPOSURE TIME

Registering a meaningful visual image is not an instantaneous process, but one that requires finite amounts of time. Just as a photograph can be taken in dim light by using a longer exposure, so can the human eye better distinguish and discriminate fine detail in poor light given time (and neglecting eyestrain). Of course, the time needed depends on the type of task, but the principle of shorter time at higher illumination, within limits, remains the same. This is particularly true when the object being viewed is not static but is in motion.

The phenomenon, however, is not linear. For one specific task tested, increasing the luminance by a factor of 6 halved the seeing time, whereas a further six-fold increase in luminance reduced the time only another 20%. Thus, as in the case of improved contrast with increasing background brightness, we have a case of diminishing returns.

With the parameter of time, as with other parameters of visual acuity, the same qualification applies. When dealing with material that does not require detail discrimination, improved performance does not necessarily result from improved illumination. It has been amply demonstrated that speed of reading and comprehension are substantially independent of illumination levels above a minimum, but are very much dependent on the contrast quality of the material.

21. SECONDARY TASK-RELATED FACTORS

Refer again to the list of factors in visual acuity in Section 11.16. We have discussed at some length the primary task-oriented factors Ia-d, and at this point wish to consider secondary factors e-h of that list. These items refer essentially to the level of concentration required. Thus, spray painting a large metal object or packing fruit are very different from inspecting the painted object for defects or the fruit for bruises. The former tasks are largely mechanical and repetitious, whereas the latter tasks require continuous judgmental decisions based on visual information. Because both of the latter tasks are frequently moving (assembly-line type of work) and both involve penalties for inaccuracy (rejection at a later inspection state or negative feedback from the purchaser), the lighting required for these tasks is several orders of magnitude better than that for related but largely mechanical tasks. Indeed, extrapolation from laboratory condition tests yields only a range of lighting recommendations. There after, considerable field testing and adjustment are required.

An observer performing a lab test has a different level of concentration and performance than that of a person at an 8-hour-a-day task. The latter compensates for an unsatisfactory seeing condition by:

1. Moving the work to a better viewing angle.

2. Moving the head and eyes to a more comfort able position.

3. Reducing the distance between the eyes and the task to the extent that the eyes can accommodate.

4. Complaining about a poor-contrast task so that something is done about it (such as fixing the photocopying machine).

5. Taking more time to perform the seeing task involved. (This item, if it affects production, frequently spurs management to make appropriate alterations in the work environment.)

The last item in the task list-peripheral pat terns-deals with the visual surround rather than the immediate area of the work. Other than glare sources, which are discussed separately, there are many items that, although outside the central field of vision, can disturb the viewer's concentration and therefore the task performance of a worker.

These include movement (vehicles, machines, per sons), to which peripheral vision is particularly sensitive; large variations in the brightness pattern of the background caused by such activities as periodic opening of an outside door or welding; and even nonvarying patterns that are disturbing because of their very nature, such as checkerboard light-dark patterns, or devices on which it is difficult to focus the eyes, such as crossed patterns of wires and bars.

None of these items is strictly lighting-oriented, but they are noted here to demonstrate that an adequate lighting design necessarily includes adjustments for particular field conditions.

22. OBSERVER-RELATED VISIBILITY FACTORS

It is a well-documented fact that the visual performance of healthy eyes decreases with age. This reduction is demonstrated principally in two areas: an increase in minimum focusing distance caused by increasing lens rigidity and a decrease in sensitivity caused by clouding of the cornea, lens, and vitreous humor. Both of these effects can be compensated for, the former with external lensing (eyeglasses) and the latter by increased task size, luminance, contrast, and exposure time, as explained previously. The point is, of course, that the age of the worker is an important parameter in the vision equation.

Furthermore, maximum performance may not be synonymous with maximum comfort or mini mum fatigue. Indeed, the reverse may sometimes be true. Most experts agree that what is normally referred to as eyestrain is a condition of the eye muscles resulting from extensive and intensive eye use.

Thus, excellent performance under excellent lighting conditions can still produce fatigue because of the demanding nature of the task. In addition, as discussed later, discomfort glare or even excessive lighting can cause fatigue without affecting performance.

The lighting designer must be concerned not only with providing adequate uniform lighting levels, but also with all of the factors involved.

Many, indeed, are beyond his or her control. Some, such as task contrast, previously thought to be outside the lighting designer's province, should be examined by the designer in an overall "lighting plus task plus observer" problem context and recommendations made. Acceptance of these recommendations is a management decision.

Recent work in field testing of visibility situations in real-work conditions considers all of these parameters in arriving at a visibility judgment and, in addition, goes far toward identifying any difficulty as lying with the task, the lighting condition, or the observer. The entire field of visibility and visual performance is undergoing very active continuing research, and definitive answers to the elusive question of how to design for optimal viewing have not yet been found-nor indeed is there any assurance that they will be. What has emerged over the years are various parameters and criteria for judging aspects of visibility, such as Equivalent Spherical Illumination (ESI), which is a contrast-related visibility criterion; Relative Visual Performance (RVP), which rates a vision situation in terms of speed and accuracy of performance; Visual Comfort Probability (VCP), which judges the visual comfort of an overall scene; and a somewhat ephemeral and entirely subjective impression called visual clarity. Other research has concentrated on the relationship between eye pupil size and visual acuity and, based on results from this research, the relationship between pupil size and various aspects of light and lighting. What has been confirmed to date is the complexity of human work-oriented vision plus our inability thus far to adequately quantify the interrelationships among the many factors involved-quantification being the basis of reliable, duplicatable design.

23. THE AGING EYE

The past few decades have seen a remarkable increase in life expectancy in modern Western countries with a resultant sharp increase in the aged population. In the United States at this writing, about 15% of the population is above 65 years of age, a proportion that is expected to reach 20% by the year 2020. As a result, lighting design must take cognizance not only of special requirements in buildings specifically intended for use by the aged but also, increasingly, in general-use public buildings. To this end, a brief review of these special requirements is presented here.

Refer to FIG. 18. Light enters the eye through the cornea, passes through the aqueous humor, and enters the lens through the pupil. After being focused by the lens, it continues through the vitreous humor and finally projects the viewed image, reversed, on the retina. As the eye ages, a whole spectrum of physiological changes may occur; some are usual and are therefore classified as normal; others, such as cataracts, are less common but are still considered by ophthalmologists to be an expected development. The unusual developments are classified as pathologies, only indirectly or partially related to aging and therefore outside the purview of normal lighting design. The normal developments and their influence on lighting design are briefly described in the following subsections.

(a) Cornea

This perfectly clear outer lens tends to become cloudy, with corresponding reduction of visual clarity and acuity. This results in a requirement for more light to overcome the reduction in light intensity on the retina. The overall effect is very similar to that of a neutral density filter on a camera lens, the difference being that such a filter is most often used to reduce excessive ambient light without excessively reducing the shutter opening (increasing the f-stop). The need for additional light for aging eyes is recognized in most modern systems of illumination specification. See TABLE 8 for the factors used in the IESNA system.

(b) Lens

The lens, which begins life as a very lightly yellow tinted flexible crystalline body, gradually thickens and yellows. As a result of the thickening, flexibility is reduced, resulting in the well-known inability to close-focus. The yellowing both reduces the overall light intensity in the eye and selectively filters the blue portion of the spectrum. Research seems to indicate improved visual acuity through pupil size control when the incident light is rich in the blue area of the spectrum. Because lens yellowing reduces the blue frequencies, the overall effect is again to require additional light.

A second and more important degenerative phenomenon of the lens is its gradual clouding.

When the opacity is confined to the perimeter, its effect is negligible because vision is unaffected.

When small, opaque areas appear within the visual axis through the lens, vision is affected in two ways:

1. The viewed image is dimmed and blurred due to opacities in the field of view.

2. Light entering the lens is scattered by inter-reflections from the opaque particles, resulting in a subjective impression of glare. This effect is particularly severe outdoors, where light enters the lens from all angles, and both indoors and outdoors as a result of concentrated high luminance (glare) sources.

The net result of these reactions is a requirement for more light but an even more pressing requirement that sources of glare and peripheral light be eliminated. (People with this condition frequently wear eyeglasses and sunglasses with large, opaque earpieces to block peripheral light and thus reduce glare.) Also, because short-wavelength (blue) light interreflects and scatters more readily than does long-wavelength (yellow-red) light, these people are more comfortable with incandescent sources and low-color temperature fluorescent lamps (2700-3100 K) than with sources rich in the blue-green spectrum. (Ophthalmologists frequently prescribe yellow-tinted eyeglass for people with this lens condition to filter out blue light.) Finally, a less common but still prevalent condition of the aging lens is the development of fluorescent particles, called fluorigens, in the vision path. In the presence of UV radiation, such as exists in daylight, fluorescent, and high-intensity discharge (HID) sources, these particles fluoresce, causing scatter, blur, and glare. The solution to this problem is a combination of yellow-tinted eyeglass lenses and a reduction of light sources containing appreciable quantities of UV.

(c) Pupil

The pupil controls the amount of light entering the eye and is therefore intimately involved in the constantly changing accommodation level of the eye.

The pupil muscles react more slowly as they age, thus lengthening accommodation time. Dark-to light accommodation is very rapid in the young eye, and is barely noticed except for extreme changes such as exiting a cinema into sunlight. With an aging eye, the slower pupil results in severe glare sensations with even much smaller brightness changes.

The net result of all of the normal conditions of the aging eye is a heightened sensitivity to glare, an intolerance to the blue-UV end of the spectrum, and an overall requirement for higher illuminance levels. For the lighting designer, these needs trans late into requirements for very careful selection and placement of luminaires, increased use of indirect lighting, and particular attention to the spectrum of the light sources used. Because some of these requirements are not only mutually incompatible but also contrary to energy-efficient design practice, it may be particularly difficult to satisfy all the requirements in spaces occupied by persons with a wide range of ages. In work areas of this type, it may be wise to provide for the possibility of readily changing lighting conditions in a limited area to accommodate older occupants rather than attempt an overall design, keeping in mind constantly the glare and color factors discussed previously.

QUANTITY OF LIGHT

24. ILLUMINANCE LEVELS

Returning to the list of factors in Section 11.16, and having discussed the task-oriented and observer oriented items (except for item IIId, psychological reactions, which is covered in Section 11.36), we turn now to item II, the lighting condition. This is frequently, if somewhat inaccurately, divided into two groups-quantity and quality of lighting-with item IIa representing quantity and items IIb to IIf representing quality. That such a division is not accurate becomes clear in our discussion of glare in Sections 11.28 to 11.31.

An understanding of the factors involved in visual acuity, as discussed previously, does not answer the most basic lighting design question, which is "How much light must I provide for the specific visual task at hand?" That this question is extremely difficult to answer is evidenced by the fact that, even today, recommendations for similar tasks vary by ratios as high as 10:1 among countries with highly developed technologies. Because this is obviously an unsatisfactory situation in an era of global markets, international construction, the European Union, and international cooperative lighting research, the trend since the late 1980s has been to attempt a degree of standardization.

The North American (IESNA) recommendations were originally developed analytically by extrapolation from extensive laboratory tests. The function of these tests was to determine the conditions under which small differences in contrast could be detected for specific degrees of accuracy, with variable parameters of task luminance, size, and exposure time. The idea behind the tests was that visual acuity could be defined as the ability to distinguish differences in contrast.

The British (and, to a large extent, the European) approach was to study specific tasks in actual and simulated field conditions. To these results, modifying factors of size, contrast, accuracy requirement, speed, and duration were added. The disadvantage of this approach was the necessity to study a large number of visual tasks individually.

In the absence of a very specific visual task description, a detailed listing of visual categories was developed, as is discussed in the next section. Recently, IESNA recommendations have been modified by a graded system-more in line with British and CIE (Commission Internationale de l'Eclairage) recommendations-that establishes a median or average requirement for a task, within a range, and then modifies the median up or down to specific conditions of speed, accuracy, error importance, task duration, background reflectances, and viewer visual capability.

In addition, because of the moral and legal pressure (ANSI/ASHRAE/IESNA Standard 90.1 is mandated by many codes) for energy conservation, the IESNA has taken additional salutary steps toward rationalization of its very influential illumination standards. They include recognition of fatigue and task familiarity (e.g., reading) as factors in determining illuminance levels, establishment of lighting power budgets, and energy standards (see Section 13.5) that encourage use of daylight as a normal component of a space's illumination and use of task/ambient lighting design as the preferred technique where high levels of task lighting are required.

25. ILLUMINANCE CATEGORY

Before discussing illuminance (lux) recommendations, it is necessary to understand the basis of their derivation, applicability, and shortcomings. As noted in the preceding section, most of the IESNA task illuminance recommendations are derived by extrapolation from threshold contrast visibility tests that yield a required task luminance. Assuming uniform, diffuse task reflectance and uniform illuminance, it is then a simple step to calculate required illuminance, since luminance is simply the product of illuminance and reflectance. In SI units,

× RF p and in I-P units, fL = fc × RF

where

L = luminance in cd/ m^2

E = illuminance in lux

RF = reflection factor

fL = luminance in foot-lamberts

fc = illuminance in footcandles (see FIG. 7)

This simple relationship is used to derive much of the current IESNA tables, with modifying factors as explained previously (see TABLE 6 for CIBSE modifiers).

A number of reservations about this method have been voiced by respected authorities. One is based on research that indicates that suprathreshold visibility requirements are more readily related to eye brightness adaptation levels than to thresh old contrast luminance levels (see FIG. 22 and the associated text discussion). Another objection is that deriving suprathreshold luminances from threshold values depends on applied criteria and can therefore vary considerably. Still another objection is based on the readily demonstrable fact that the sensation of vision is not mathematically related to photometric luminance. Thus, a black surface of 10% reflectance illuminated with 900 lux and a white surface of 90% reflectance illuminated with 100 lux both have exactly the same luminance, yet the eye always sees the white surface as lighter than the black one by a large margin. All of these reservations add up to a recommendation to the lighting designer to use the IESNA illuminance recommendations as one aspect of the overall design and not dogmatically.

Returning to recommendations, visual task studies indicate that assuming good contrast, the required luminances, categorized by type of task, are roughly as follows:

Dependence of required illumination upon task reflectance (RF) can be seen by a glance at the following tabulation, which shows quantitatively the illuminance requirements in the previous categories for tasks of radically different reflectance.

This illustrates that a single illumination scheme is often inadequate for an area containing widely differing visual tasks. Note that a 10% RF makes all tasks difficult and that casual seeing comprises only outline recognition.

TABLE 4 Typical Illuminance Recommendations, CIBSE (UK).

Reference 25: CIBSE Lighting Guide LG3: Areas for Visual Display Terminals Notes: The information in the table will be influenced by reference to the "core recommendations" in Sections 4.3 to 4.5 of the Code.

Check that installations designed to meet the needs of visual display screen tasks also have the task-to-wall and task-to-ceiling ratios recommended in Section 4.4 of the Code.

Where air conditioning or mechanical ventilation is required, air-handling luminaires may be appropriate.

TABLE 5 Examples of Activities/Interiors Appropriate for Each Maintained Illuminance

TABLE 6 Design Maintained Illuminance Flowchart a Source: Reproduced with permission from the CIBSE Code for Interior Lighting (1994).

a. To use the chart, follow the horizontal path from the "standard" maintained illuminance in the schedule until the answer to a question is "yes." If the "yes" is strong, follow the solid arrow; if moderate, follow the dashed arrow.

b. The flowchart should be used for all standard maintained illuminance recommendations from 200 to 750 lux for general activities and interiors in the lighting schedule. For recommendations of 150 lux or less, the modifying factors are not relevant (see TABLE 5). Where a standard maintained illuminance of more than 750 lux is recommended, this always applies to a stated task for specific industries or activities where the modifying factors have usually been taken into account.

c. The standard maintained illuminance given in the schedule assumes that the task is representative of its type. If the task is much more visually difficult than usual (e.g., smaller size, lower contrast), then an increase in the maintained illuminance is appropriate. Reduced contrast may arise from the use of safety lenses or safety screens because they reduce the transmission of light. Increase illuminance to take account of the age or eyesight of the operator. Conversely, if the task detail is such that the task is easier to see than usual (e.g., larger size, higher contrast), a reduction in maintained illuminance can be made.

d. The standard maintained illuminance given in the schedule assumes that the task is to be undertaken over a conventional working period.

If the work is to be undertaken continually for a much longer period than usual, the maintained illuminance should be increased in order to diminish the risk of visual fatigue. Conversely, if the work is to be carried out over a much shorter period than usual, the maintained illuminance may be reduced.

e. The schedule assumes that the consequences of any errors are typical of the activity. However, if errors have unusually serious consequences for people, plant, or product, an increase in illuminance may be appropriate.

f. If the design maintained illuminance is more than two full steps (e.g., 300-750 lux) on the illuminance scale above the standard maintained illuminance, consideration should be given to whether changes in the task details or organization of the work are more appropriate than substantial increases in the maintained illuminance.

26. ILLUMINANCE RECOMMENDATIONS

Because American architects and engineers are involved in many construction projects outside of the United States, this section presents the essentials of British illuminance recommendations, which are similar to those of the CIE and many European countries, in addition to the American IESNA recommendations. In any specific design, the standards of that country should be consulted.

Most developed nations, including Australia, Brazil, China, France, Germany, Japan, and others, have their own published lighting standards.

(a) British Lighting Standards

These standards are published by the Chartered Institution of Building Services Engineers (CIBSE: cibse.org). The particular publication in which illuminance recommendations appear is the Code for Lighting, 2006. The method of use is to determine the recommended average illuminance level, called the standard maintained illuminance, from either the very detailed, extensive listing of specific tasks in the previously mentioned publication (a small sample of which is reproduced in TABLE 4) or, if only the representative type of task is known, from the task category chart in TABLE 5. Having established this recommendation, the designer then modifies it (if necessary) by using the flowchart shown in TABLE 6, which either increases or decreases the recommended illuminance to suit the task size, contrast, duration, and error risk. Explanatory usage notes accompany TABLE 6.

TABLE 7 Determination of Illuminance Categories

(b) North American Illuminance Recommendations (IESNA)

The North American approach to illuminance recommendations is similar in context to the CIBSE approach described in Section 11.26(a). The Illuminating Engineering Society of North American (IESNA) provides illuminance selection categories (see TABLE 7) organized around three sets of visual tasks (orientation and simple tasks, common tasks, and special tasks). Specific illuminance recommendations are provided with reference to a defined application in a matrix called the IESNA Lighting Design Guide. The Guide suggests recommended illuminance values (horizontal, vertical, or both) that represent best practice for a typical application. In addition to the lighting quantity recommendations inherent in the illuminance categories, the Lighting Design Guide provides recommendations regarding lighting quality issues.

Additional factors related to the lighting condition can be factored into the choice of illuminance.

These include scene geometry, shadowing and modeling, luminance ratios, presence of daylight, glare considerations, and other considerations. The Lighting Design Guide is a substantial document and cannot be reproduced herein. Anyone concerned with lighting design should use the current IESNA Lighting Handbook and the illuminance recommendations presented therein.

Earlier IESNA illuminance recommendations provided fairly explicit advice on adjustments to published illuminance values to accommodate differential conditions (such as the aging eye). Although no longer presented as explicitly by IESNA, such information promotes an understanding of the practical applications for illuminance recommendations. The following are excerpts addressing this adjustment procedure:

1. An illuminance category is initially selected, based either on a general description of the activity involved (TABLE 7) or, if known, on a specific activity in a specific setting. (The extensive tables for task-specific selection are published in the IESNA Lighting Handbook, 9th ed. and are not reproduced here.)

2. Following the initial illuminance selection, three adjustment factors are introduced.

Age under 40 reduces the lighting requirement, unusual demand for speed or accuracy increases it, and particularly low or high background reflectance (below 30% or above 70%) increases or reduces it, respectively.

Adjustment of recommendations should not be automatic. At the higher illuminance recommendations (categories D through G), the designer must become thoroughly familiar with the importance, duration, and visual difficulty of the specific task involved, because the design aim is task lighting and not simply general room illumination (see TABLE 5). Having determined appropriate adjustment factors from Tables 11.8 and 11.9, the designer can then establish recommended (target) illuminance values that better reflect project-specific conditions.

TABLE 8 Suggested Adjustments. a to General Lighting Illuminance Values, for Illuminance Categories A, B, and Cb

A recommended illuminance resulting from this procedure is a "raw" or conventional illuminance, that is, average maintained lux (divide by 10 for approximate fc values). This illuminance is under stood to be on the task for categories D through G, and in the room for categories A through C. There fore, the lumen method is an appropriate calculation procedure for categories A through C, and a point-by-point method is appropriate for the task illuminance related to categories D through G.

For a space with several tasks of varying visual difficulty, the designer is expected to so design the lighting and controls that task requirements are met without overlighting. A uniform layout keyed to the most severe task is visually tedious and energy-wasteful, and is strongly to be discouraged.

The IESNA-recommended illuminance values are not applicable to installations where a visual task is not the deciding factor. Such installations include merchandising spaces, displays of all sorts, theatrical and artistic lighting, lighting for mood, lighting for safety, light used as part of an industrial process, and so on. Recommended illuminances for exterior spaces are found in the same group of tables in the IESNA handbook referred to previously.

TABLE 9 Suggested Adjustments to Task Lighting Illuminance Values for Illuminance Categories D through F

QUALITY OF LIGHTING

27. CONSIDERATIONS OF LIGHTING QUALITY

Quality of lighting is a term used to describe all of the factors in a lighting installation not directly connected with quantity of illumination. Certainly it is obvious that if two identical rooms are lighted to the same average illuminance, one with a single bare bulb and the other with a luminous ceiling, there is a vast difference in the two lighting systems. This difference is in the quality of the lighting, a term that describes the overall scene-that is, the luminances, diffusion, uniformity, and chromaticity of the lighting.

Excessive luminances and/or excessive luminance ratios in the field of vision are commonly referred to as glare. The quality of the lighting sys tem must also include the visual comfort of the sys tem, that is, the absence of glare. When the glare is caused by light sources in the field of vision, it is known as direct glare. When the glare is caused by reflection of a light source in a viewed surface, it is known as reflected glare or veiling reflection (see FIG. 23). Factors affecting the severity of glare are adaptation level of the eyes, the apprehended size of the glare source, luminance ratios, room size and surface finishes, and size and position of lighting fixtures and windows. Light sources in the far-field and peripheral-vision areas beyond the central 90º cone are less troublesome as glare sources.

FIG. 23 Glare zones. Direct glare presupposes a head-up position, whereas reflected glare assumes eyes down at a reading angle.

28. DIRECT (DISCOMFORT) GLARE

The factors involved in producing discomfort glare are the luminance, size, and position of each light source in the vision field plus the adaptation level of the eye. The discomfort of direct glare stems from two facts: first, the eye adapts (rapidly) to the aver age brightness of the overall visual scene; and second, the eye is attracted to the highest luminance in that scene. (The latter fact is used effectively in merchandising displays.) Thus, if an area of high brightness, such as a window or a lighting fixture, exists in the visual scene and we are looking at an area of lower brightness, such as a work task, three visually disturbing things occur:

1. The eye adapts to a higher luminance level, thus effectively reducing the subjective brightness of the task-or, put more simply, making it harder to see what we are looking at (see FIG. 22). This is readily demonstrated by alternately blocking and unblocking a direct glare source with one's hand while trying to perform a moderately difficult visual task and noting the immediate improvement in visibility when the glare source is obscured.

2. The eye is drawn simultaneously in two directions: involuntarily to the source of high luminance and volitionally to the object we are looking at. The resultant tension causes considerable visual discomfort.

3. The adaptation level is continuously varying as the eye is drawn to the glare source and away again.

Glare is proportional to a source's luminance and its apprehended solid angle. Therefore, a small, bright source is usually not a problem, whereas a large, low-brightness source (such as a luminous ceiling) may be. Indeed, a small, bright source adds sparkle to the field of vision, and many observers find it a pleasant addition in a monotonous lighting environment. Although discomfort glare from a scene is cumulative, source luminance is more important than the number of sources. For instance, if the luminance of a number of sources is halved, the reduction in glare is greater than is achieved by reducing the number of such sources by half. Indeed, the latter procedure has little effect on discomfort glare.

The remaining two factors are less self evident. Glare decreases rapidly as the brightness source is moved away from the direct line of vision; thus, the glare produced depends on the source's position in the field of view. The amount of discomfort glare produced by a source is inversely proportional to the background luminance (eye adaptation level). Thus, a ceiling fixture with a luminance of 4000 cd/ m^2 at 65º might easily constitute a source of discomfort glare in a space with an eye adaptation level of 150 cd/ m^2. The same fixture would not be objectionable in a daylight condition, where the eye adaptation level might be 1500 cd/m^2. A more striking example is that of an automobile's headlights, which at night are so severe a source of glare as to constitute disabling glare, whereas in daylight, with its concomitant high eye adaptation level, these lights are very noticeable but not usually disturbing.

Keeping in mind the dependence of direct glare on eye adaptation level, a useful rule of thumb is that the luminance of large sources should not exceed 2500 cd/ m^2 and that of small sources should not exceed 7500 cd/ m^2. The former is roughly the luminance of blue sky; the latter approximates that of a fluorescent lamp. The terms "large" and "small" depend not only on the actual physical dimensions of the source but also on the distance from the observer. That is, the actual criterion is apprehended size, or subtended visual angle, as shown in FIG. 20.

FIG. 24 Determination of direct glare. The glare contribution of each source depends on its size (subtended or apprehended solid angle), luminance, and location in the field of view. Note that the apprehended solid angle of a small source is such that even with high luminance, it is not objectionable. Such sources are normally called sparkle. Glare is much more objectionable with a dark background than with a light one; therefore, light-colored paints on ceilings and upper walls are recommended.

The glare effect of a number of individual glare source contributions in an interior space can be quantified by the criterion called visual comfort probability (VCP), which is defined as the percentage of normal-vision observers who will be comfortable in that specific visual environment. The IESNA has established a set of standard conditions for which VCP of sources can be calculated. These include a 1000-lux illuminance, representative room dimensions, fixture height and observer position, and a head-up field of view limited to 53º above and directly forward from the observer (FIG. 24).

With these conditions, direct glare will not be a problem if all three of the following conditions are satisfied:

1. The VCP is 70 or more.

2. The ratio of maximum-to-average luminaire luminance does not exceed 5:1 (preferably 3:1) at 45º, 55º, 65º, 75º, and 85º from the nadir, crosswise and lengthwise.

3. Maximum luminaire luminances crosswise and lengthwise do not exceed the following:

A typical set of manufacturer's luminance and VCP data is shown in FIG. 25 for a ceiling mounted fluorescent fixture with 4-40WT12 lamps. Note that all VCP values are considerably above the 70 minimum criterion. If full VCP data of this type are not available, they can be calculated with almost any lighting calculation program, given the luminaire luminance data.

Despite the usefulness of the VCP criterion, it is inherently limited by its own standard conditions, which are not easily applied to other situations:

1. In small spaces, VCP has little significance.

2. Tabulated VCP figures are given for the worst viewing position in the room. Because VCP varies dramatically with observer position, the VCP values given are always lower than the space's average VCP.

In view of these (and other) reservations, most of which tend to make the actual direct glare situation better than the VCP calculation would indicate, it is recommended that layouts giving a VCP somewhat below 70 not be discarded out of hand. Instead, they should be examined carefully and, if possible, several observer positions calculated using one of the many readily available PC-based computer pro grams. With these as a guide, the designer can usually rearrange and substitute equipment to obtain the desired condition.

See Section 15.4 for a comparison of the direct glare characteristics of lighting fixture diffusers.

29. VEILING REFLECTIONS AND REFLECTED GLARE

FIG. 25 A typical set of manufacturer's published VCP and luminance data.

Although there is no generally accepted convention with respect to nomenclature, many people refer to reflected glare when dealing with specular (polished or mirror) surfaces and to veiling reflections when considering source reflections in dull or semi matte finish surfaces, which always exhibit some degree of specularity. This discussion uses the terms interchangeably.

(a) Nature of the Problem

The problem of veiling reflections is much more complex than that of direct glare because it involves both the source and the task and is inherent in the act of seeing (FIG. 26). Vision is produced by light being reflected from the object seen.

Thus, if the object being viewed were replaced by a mirror, we would see the source(s) of light clearly (FIG. 26a). In commercial spaces there are usually one or more lighting fixtures near the observer that furnish most of the light by which to see. These principal sources are the main contributors to reflected glare. Other, more remote fixtures in the room are lesser sources of veiling reflections (FIG. 26b).

To the extent that the sources can be seen in the vision task, glare exists. It is imperative to an understanding of this problem to appreciate the importance of the reflection characteristic of the object being viewed. If the object were perfectly absorbent-that is, if it had a reflection coefficient of 0%-it would appear completely black, as no light would be reflected into the eye (FIG. 26c). Conversely, if the object were perfectly specular, like a clean mirror, and no light source were within the geometry of reflection, it too would appear black (FIG. 26d). Thus, if we took a clean mirror out on a moonless night and shined a light on it from over our shoulder, it would be practically invisible because no light would be reflected back into our eyes.

The reader might try this experiment: in an inside space with a single overhead luminaire, try to examine the surface of a very clean, dust-free mirror. You will find that the best angle to hold it is almost at the angle at which the light source is seen. This is because the mirror is almost completely specular, and it is the slight diffuse reflection near the viewing angle that permits us to see the surface.

This means that reflected glare is due to task surface specularity, whereas object definition, that is, the ability to see the task itself, is due to task surface diffuseness. A corollary of this conclusion is that veiling reflections, which are caused by mirroring of a source in the task, are proportional to source luminance and substantially independent of illuminance level. The brighter the source, the more troublesome its reflection.

Glare sources within the geometry of reflected vision are shown in FIG. 27, and the effects are shown in FIG. 28. FIG. 27 clearly shows that although large sources are difficult to avoid, small sources can usually be easily avoided by a small change in the source-task-eye geometry, such as by moving the head or tilting the task. TABLE 10 lists a few sample reflectance figures to demonstrate that most materials exhibit both specular and diffuse reflectance. In studying FIG. 27, it is important to note that a majority of visual work is done in the zone of 20º to 40º from the vertical, below the eye, with a maximum at the 25º reading angle (FIG. 29).

FIG. 26 (a) The nature of the seeing process requires that light from the source(s) be reflected by the task into the eye. (b) The light entering the eye is the sum of all of the reflected light, specular and diffuse, from all sources in the direction of the eye. If the task is specular, all of the sources will be seen reflected in the task. (c) A perfectly absorptive object is jet black because it reflects nothing. (d) A perfectly reflective object positioned as shown is also black because geometrically it cannot reflect light into the eyes.

(b) Contrast Reduction

The principal effect of the reflection of a light source in a visual object is to reduce contrast between the object and its background, thus reducing visibility. It is as if a bright veil were spread over the object being viewed, which accounts for the term veiling reflection. As the angle of the incident light approaches the viewing angle, the specularly reflected component of this light becomes more and more pronounced, and task contrast drops. This is clearly visible in Figs. 11.28 and 11.29. The worst situation occurs when the incident angle equals the viewing angle.

When the specular reflectance of the task and background is high, as with the glass screen of a visual display terminal, for instance, an image of the source is superimposed on the object, making viewing impossible (FIG. 29). However, even with the highly specular finish of "slick" magazine paper, vision is still possible because of the very high contrast between black ink and white paper, although with much reduced clarity and considerable annoyance.

When considering specular and diffuse reflectance, the equation for contrast given in Section 11.19 must be rewritten as

(eqn. 9)

where LB and LT are background and task luminances caused by diffuse (D) and specular (S) reflectance; that is, LBS is the background luminance due to its specular reflectance, and so forth. If we were to rework the calculation of contrast of Section 11.19, including specularity, the result would be much different.

FIG. 27 Geometry of reflected glare. (a) Because normal desktop, head-down viewing angles vary from 20º to 40º from the vertical, the offending zone is the area on the ceiling corresponding to specular reflection between these two angles. Note that the higher the ceiling, the larger this area becomes. (b) In an office situation, the draftsman would see ceiling fixtures in the offending zone reflected in his instruments, parallel straightedge, and work. Note the important fact that the offending zone moves back and becomes smaller as the table tilts up.

FIG. 28 Note the effect of light in the area of the ceiling that causes reflected glare; contrast is reduced, and the print washes out.

The usual viewing angle to a horizontal surface is between 20º and 40º from the vertical; we show 25º because it is the most common viewing angle. With vertical incident light on a diffuse surface (a), such as the pages of a textbook, the print is dark and clear. When the angle of light incidence is equal to the viewing angle (b), we have a mirror reflection situation. Even with diffuse paper, the print is light at best and almost invisible at worst. As the angle of incidence becomes larger (c), reflected glare decreases. When the incident light is at a very low angle (d), there is little reflected glare, but all the print appears lighter.

FIG. 29 (a) Normal viewing angle of a task on a horizontal surface ranges from 20º to 40º from the vertical. Most common viewing angle is 25º. (b) Graph showing contrast reduction of a task with a specular background (such as a sheet of glossy paper) as a function of the angle between incident light and the normal to the task surface. Viewing angle is assumed to be 25º from normal. Note that between 22º and 27º the contrast is negative. This indicates that background luminance exceeds that of the task, making the task essentially invisible. What is visible, clearly, is a reflected (mirrored) image of the source.

FIG. 30 (a) This unit is a high-precision electronic digital instrument that measures luminance, luminance ratio, contrast, and contrast reduction. Use of the unit is shown in (b), (c), and (d). Luminance measurement range is 0 to 200,000 cd/ m^2 (56,400 fL) at high accuracy. (b) The meter measures contrast reduction at a specific viewing angle. (c) Note that with a lighting fixture directly in the offending zone-that is, above and in front of the viewer-a severe loss of contrast occurs over much of the work surface. (d) By shifting the relative position of the viewer and the source so that no source exists in the offending zone, contrast reduction is held to 3% to 4% over most of the work surface. Contrast reduction in the normal work area should not exceed 15%. (Brüel & Kjær.)

TABLE 10 Typical Reflectances

A similar calculation for clear, black typewritten material on good white bond paper yields a contrast reduction from 94% to 77%, or R = 17%. This 77% figure simply serves to emphasize the fact that with high task to background contrast, effective seeing is possible almost regardless of the lighting condition. However, this most emphatically does not relieve the lighting designer of the responsibility to provide a comfortable and efficient lighting environment in which pronounced veiling reflections do not exist. In general, any contrast reduction of more than 15% is undesirable.

Because both specular and diffuse reflectances frequently vary with the angle of view, and exact figures are rarely available, accurate calculation is difficult. If a lighting system exists or a mock-up can be made, measurements of contrast reduction can be made accurately with a contrast/luminance meter of the type shown in FIG. 30. A standard contrast device that is designed to correspond to a normal office task (black typeface and white paper background) is positioned on the work surface and exposed to the ambient illumination. The task contrast is then measured at the same angle at which it would normally be viewed. Contrast reduction is automatically calculated and displayed.

A contrast reduction map of the work surface can thus easily be made (FIG. 30c). Thereafter, changes can be made in the position of the work, viewer, or illumination sources to minimize contrast reduction. Note the pronounced effect of simply shifting the source out of the ceiling glare source zone (offending zone). Contrast reduction in the primary work area should not exceed 15% for good task visibility with specular work items such as glossy papers.

FIG. 32 A test classroom illuminated by three widely spaced rows of four-lamp fixtures with lens-type wraparound diffusers.

Observer positions are shown by arrows. The row of fixtures in front of position M4 is too far forward to be in the offending zone.

30. EQUIVALENT SPHERICAL ILLUMINATION AND RELATIVE VISUAL PERFORMANCE

(a) Equivalent Spherical Illumination

Another way of approaching the problem of contrast reduction is to define a reference lighting system that is effectively free of veiling reflections and then relate actual lighting systems to it with a figure of merit. Conversely, one can measure the effectiveness of a given lighting system in terms of the equivalent glare-free system. Both of these ideas, which are essentially the same, are the basis of the concept of equivalent spherical illumination (ESI).

In order to achieve a lighting system almost free of reflected glare, it is necessary to construct an enclosed volume whose surfaces are uniformly diffusely reflective and whose primary source is obscured to the maximum extent possible. As illustrated (FIG. 31), the integrating sphere is such a device. Light is introduced from the outside, split by a deflector, and evenly distributed throughout the sphere by the multiple reflections from the white painted walls. The result is an evenly illuminated volume. When a task is introduced, the illumination falling on it is entirely uniform; that is, there are no high-luminance sources reflected in it. It is therefore termed spherically illuminated. (Note the parallel to diffuse sky illumination.) The extent to which any other illumination system can duplicate this glare-free environment is that system's equivalent spherical illumination (ESI), representing the portion of its total illumination that is spherical- that is, diffuse and glare-free. ESI is determined by comparing contrast rendition in the spherical and test systems.

A study of school lighting (FIG. 32) gave the illustrated results for four viewing positions in a classroom lighted with ceiling-mounted continuous rows of 2-ft × 4-ft, 4-lamp, 40-W fluorescent fixtures with lens-type wraparound diffusers on 10 ft centers. Carefully note that:

1. ESI depends entirely on the viewing position and viewing angle, other factors in the space being equal.

2. In an ostensibly very well lighted (215 fc) position (M1), the glare-free illuminance is only 28 fc!

This does not mean that visual work in this position is impossible. It does mean, emphatically, that in position M1, a pronounced veiling reflection exists on all specular objects. (Because of the size, orientation, and location of the glare source, this reflected glare is difficult to avoid.) It further means that a large amount of energy is being utilized (effectively wasted) to produce essentially negative results.

FIG. 31 Spherical illumination is produced by illuminating an object by diffuse reflection from the inside walls of an integrating sphere. The light source and observer are normally external.

The results could have been anticipated, at least qualitatively, by examination of the observer positions vis-à-vis the layout. Positions M1 and M3 have bright sources in the offending zone-M1 more so than M3-as is borne out by the results.

M2 is an excellent position in that it receives light contributions from the two sides, its illuminance value being lower than the others due to wide row spacing. M4 is ideally placed; no glare sources are in the offending zone and a row of fixtures is positioned behind it, which makes it geometrically impossible to act as a glare source. The ESI analysis gives quantitative expression to our qualitative judgment and, as such, is a valuable design tool.

Note particularly that the ESI results shown in FIG. 32 clearly correspond to the results of a similar test made with the contrast meter, as shown in the charts of FIG. 30.

As with VCP criteria for direct glare, so with ESI; there are ameliorating factors that generally make a given lighting system better than these criteria figures would indicate. Some of these factors have already been mentioned but bear repetition.

1. ESI is critically dependent on observer position and viewing angle. Although position is generally fixed by chair location, observers can and do change their viewing angle and head aspect to correct for glare situations.

2. The nature of the task (i.e., its specularity) is assumed to be fixed and unique. In some situations the task nature varies, and thus also the contrast. When tasks are constant, severe veiling reflections frequently lead to measures being taken that improve the task, the lighting, or both.

3. The lighting distribution characteristic of the fixture involved is a critical factor in glare production. The characteristic of a wraparound lens diffuser is such (see Section 11.31c[4]), and Section 15.2) that considerable light falls in the glare zone. Other diffuser characteristics yield different results.

The concept and use of ESI have come under considerable criticism in the professional lighting literature because ESI addresses contrast, which is not identical with visibility; it requires recalculation for even the small geometry changes that can change contrast dramatically; it tracks raw footcandles; and it is based on extrapolation from threshold conditions, the efficacy of which has been called into doubt. These published reservations and criticisms of ESI have disturbed the lighting profession sufficiently to result in wide abandonment of its use and its disappearance from most modern computer lighting programs.

The ESI procedure, however, does exactly what it is designed to do: to point out locations of poor lighting geometry immediately and quantitatively, and to flag luminaires with unsuitable distribution characteristics for the proposed use. That is, as stated in the 1981 IES Lighting Handbook, it is "used as a tool in determining the effectiveness of controlling veiling reflections and as part of the evaluation of lighting systems."

(b) Relative Visual Performance

In more recent years, a metric called relative visual performance (RVP) has appeared in the literature and has gained considerable acceptance. It tests (also via computer calculation) the effectiveness (i.e., the relative [to perfection] visual performance of a given visual environment) of task accomplishment in regard to speed and accuracy. Like ESI, it is based on luminance and contrast, but unlike ESI, it judges the relative performance of a task rather than simply contrast reduction. It seems to ignore the discomfort of veiling reflections if the task can be performed efficiently. The author has computed RVP for a number of common lighting layouts with various types of lighting fixtures.

The results give uniformly high values for RVP varying between 0.95 and 0.99, which to our mind makes effective judgment of glare situations very difficult. The reader is encouraged to use the available lighting programs to calculate ESI and RVP for proposed lighting layouts and, where possible, to compare the results with completed construction.

FIG. 33 (a) If luminaires are kept out of the trapezoidal offending zone, contrast will be excellent. If the bulk of one or more luminaires projects into this zone, and in particular into the critical zone, contrast will drop sharply. The dimensions shown are for a flat desk 3 ft × 5 ft (0.91 m × 1.5 m) and a 9-ft (2.7-m) ceiling height. (b) The dependence of the glare zone on table tilt is illustrated. The offending zone becomes smaller as the table is raised, so that with a table near the vertical position, glare is all but eliminated.

31. CONTROL OF REFLECTED GLARE

Because the causes of veiling reflections are well understood, it would seem that a solution to the problem should long since have been developed.

Unfortunately, this is not the case. Although there is no known lighting method or material that completely eliminates veiling reflections, there are a number of techniques that minimize contrast loss due to veiling reflections while maintaining adequate illumination. These are:

• Physical arrangement of sources, task, and observer so that reflected glare is minimal.

• Adjusting brightness (eye adaptation level) so that objectionable brightness is minimized.

• Design of the light source so that it causes mini mal reflected glare.

• Changing the task quality.

(a) Physical Arrangement of System Elements

Arrange the lighting geometry to avoid sources of high luminance at reflection angles when dealing with specular tasks. This is often difficult to accomplish in modern offices, which frequently have both horizontal and vertical work surfaces, the latter being the specular screen surface of a visual display terminal (VDT). The latter problem is so widespread that today it effectively governs the design of office lighting. As should be clear from Figs. 11.30 and 11.32 and the related discussion, in a space using multiple sources, particularly in continuous rows, placing the work between rows with the line of sight parallel to the long axis of the units is an effective technique (see FIG. 32, position M2). Position M4 is dangerous in that the center row can be a source of reflections. In this case it is not, due to the 10-ft row spacing. Note that the offending zone for horizontal tasks depends on the tilt of the desk. Thus, for a horizontal desk, the offending zone is forward of the desk, as in FIG. 33a; with an elevated table, the ceiling glare source zone may well be behind the source, as in FIG. 33b.

All of the geometric solutions mentioned pre suppose a detailed, fixed furniture layout, a situation that obtains in many but certainly not all cases. In the absence of such data, two alternatives are possible: a uniform layout with furniture adjusted to it, or vice versa. In practice, a combination of both is the most practical approach. Because low watts per square foot budgets have made ducted lighting fixture heat removal systems (air troffers) much less prevalent, fixtures are easily shifted. This mobility is further enhanced by the extreme flexibility of lighting fixtures fed from ceiling plug-in raceways.

FIG. 34 shows such a rearrangement, which results in saving five fixtures, a load reduction of 800 W, and an improvement in visibility.

FIG. 34 The original uniform fixture layout utilized three rows of six 2 × 4 ft (0.6 × 1.2 m), four-lamp fixtures, giving a total load of 2880 W, a load density of 2.6 W/ft^2 (28 W/m^2), and a uniform illumination level of approximately 90 (raw) fc (900 lux). The original layout is shown dotted and numbered. The rearranged layout uses 13 fixtures (shown shaded) for a total 2080 W, a load density of 1.9 W/ft ^2 (20.5 W/ m^2), and more than 100 ESI fc (1000 lux) on each work surface. In addition, five fixtures are saved. Note: This level of illuminance is justified only for difficult visual tasks.

(b) Control of Area Brightness and Eye Adaptation Level

As discussed in Section 11.19, loss of contrast can be compensated for (and glare reduced) by increased overall non-glare illumination. In so doing, we are simply making the task brighter to override the detrimental veiling reflection. The problem with this technique, however, is that a large increase in illuminance is required to overcome the glare. This increase can, in many instances, be most practically accomplished not by increasing overall room illumination, with the associated extremely high energy consumption, but by adding a supplementary task-lighting source so arranged as to be free of reflected glare. By making this supplementary source's position adjustable (as in FIG. 27b), we accomplish three things:

1. Veiling reflection is overcome.

2. The high level of illumination needed for exacting tasks is provided with minimum energy expenditure.

3. The observer is granted complete control, with resultant optimum lamp placement plus psychological satisfaction that generally prevents worker complaint. (The optimum position is generally to the left and slightly forward of the task.)

We can demonstrate the effectiveness of a supple mental desk lamp by returning to Example 11.5.

FIG. 35 A concentration of light in the glare zone (a) produces the largest amount of reflected glare. As the number of light sources is increased (b) in the glare zone and luminance is decreased, reflected glare is decreased. The least glare is from an all-luminous ceiling, which also has the lowest luminance (c).

FIG. 36 With a high-reflectance diffuse finish ceiling, this semi-direct perimeter lighting installation yields a higher ESI illuminance than raw illuminance at the viewing location illustrated, indicating excellent contrast rendering. The plastic lens at the bottom of the fixture serves to provide perceived light source luminance and to avoid an impression of gloominess, despite the satisfactory overall luminance level.

(c) Control of Source Characteristics

The reflected luminance that causes loss of contrast is proportional to the luminaire's luminance at a given viewing angle, and therefore may be reduced by reducing luminaire luminance at that angle. This can be accomplished in four ways:

Dimming or switching lamps (see number 1 following)

Using luminaires with lower overall luminance (see number 2 following)

Using the luminaire as a primary source to illuminate a large, low-brightness secondary source (see number 3 following)

Reduce the luminaire luminance only at the offending angles (see number 4 following)

1. Reducing the total output of a fixture also reduces its output in the critical portion of the ceiling glare zone and can actually increase the ESI illuminance (i.e., improve task contrast).

2. In lieu of using a few small high-output sources, utilize larger-area, low-output sources (FIG. 35). This has the effect of reducing the source luminance in the ceiling glare zone while increasing the illumination contribution from outside the glare zone, resulting in better contrast for the same or lower illuminance level (lux). The disadvantage of this technique is an increased lighting fixture cost.

3. To overcome the economic disadvantage of multiple low-output, low-luminance sources, the ceiling can be used as a secondary source illuminated from high-output indirect or semi indirect fixtures. These sources, which can be fluorescent or HID (e.g., metal-halide), have the advantage of high efficiency. The space's ceiling height must be sufficient to permit suspending the unit while avoiding "hot spots" on the ceiling. The minimum suspension length depends on the luminaire characteristics and is normally provided by the manufacturer.

To ensure high efficiency, the ceiling should be painted with a high-reflectivity matte white paint and kept clean. Results obtained from a semi-indirect installation using 1500 mA, very-high-output lamps are shown in FIG. 36.

4. Because most horizontal task vision takes place between 20º and 40º from the vertical (see Figs. 11.27 and 11.30), any fixture that emits little or no light below 40º from the horizontal cannot produce veiling reflection, regardless of its position in the field of view (FIG. 37). As a result, diffuser manufacturers produce prismatic diffusers the output of which is diminished below 30º and above 60º in order to minimize both reflected and direct glare. Due to the characteristic shape of the distribution curve, they are known industry-wide as batwing diffusers. For observers positioned so that their sight lines are parallel to the longitudinal axis of the ceiling fixtures, lenses with linear (side-to-side) batwing characteristics perform well. If the observing position varies in aspect with respect to the fixture, a radial batwing curve (in all directions) is required. Note carefully that these diffusers have only limited usefulness in reducing reflected glare in specular vertical surfaces (VDT screens). All types of diffusers and their characteristics are discussed in Section 15.4.

FIG. 37 Glare zones are 0º to 45º and 45º to 85º for reflected and direct glare, respectively. Therefore, a diffuser that emphasizes the 30º to 60º zone will be least objectionable on both counts.

(d) Changing the Task Quality

At this point, it should be clear that reducing the task specularity is at least as effective a means of reducing veiling reflections as changing the lighting system characteristics, if not more so. It is therefore recommended that task contrast and specularity be actively considered and recommendations made in a framework of energy- and cost-effectiveness. Thus, to produce adequate visibility it is often cheaper and always more energy economical to upgrade the task (in the visibility sense) than to change the lighting system.

32. LUMINANCE RATIOS

As explained previously, other factors being equal, visual performance increases with contrast-that is, with the difference in luminance between the object being viewed and its immediate surroundings. Conversely, however, the difference between the average luminance of the visual field (task) and the remainder of the field of vision should be low to avoid the discomfort of large, rapid changes in eye adaptation level. Restated, contrast is desirable in the object of view but undesirable in the wider surrounding field of view.

Providing reflectances of 50%, 30%, and 80% for walls, floor, and ceiling, respectively, and 35% for furniture, establishes a fairly high eye adaptation level so that direct glare (which results from excessive luminances in the field of view) is minimized. Recommendations for maximum luminance ratios to achieve a comfortable environment are presented in TABLE 11. Effective visual performance is entirely possible in environments with much higher ratios. They are simply not as visually comfortable and may be fatiguing.

TABLE 11 Recommended Maximum Luminance Ratios

FIG. 38 The reflected glare from luminaires disappears when a piece of light, diffuse linoleum is placed over the dark, polished desktop. Light-colored desktops with 35% to 50% reflectance result in task-to-background ratios within the 3:1 recommended range. Before the linoleum was placed, a reflection similar to the one seen on the right also existed on the left of the desk due to another luminaire.

To achieve the recommended luminance ratios, it is obviously necessary to carefully control the reflectances of the major surfaces in a room. The reflectance figures given are averages for work-type commercial and educational spaces. The marked difference between a background with proper reflectance and one with excessive brightness ratios caused by the low surrounding reflectances is shown in FIG. 38.

FIG. 39 Totally diffuse lighting (a) destroys texture, whereas a combination of diffuse and directional lighting (b) produces the required modeling shadows.

33. PATTERNS OF LUMINANCE: SUBJECTIVE REACTIONS TO LIGHTING

In the list of characteristics in Section 11.16, we included among the secondary factors in illumination patterns of luminance-that is, the patterns of light and shadow in a space resulting from the illumination. Thus, a single source may produce sharp shadows, whereas a luminous ceiling or a completely indirect illumination system produces almost completely diffuse light. Diffusion is the degree to which light is shadowless and is therefore a function of the number of directions from which light impinges on a particular point and their relative intensities.

Perfect diffusion, rarely obtainable (or desirable), would have equal intensities of light impinging from all directions, therefore yielding no shadows. The only naturally occurring example of perfectly diffuse lighting is a daytime fog, which we know to be extremely disturbing to the eye, demonstrating that some directivity is desirable. Diffusion can be judged by the depth and sharpness of shadows. A room with well-diffused illumination resulting from multiple sources and high room surface reflectances yields soft multiple shadows that do not obscure the visual task. Because purely diffuse lighting is monotonous and not entirely conducive to extended periods of work effectiveness, some directional lighting is often introduced as an adjunct to diffuse general lighting, to lend interest by producing shadows and brightness variations. Where texture must be examined or surface imperfections detected by grazing angle reflections, highly directional lighting is required.

Indeed, as seen in FIG. 39, directional light is what creates shape and is precisely the characteristic best used to influence architectural space and form.

Sections 13.10 through 13.15, which deal with systems of lighting, illustrate a few of the light/dark patterns produced by different lighting arrangements. The combinations of uplighting and down-lighting are legion; each produces its own shadows and modeling, and each has a quality of its own. It is very much in the interest of the lighting designer to be familiar with these effects so that he or she can mentally visualize them as the design progresses. Indeed, it would be well for a designer to prepare a reference sketchbook of such shadow diagrams. It is these patterns of light and darkness that give the ambience and the subjective reactions of sociability/isolation, clarity/fuzziness, spaciousness/ crampedness, simplicity/clutter, formality/informality, boredom/excitement, definition/shapelessness, and so on. Indeed, many lighting designers begin a lighting design by sketching pictorially the area to be lighted, showing the patterns of brightness and shadow desired to achieve their objective.

This technique is obviously most useful in non-office areas such as lobbies, waiting areas, all types of recreational spaces, restaurants, merchandising spaces, and so on. In office areas the technique can also be applied to provide the points of visual interest referred to in our discussion.

Color has a great deal to do with subjective reactions and is discussed separately. The subject of psychological reactions to the lighting environment is extensive and complex, and can be touched on here only to the extent of mentioning a few of the salient lighting techniques and their usual subjective responses.

In addition to modeling and texture accent, small, high-brightness sources, usually called sparkle, create points of interest and visual excitement. Lighting installations generally yield a sense of vividness or activity proportional to the level of illumination. This is not the case with very diffuse lighted areas, which, even at high illuminance levels, are tedious. This is particularly noticeable in large, luminous ceiling installations that are especially oppressive when the ceiling is low. Small, exposed incandescent lamps, a brightly lighted, rough-textured wall, and pendant fixtures with pierced reflectors are some of the techniques used to create visual interest.

Visual attention can be drawn by high brightness. This well-known fact is used constantly in displaying merchandise. Note the following usual reactions:

• A 3:1 luminance ratio between object and surround will be noticed, but usually will not affect behavior or draw attention.

• A 10:1 luminance ratio will attract attention and, if interesting, hold it.

• A 50:1 luminance ratio or higher will highlight the object thus illuminated, practically to the exclusion of all else in the field of view.

Because areas of high luminance draw the eye's attention, all of the individual brightness sources in the field of view produce an overall impression.

If there is some form or order or pattern to them (as a pattern of lighting fixtures), then the overall impression is not disturbing-it can be thought of as visually harmonious. If, however, they are in disarray, they produce a discordancy in the eye precisely as noise produces discordancy in the ear. This visual "noise" is frequently referred to as visual clutter and can be very disturbing. The designer is well advised to keep this important fact in mind when arranging light sources that are the primary sources of luminance in an enclosed space. Aspects of fixture patterning are shown in Section 13.16.

Other subjective reactions to lighting on which there is wide consensus are:

1. Bright walls (about 25% of the horizontal light level) increase the impression of spaciousness.

Conversely, dark walls diminish a space. As a corollary, high fixture luminance attracts the eye away from the walls and diminishes spaciousness.

2. Worker-adjustable task lights increase the feeling of control and therefore comfort.

3. Downlights (and color highlights) increase feelings of relaxation and comfort.

4. Hidden-source indirect lighting and very-low brightness lighting fixtures cause discomfort because of the inability to locate the source of light.

FIG. 40 Approximate color temperatures of common illuminants.

FUNDAMENTALS OF COLOR

The subject of color is vast. Here we consider only a few aspects of color that it is imperative that the lighting designer understand. Furthermore, because it is difficult to discuss color without actually using it, the coverage is brief.

34. COLOR TEMPERATURE

A light source is often designated with a color temperature, such as 3400 K for halogen lamps, 4200 K for certain fluorescent tubes, and so on. This nomenclature derives from the fact that when a light-absorbing body (called a blackbody) is heated, it first glows deep red, then cherry red, then orange, until it finally becomes blue-white hot. The color of the light radiated is thus related to its temperature.

Therefore, by developing a blackbody color temperature scale, we can compare the color of a light source to this scale and assign to it a color tempera ture-that is, the temperature to which a blackbody must be heated to radiate a light similar in color to the color of the source in question. Temperature is measured in Kelvin, which is a scale that has its zero point at -460ºF. FIG. 40 shows the assigned color temperature of some common light sources.

Strictly speaking, a color temperature can be assigned only to a light source that produces light by heating, such as the incandescent lamp. Other sources, such as fluorescent lamps, produce light by processes that are detailed in Section 12. Such sources are assigned a correlated color temperature (CCT), which is the temperature of a blackbody whose chromaticity most nearly matches that of the light source. For such sources there is no relation whatever between their operating temperature and the color of the light produced.

Any non-spectral color illuminant is composed of two or more component color illuminants. When such a composite light-for example, white-falls on a surface other than black or white, selective absorption occurs. The component colors are absorbed in different proportions so that the light reflected or transmitted is composed of a new combination of the same colors that impinged on the surface. Thus, a white light reflected from a red wall acquires a red tint because the component colors of the white light other than red were absorbed in greater proportion than the red. When reflected, the red light takes prominence, thus giving the reflected light a red tint. This is illustrated in FIG. 41a.

Similarly, a white light passed through a piece of red glass emerges as a reddish light because the other components are absorbed in much greater proportions than the red. This well-known phenomenon is illustrated in FIG. 41b.

It is this phenomenon that allows us to see color at all; the individual object pigmentation absorbs other colors of light and reflects or transmits to the eye only its own hue in greater concentration than in the incident light.

35. OBJECT COLOR

The color of the illuminant (light) and, correspondingly, the coloration of the objects within a space constitute an important facet of the lighting quality. The two factors, however, must not be considered separately because by definition the color of an object is its ability to modify the color of light incident on it by selective absorption. The color reflected or transmitted is apprehended by the eye as the color of the object. An object is technically said to be "colorless" (not transparent) when it does not exhibit selective absorption, reflecting and absorbing the various components of the incident light non-selectively. Thus, white, black, and all shades of gray are colorless, neutral, achromatic, or, more precisely, lacking in hue.

Hue is defined as that attribute by which we recognize and therefore describe colors as red, yellow, green, blue, and so on. Just as it is possible to form a series from white to black with the intermediate grays, it is possible to do the same with a hue.

The difference between the resultant colors of the same hue so arranged is called brilliance or value. White is the most brilliant of the neutral colors and black the least; pink is a more brilliant red hue than ruby; and golden yellow is a more brilliant (lighter) yellow hue than raw umber.

Colors of the same hue and brilliance may still differ from each other in saturation, which is an indication of the vividness of hue or the difference of the color from gray. Thus, pure gray (black plus white) has no hue; as we add color, we change the saturation without changing the brilliance. The three characteristics, then, that define a particular coloration are hue, brilliance, and saturation. Using these terms, we may define "bay" as a color red- yellow in hue, of low brilliance and low saturation, whereas carmine is a color red in hue, of low brilliance and very high saturation.

FIG. 41 Selective absorption of materials relative to (a) reflected light and (b) transmitted light.

Various systems of color classification have been devised, including the ISCC-NBS (Inter-Society Color Council-National Bureau of Standards) color system, the Munsell Color System, the Ostwald Color System, and the CIE Chromaticity Diagram. In the well-known and widely used Munsell Color System (FIG. 42), brilliance is referred to as value and saturation as chroma; thus, a color is defined by hue, value, and chroma. The brilliance (value) of a pigment or coloration is related to its reflectance to white light. The higher the brilliance or value, the higher the reflectance, as might be expected when one considers that white and black are the poles of brilliance. Chroma or saturation may be thought of as either the difference from gray or the purity of the color. Spectral colors have 100% purity and therefore maximum chroma.

When white is added to a pigment, it produces a tint; adding black produces a shade. When pigments are mixed to produce a particular color, the color is created by a subtractive process. That is, each pigment absorbs certain proportions of full spectrum white light; when mixed, the absorptions combine to subtract (absorb) various colors of the white spectra and leave only those colors that finally constitute the hue, value, and chroma of the pigment. This subtractive effect is also utilized when producing colors by filtering white light. Each filter selectively absorbs component colors, transmit ting only the component desired. Thus, a blue filter transmits only blue, and so on (see FIG. 41b). Conversely, when lights of the three primary colors-red, green, and blue-are combined, they form white by an additive process (FIG. 43). FIG. 42 The Munsell Color System defines a color by three characteristics: hue (color), chroma (saturation), and value (grayness).

The additive and subtractive primary colors are complementary; they combine to give a white or neutral gray, respectively. Thus, red and blue green, blue and yellow, and green and magenta are complementary. Therefore, if a red object is illuminated with blue-green light, the object's color appears gray because the red pigment absorbs the blue-green and reflects nothing; hence the gray.

This accounts for the once common "lost red car" in parking lots illuminated with clear mercury lamps with their characteristic blue-green color. This phenomenon is today rare, as clear mercury lamps have fallen into disuse in favor of metal-halide and sodium lamps.

FIG. 43 Primary and complementary colors. Complementary color pairs are shown by arrows. Pigments form color by an absorptive (subtractive) process; colored lights form colors by an additive process.

36. REACTIONS TO COLOR

Light of a particular hue (other than white) is rarely used for general illumination except to create a special atmosphere. When a space is lighted with colored light, the eye adapts by a phenomenon known as color constancy so that it can, to a considerable degree (depending on the chromaticity of the light), recognize colors of objects despite the spectral quality of the illuminant. Thus, even when wearing heavily tinted sunglasses, we can still distinguish the color of objects quite easily. Indeed, after only a very short while, we no longer notice the green, yellow, blue, amber, or other color cast caused by the tinted lenses. However, the eyes become more sensitive to the missing colors that would make up white light. This phenomenon can be used to make meat look redder on a butcher's counter by using blue-rich, red-poor, cool white lighting in the remainder of the store.

A similar phenomenon occurs when the eye is exposed to a monochromatic scene, where the chromaticity is due to coloration of the objects rather than the illumination. The eye in such a situation becomes sensitized to the complementary color; thus, if after looking at a green surface one shifts the gaze to a white surface, one sees the complementary red color. Returning to our meat market, the use of green paint on the walls also enhances the redness of the meat. This effect in reverse also partly accounts for the extensive use of green for paints, linens, gowns, and so on in operating rooms. The eyes of the surgeons and nurses when diverted from the redness of the surgical area are more comfort able seeing green on a green background than on a white one.

By a process known as lateral adaptation, the apparent color of an object changes when the background color is changed. Thus, a green object looks somewhat blue-green on a yellow back ground because the eye is supplying the complementary color to yellow-that is, blue. Similarly, the same green object looks slightly yellow-green when on a blue background, the eye supplying the yellow.

Apparent brightness of a color is a function of its hue, in that light colors appear lighter than dark colors even when measured luminance is the same.

Thus, spaces may be defined by color within an area of equal illumination. Also, all colors tend to appear less saturated; that is, they appear "washed out" when illumination is high. Thus, pigments of high saturation (chroma) must be used in well-lit spaces if they are to be effective, although extensive use of saturated colors is generally best avoided.

Other well-known psychological effects of colors are the coolness of blues and greens and the warmth of reds and yellows. Thus, cool colors might well be used in a fur salon and warm colors in a display of summer wear. Red and yellow are "advancing" colors because objects lit with them tend to advance toward the observer, giving the appearance of becoming larger. The opposite effect is noted with blue and green, accounting for their being known as "receding" colors.

A practical, energy-saving application of these color phenomena would be to use warm colors to compensate somewhat for lowered thermostats in the winter and cool colors for the opposite effect in summer. How to accomplish this without the expense of repainting twice a year is left to the ingenuity of the architect and interior designer. In an atmosphere designed to be calm and restful, greens should generally predominate either in illuminant color, object color, or both, except in eating areas, which should be lighted with reds and yellows because cool colors are generally unappetizing.

Yellows and browns emphasize motion sickness, whereas blues and greens tend toward the reverse.

Warm and saturated colors produce activity; conversely, cool, unsaturated colors are conducive to meditation. Cool colors also seem to shorten time passage and are well applied in areas of dull, repetitive work.

A further discussion of color control, illuminant colors, color measurement, and color matching is found in the next sections, dealing with spectral energy distribution of sources and the color rendering index.

37. CHROMATICITY

The CIE color system is the internationally accepted standard for designating illuminant color. In this system, the relative proportions of each of the three primary colors (red, green, and blue) required to produce a given illuminant color are calculated.

These values are called the tri-stimulus values for that color and are designated by capital letters: X (red), Y (green), and Z (blue). See FIG. 47 for an example of a chromaticity diagram.

FIG. 44 Spectral energy distribution of several fluorescent lamp types with their correlated color temperatures (CCT) and color rendering indices (CRI). The curves are not truly continuous, but consist of individual color lines; they are shown connected for simplicity.

Because only a radiating black body has a true color temperature, a source with mixed color illuminants is assigned a CCT, which is the temperature of a blackbody radiator whose chromaticity most nearly matches that of the light source. (These graphs are generic and show patterns of wavelength distribution; consult manufacturers' data for output spectra for a specific lamp.)

38. SPECTRAL DISTRIBUTION OF LIGHT SOURCES

In addition to providing sufficient light of adequate quality, the lighting designer must be concerned with the spectral content of the selected illuminant, because perceived object color depends heavily on the illuminant. As discussed earlier, perceived object color is the result of selective absorption and reflection of components of the illuminating light by the pigments of the object being viewed. It is therefore necessary that the illuminant contain the color of the object in order for us to see the object's color. It is not so obvious that the relative energy of an illuminant at a particular wavelength deter mines the saturation and brilliance with which we see a color. To understand this, refer to FIG. 44.

In graphic form, the relative spectral energy distributions of a few common light sources have been plotted (as a function of wavelength-that is, color). If we compare the graphs in Figs. 11.44a and 11.44c, which show the spectral content of two of the most common light sources, cool white and warm white fluorescent lamps, respectively, we note that the principal difference lies in the amount of blue in their spectrum. As a result, a blue object will be bright under cool white light and dull (grayed) under warm white light. The situation is more pronounced with the standard high-pressure sodium lamp (FIG. 45c), compared to the clear metal-halide lamp (FIG. 45b). A blue object will be gray under the sodium lamp and unrecognizable as blue, whereas under the metal-halide lamp its blue color will show clearly, and so on.

FIG. 45 Spectral energy distribution of typical high-intensity discharge (HID) lamps with their CCT and CRI. (These graphs are generic and show patterns of wavelength distribution; consult manufacturers' data for output spectra for a specific lamp.)

This concern for perceived object color, which relates not only to furnishings but also to paints and prefinished construction materials such as carpets or floor tiles, is quite properly the province of the architect and lighting designer, who in turn must possess the necessary knowledge and information to make the appropriate choices of both illuminant and object color. Spectral composition graphs of the types shown in Figs. 11.44 through 11.46 are available from manufacturers for all light sources, and they should be examined when considering the characteristics of a particular light source.

FIG. 46 (a) Standard photopic (cones) eye sensitivity curve. Note that maximum sensitivity occurs in the daylight range of 500-750 nm. (b) Spectral energy distribution of two specific types of daylight. North light, with a color temperature of 8000-10,000 K, peaks in the blue range, whereas noon daylight contains all spectral colors in roughly equal proportions. (c) Tungsten-halogen lamps are incandescent light sources and therefore contain all spectral colors. As wattages increase, the color changes from orange-red to white, and the CRI drops slightly. (d) A simple filament-type incandescent lamp is very close to being a blackbody radiator (i.e., its actual temperature and CCT are almost the same). This is indicated by its high CRI (97).

One of the best ways to compare illuminants is first to expose a dull white surface to the illuminants, side by side but separated by an opaque divider, to get an impression of the illuminant color, and then expose a series of colored chips-again, side by side-to see which colors are brightened and which are grayed. The intensity of illumination also influences the appearance of colors, and it must be considered in choosing object colors. As intensity is increased, reflection increases, particularly with pale tints (high value) that contain much white pigment and thus tend to wash out color. Therefore, with high-intensity lighting, saturation of colors should be high for true, brilliant color rendition. Refer now to FIG. 46. Note from diagrams b, c, and d that the spectrum of a light source that produces light as a result of heating is continuous.

Sunlight is equal in spectrum to a blackbody radiator at 5500 K; north light is equal to one at about 8000 to 10,000 K; a 500-W incandescent lamp is approximately equal to one at 2850 K-and so on.

If the spectrum of a blackbody radiator is plotted on a chromaticity diagram, its locus is a continuous curved line, as seen in FIG. 47. The chromaticity of all true blackbody radiators falls exactly on this line, with the location depending on tempera ture. Daylight, for most purposes, falls on this locus, although because of selective atmospheric absorption and other phenomena, it is actually slightly off.

Incandescent lamp chromaticity is very close to this locus because it is also a heat-light radiator.

A source that produces light by means of individual phosphors can also have chromaticity on this locus if the phosphors are selected to produce a continuous spectrum similar to that of a blackbody radiator. Thus, we see in FIG. 47 that tri-phosphor fluorescent lamps (FIG. 44d) and metal-halide lamps (FIG. 45d) have spectral components over the entire spectrum, yielding chromaticities fairly close to the blackbody locus. For other sources, the CCT is established by their chromaticity locus in relation to the diagonal lines crossing the blackbody locus, as seen in FIG. 47. Each of these lines is isothermal-that is, all chromaticities on it have the same CCT. Thus, the reader can see that two sources with widely differing spectral content and therefore object color rendering can have the same CCT.

FIG. 47 Portion of a chromaticity diagram showing the relation of common illuminant chromaticities to that of the blackbody locus.

Illuminants whose coordinates fall on the same line crossing the blackbody locus have the same CCT but may have entirely different component colors.

39. COLOR RENDERING INDEX

Color rendering is defined as the degree to which perceived colors of objects illuminated by a test source conform to the colors of the same objects as illuminated by a reference source. The color rendering index (CRI) of a source is a two-part concept, comprising a color temperature that establishes the reference standard and a number that indicates how closely the illuminant approaches the standard. The standard is always daylight at that color temperature.

Therefore, the CRI of a lamp is really a measure of how closely it approximates daylight of the same color temperature. Two sources cannot be com pared unless their color temperatures are equal or quite close. A CRI of 100 indicates an illuminant whose spectral content is equal to daylight of that temperature. CRIs for typical common lamps are given in Figs. 11.44 through 11.46.

TABLE 12 lists the color characteristics of a few of the major sources. An illuminant's own color appearance on a neutral surface depends on its own spectral content, but if the observer is placed in a space illuminated with this source, after a short exposure time the eye becomes adapted to the source color and detects only a degree of whiteness rather than an actual tint.

Where it is necessary to detect small color differences between two objects, a light poor in object color or complementary to the object color should be used at a relatively high illumination level, followed by a light high in object color at the same illumination level. If this is not possible, two widely different but broad-spectrum illuminants should be used, preferably at the same illumination level.

Another technique is the use of a special, fixed-color source. For a full discussion, see the current IESNA Lighting Handbook.

FIG. 48 A hand-held lightweight (˜8 oz [225 g]) chromaticity meter for determining the chromaticity and color temperature of a light source. The unit reads X and Y coordinates, color temperature in degrees Kelvin, and illuminance in lux. (Minolta Corp.)

TABLE 12 Effect of Illuminant on Object Colors

It should be remembered in all considerations of color, comparison, matching, and rendering that object color depends on the spectral energy distribution of the light source (illuminant), and therefore any change in the spectral content changes the object's appearance. Two sources of the same color temperature and, therefore, apparent whiteness can have quite different spectral content and therefore render object colors differently. A case in point would be a 3000-K warm white fluorescent tube and an incandescent lamp (500-W photoflood) of approximately the same color temperature. Color temperature is an expression of dominant color, not spectral distribution.

A convenient hand-held chromaticity meter is illustrated in FIG. 48. This unit measures the X and Y coordinates of an illuminant, which can then be plotted on a standard CIE color diagram to determine absolute chromaticity. This is very useful in comparing illuminants to predict the color response and avoid color metamerisms. The meter also reads color temperature in kelvin and illuminance in lux.

References and Resources

CIBSE. 1994. CIBSE Code for Interior Lighting. Chartered Institution of Building Services Engineers. London.

CIBSE. 1996. The Visual Environment for Display Screen Use (CIBSE Lighting Guide LG03). Chartered Institution of Building Services Engineers. London.

CIBSE. 2006. CIBSE Code for Lighting. Chartered Institution of Building Services Engineers. London.

Cotton, H. 1960. Principles of Illumination. John Wiley & Sons. New York.

IESNA. 1963. "How to Make a Lighting Survey." Illuminating Engineering, Vol. 57, No. 2, 87-100.

IESNA. 1993. IESNA Lighting Handbook, 8th ed. Illuminating Engineering Society of North America. New York.

IESNA. 2000. IESNA Lighting Handbook, 9th ed. Illuminating Engineering Society of North America. New York.

Light and Color. The Franklin Institute: International Technologies, The Light Measurement Handbook: Lighting Design Lab ( Seattle): Lighting Research Center: Murdoch, J. B. 1985. Illumination Engineering. Macmillan. New York.

Pacific Energy Center, Online Resource Center (see Lighting Fact Sheets and Lighting Application Notes): Ross and Baruzzini, Inc. 1975. "Energy Conservation Applied to Office Lighting" (Conservation Paper No. 18). Federal Energy Administration. Washington, DC.

Sampson, F. K. 1970. Contrast Rendition in School Lighting. Educational Facilities Laboratory. New York.

Prev. | Next

Home  Similar articles   top of page