Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
5 Lighting and Reflections Experience reveals that lighting and reflections may cause problems for video display terminal operators. The reflection of a lamp or a bright window on any viewing screen makes it difficult or impossible to see the picture. Even if the picture is visible through the glare, the reflection may be distracting and annoy- ing. Almost everyone who has watched television or worked at a VDT, microfilm viewer, or similar display device has probably had such experiences. Many complaints reported by VDT operators are specifically related to workplace lighting and reflections. Several studies involving such complaints are reviewed and analyzed later in this chapter. The basic principles and most of the details of how lighting and reflections affect visual performance and comfort are known. Lighting specification systems based on principles of geometry, the physics of light and reflectance, and characteristics of the human visual system are routinely used by illuminating engineers and other lighting specialists to design appropriate lighting for workplaces. Good lighting design also includes esthetic and other considerations intended to promote appropriate psychological and social reactions. Although lighting specification systems differ in their specific assumptions, criteria, and spatial resolution, they are fundamentally the same. An experienced, well-trained light- ing specialist could use these systems to design appropriate lighting for VDT workplaces and to predict or explain problems that may result from inappropriate lighting design. Workplaces in which VDTs are used require lighting designs that differ in simple ways from those required in no - VDT workplaces. The differences primarily involve geometrical relationships--the presence of VDTs may complicate lighting design by adding (or substituting) work surfaces in different positions and planes.
112 Just as some workplaces are poorly designed in terms of lighting and workstation arrangements for non-VDT work, so some are poorly designed for VDT work. It is likely that many offices designed for desk-top paperwork are now being used for VDT work without appropriate modifications of the lighting. Higher rates of complaints for work involving VDTs compared to other types of work may largely be due to inappropriate lighting for the VDT situations. As lighting is improved for VDT installations, many of the problems attributed to VDTs may vanish. ILLUMINATION Illumination in offices and other workplaces comes from light sources, windows, and reflections from a variety of objects and surfaces. VDTs differ from most other task objects or surfaces in that they emit light, and they usually have a highly specular curved glass surface in a more vertical plane. These differences have important consequences with respect to illumination and reflections. For example, VDT operators in some situations see reflections of their faces and clothing on the display screen. Such reflections can be annoying and distracting, and in some cases may reduce task visibility enough to affect performance. An analysis of the basic characteristics of illumination should be helpful in determining the possible role each may play in causing problems for VDT operators. The four major characteristics of illumination are spectral composition, temporal changes, intensity, and spatial or directional aspects. There is little indication or reason to believe that spectral composition or temporal changes of illumination are responsible for complaints or problems peculiar to VDT use. Illumination having unusual or extreme color or flicker character- istics may interact with display characteristics in special ways that may cause problems. (Display characteristics, including color and flicker, are discussed in Chapter 4.) It is unlikely that such conditions were involved in any of the studies reviewed in this report. Although some people have negative attitudes and reactions to commonly used illumination (certain spectral compositions and flicker frequencies), there is no indication that these conditions have special importance in VDT situations. For the other two major characteristics of illumination, intensity and spatial aspects, there are important differences between VDT and non-VDT work situations. Our review and analysis of these characteristics of illumination is divided into three parts: problems caused by successive viewing of different luminances (which can lead to transient adaptation), reflections, and glare.
113 Transient Adaptation Transient adaptation refers to the temporary loss in visibility that occurs when a person changes his or her point of regard to sur- faces having different luminances or when illumination changes occur naturally in the visual environment. In general, the greater the ratio of change or difference in luminance levels, the greater the loss in visibility (Boynton et al., 1969; Rinalducci and Beare, 1974, 1975~. The results of research on transient adaptation have typically been interpreted in terms of the ratio of the contrast threshold of the target in the transient state of adaptation to the contrast threshold after complete adaptation to the new prevailing lumi- nance level. This ratio (symbolized by t) represents the increased amount of light needed to see the target in the transient versus the steady state and thus is indicative of loss in visibility as a function of luminance change. Figure 5.1 shows ~ as a function of the ratio of background field change. Figure 5.2 shows log ~ and plotted as a function of the log ratio of background field change and compares data obtained by Boynton and coworkers (1969) and Rinalducci and Beare (1974~. Here, as with most research of this nature, the transient state threshold is measured 300 msec (~) after the change from the prevailing luminance level, B1, to the new luminance level, B2. Visibility losses due to transient adaptation in VDT operations occur when an operator looks toward a glare source (e.g., a window or a luminaire) and then back to the display screen. The decrement in visibility should be particularly large for a positive- contrast display (light characters on a dark background). A second situation that could involve losses in visibility due to transient adaptation is when a positive-contrast display is combined with a negative-contrast source document, such as a typewritten page. Visibility losses may also occur when secondary task lighting is used on the source document. In this situation, a VDT operator using a positive-contrast display may be particularly prone to the effects of transient adaptation, especially when a negative- contrast source document is used. Rupp (1981) has reviewed a number of European and Canadian documents that recommend standards for VDT design and use. Two of the documents reviewed- - report of the Technical University of Berlin (Cakir et al., 1978) and the German DIN draft Standard 66234- - xpress some concern with transient adaptation effects that may occur when an operator continually looks back and forth between a positive- contrast display and a negative-contrast source document. Rupp appears unconvinced that such effects are significant, citing the review of MacLeod (1978) and the research on scotopic adaptation
114 8 7 6 l Maximum for / Al I Conditions / Investigated / 5 _ 4 3 _ 2 1 Overage / / ~ / // _' / ,o'~Minimum _ ~ 3 1 0 30 1 00 300 1 ,000 FACTOR BY WHICH PREVAILING ADAPTING LEVEL IS CHANGED FIGURE 5.1 Phi as a function of the factor of change from one luminance to another. SOURCE: Boynton and Miller (1963~. by Barlow and Andrews (1973), and suggests that the visual system's level of adaptation is determined by the luminance of the light symbols and not by an integrated luminance level or background luminance level. Thus, if this hypothesis is correct, one might be more concerned with matching the luminance of the light symbols with the source document background. However, the evidence presented in support of Rupp's hypothesis is either unconvincing or inappropriate, and we believe the situation needs to be examined further (also see the discussion in Chapter 4~.
115 Tau = +300 ms 0.7 0.6 0.5 0.4 o 0.3 1 0.2 _ ~~ %~% 0.1 _ In_ Q 1 1 - Log (B2/B1 ): -2 (B2/B1 ): 0.01 FIGURE 5.2 Log phi as a function of log ratio of backgrounds. Data from an experiment employing the same ratios by Boynton et al. (1969) are included to allow comparison with effects of similar changes from higher initial luminances. NOTE: The right-hand ordinate and lower abscissa scales make the figure direct-reading for phi as a function of the ratio of the backgrounds. SOURCE: Rinalducci and Beare (1974~. o ._ B1 = 0.02 fL Size = 10.6' 0~0 B1 =4Q.OfL ~ B 40 f J Size = 12' (from Boynton etal., 1969), 5.01 _ 3.98 p ,~ / _ 2.51 "'/ ~ / ,' / ,' ~ / ,' /, d, At _-// ~/ 1 +1 10 -1 o 0.1 1 3.16 1.99 1.58 1.26 +2 100 _ 1 .00 v, *" o - - Reflections CRT screens are usually convex, spherical shells of glass with a radius of curvature of approximately 63.5 cm. Reflections from the mirrorlike front surface of the screen form images. A very distant object seen by reflection in the screen will appear to be located at one-half the distance from the screen to its center of curvature, i.e., at 31.75 cm behind the screen. For an operator at 70 cm from the screen, dioptric accommodation levels for the screen and the reflected image are 1.43 diopters and 0.98 diopter, respectively. The reflected image of an object located close to the screen will also be closer to the plane of the CRT face. For example, if an operator at 70 cm from the screen sees his or her face reflected from the screen, the image will appear to be located about 22 cm behind the screen. Reflected images of luminaires or windows can produce a veil of light (reflected glare) over a portion of the screen. They can also serve as distracting or annoying stimuli that may cause discomfort or affect perfor- mance indirectly by distracting or changing the motivation of the
116 operator (see Petherbridge and Hopkinson, 1955, for evidence from a non-VDT study). Since reflected images form at distances other than that of the screen surface, accommodation and convergence may fluctuate or otherwise be inappropriate for viewing the screen. This effect may be annoying, induce discomfort, or affect performance. Reflected images may also affect performance if an operator looks directly at them (possibly due to phototropism), which could cause transient adaptation problems (DeBoer, 1977~. Reflected images could also induce binocular rivalry, which might cause - discomfort or affect task performance (Reitmaier, 1979)e The inside phosphor surface of a CRT screen reflects light in a diffuse manner rather than imaging it. Light also excites the phosphor, increasing its illuminance. Both effects can reduce contrast. Many VDTs have adjustments for screen brightness (illuminance) and contrast that can compensate for this type of effect except in extreme cases. An analysis and discussion of relationships between problems reported by VDT operators, such as ocular discomfort and fatigue, and physiological optics variables, such as accommodation, fixa- tion, convergence, and binocular rivalry, is included in Chapter 7. Glare Glare is the sensation produced by luminances within the visual field that are sufficiently greater than the luminance to which the eyes are adapted to cause annoyance, discomfort, or loss in visual performance and visibility (Kaufman and Christensen, 1972~. The magnitude of the sensation of glare depends on factors such as the size, position, and luminance of the light source or reflecting surface, the number of light sources, and the luminance to which the eyes are adapted. Reflected glare is the result of specular reflections from polished or glossy surfaces or diffuse reflections that produce a veil of light that reduces contrast. Disability gore, which may be caused by light scattered within the eye (reducing contrast at the retina), or by reflected glare, reduces visual performance and visibility. Discomfort glare produces discomfort, and it may, but does not necessarily, interfere with · . ~ . .. ... visual performance or VlSlDlllty, Just as disability glare may or may not be accompanied by discomfort. The large reported differences among individuals in sensitiv- ity to glare, as well as the great variability among studies, may be due to problems in methodology, especially in studies of discom- fort glare (see, e.g., Lulla and Bennett, 1981~. The lack of a clear understanding of how glare induces discomfort (see Chapter 7j
117 also makes the analysis and interpretation of problems attributed by VDT operators to glare difficult and uncertain. Nevertheless, several models and mathematical expressions for describing the effects of glare on visual comfort probability (VCP) have been formulated (see, e.g., Kaufman and Christensen, 1972~. Extensive research dating from the 1 920s (Holladay, 1926; Nowakowski, 1926) has provided the basis for such models. References to much of this research and other important issues related to discomfort glare are included in a report of the Commission Internationale de l'Eclairage (1 980~. The report also includes a proposed CIE glare formula by H. D. Einhorn with a discussion of its rationale, quantitative aspects, and the significance and choice of scaling factors. The formula identifies the important variables and indicates how they are related: CGI = 10 log 0.1 (L w) X Ed/500 (Ee) where CGI = CIE glare index (provisional name) L = Luminance of a glare source, in cd/m2 w = Solid angle of source, in steradian P = Guth position index Ed = Direct vertical illuminance at eye due to all sources, in lux Ee = Vertical illuminance at eye, in lux. Ee includes the indirect illuminance: Ee = Ed + Ei The position index P is based on Luckiesh-Guth's research. For computer work it is best expressed as: 1 d2E + 0.12 (1-E) P d2 + 1.5d + 4.6 with E = exp (- 0.18 s2/d ~ 0.011 s3/d) where d = forward distance of source/height s = sideways distance of source-height (forward) means in the direction of the line of sight, sideways means perpendicular to it, height means height above eye level.)
118 In general, the higher the luminance of a glare source, the larger the source, the lower the background luminance, and the closer the source is to the line of sight, the greater is the capacity of the source to produce discomfort. The position index, P. is direc- tional; a glare source located horizontally to the line of sight has greater potential for producing discomfort than an equivalent source located the same angular distance directly above the line of sight. The use of comprehensive formulas of this type has not been reported in studies of VDTs. However, the potential for VDTs to induce discomfort glare can be estimated from reported lumi- nance measurements and data from basic research on discomfort glare. Table 5.1 shows data (Gush, 1951) relating background luminance and glare-source size to glare-source luminance at the borderline between comfort and discomfort (BCD). Assuming that these data accurately represent discomfort glare thresholds for at least some VDT operators, it can be seen from Table 5.1 that some situations would induce discomfort. Fellmann and coworkers, using an illumination of 150 lux on eight different brands of VDTs, reported !uminances ranging from 2-7 cd/m2 for screen backgrounds, 8-110 cd/m2 for consoles, and 8-45 cd/m2 for keyboards (Fellmann et al., 1981~. These values represent the background luminances (the approximate levels to which th operators would be adapted), that are best represented in the table by the 3.4 and 34 cd/m values. Luminances in excess of 1,000 cd/m2 from potential glare sources (e.g., windows and luminaires) have been measured in actual VDT workstations (Cakir et al., 1978; National Institute for Occupational Safety and Health, 1981~. These values are glare source luminances. Values above the BCD values in the table would induce discomfort. (Note, however, that Guth used a flashing glare source. It is not clear how much the BCD values from steady sources in natural settings would differ. Eye movements and blinks would interrupt the retinal images of steady sources.) Several combinations of the measured or assumed conditions would produce discomfort glare. A lighted environment that is properly designed and therefore comfortable for workers performing traditional desk-top tasks may not be comfortable for workers performing tasks involving VDTs for two reasons. First, the design of general office lighting assumes a depressed line of sight; however, when a VDT screen is viewed, the line of sight is at or near horizontal. The higher line of sight needed to view the screen brings ceiling luminaires closer to the line of sight, resulting in a higher glare index and a greater likelihood of discomfort glare. Second, the temporally and spatially averaged luminance is lower for positive~ontrast VDTs, which results in a higher glare value.
119 In In - CD 3 a: .E o A: ._ Cm: $- E An a o 3 m - o m LO m o Cal o _ 00 ~ - U. ._ o U. a, 00 O Cal o V, Cd - at v Cal o o A: Cal ._ To - ._ I A: o Cal ~0 Go ~0 lo =1 1- O ~ O 00 ~0 00 O O 00 ~4 ~ 00 - ~ - o ~ - oo - ~ - ~ oo - ~ oo o ~ ~ oo c~ c~ ~ ~ - o ~ ~4 o ~ oo ~ - o oo ~ - oo - o ~ c~ ~ ~> a~ ~ a5 ~ Co E o - $ - Ct C~ ._ s~ o - ._ o C~ - C~ C~ ._ - s~ o o o ._ Cq U) ._ ~' s ._ 3 ._ ~_ - - - - . . o CQ
120 Glare problems have been reported in several studies involving VDTs. Although there are problems in the methods used and the interpretation of results in some of these studies (see Chapters 2 and 7 and Appendix A), there is little doubt that lighting and reflections were responsible for some of the reported operator complaints and problems. A general analytic model for predicting task visibility under different lighting conditions may be useful in analyzing VDT task situations. This model is ~ n - S VL = Cref X - of-- X CRF X DGF X TAF 0.0923 where (as defined in Commission Internationale de l'Eclairage, 1981 ): VL = Visibility level, a measure of the extent to which the equivalent contrast of a task visual display exceeds the visibility threshold of an observer for the same display at the same level of task background luminance, measured in units of the observer's threshold contrast. - Cref = Reference equivalent contrast, the value of equivalent contrast of given task details under reference lighting. RCS = Relative contrast sensitivity, proportional values of contrast sensitivity expressed relative to the value , · ~ obtained with a luminance of 100 cd/m~. CRF = Contrast rendering factor, the measure of the visibility of a task in a real lighting installation in comparison with its visibility under reference lighting conditions, account being taken of lighting geometry and polarization of illuminance. DGF = Disability glare factors a measure of a task in a _ given lighting installation in comparison with its visibility under reference lighting. The measure takes account of the two effects of the ocular stray light produced by the pattern of luminance in the surround of the task: (a) the reduction in image contrast, and ~) the increase in RCS due to visual adaptation to the sum of the focused light and stray light.
121 TAF = Transient adaptation factor, a measure of task detail visibility in a given lighting installation in comparison with its visibility under reference lighting, account being taken of the transient adaptive effect that occurs when the eyes of the observer view luminances in the environment different from the task background luminance. 0.0923 = The value of the visibility threshold for the 4-minute disc task obtained by the reference observer at a level of task background luminance equal to 100 cd/m2. This model is based on extensive laboratory data and has been used in analyzing complex realistic visual performance, for example, proofreading, visual search, and numerical verification. While it has not yet been applied to VDT tasks, there is no obvious reason why it should not be applied to such tasks. It incorporates several factors, including transitional adaptation and disability glare effects. Although this model has been criticized (see, e.g., Yonemura, 1977; Padmos and Vos, 1980) and may have important limitations, it seems worthy of further validity tests, some of which could include VDT applications. Some aspects of an earlier version of the model incorporated a visual discomfort formula (Commission Internationale de l'Eclairage, 1975), and some recommended standards have been used by player and Barlier (1981) to evaluate 73 VDT work- stations. Luminance measurements were made of screens, keyboards, and documents. Contrasts of screen characters and their backgrounds for 40 of the VDTs are plotted with CIE visibility curves in Figure 5.3. Of the 40, 2 VDTs were above the comfort limit; 3 were approximately at threshold (VL = 1), which should make task performance very difficult or impossible; and 15 were in the range that would be expected to have a significant effect on the performance of some tasks (VL < 8~. In summary, VDTs differ from objects and surfaces in non-VDT workplaces because they usually have a highly specular, curved glass surface in a more vertical plane. Consequently, workplaces in which VDTs are used require lighting designs that differ from those required in non-VDT workplaces. Lighting and equipment arrangements that are appropriately designed for a particular task and working situation should prevent glare and most other prob- lems arising from lighting and reflections. Application of established knowledge and principles of design in the field of illuminating engineering can be expected to alleviate most of the difficulties related to lighting and reflections in VDT-related work.
122 100 at o J A UJ A: UJ IL 11 ~ 0.1 10 1 \\ \ \\ Comfort Limit \ No Visibility r Satisfactory Visibility - ~ ~ VL8 VL 4 l Poor Visibility _ VL 1 o.olL I I I I I I I I I 0.001 0.01 0.1 1 10 100 1,000 J Lum inance Characters Lum inancegackground Luminance Background BACKGROUND LUMINANCE (cd/m2) FIGURE 5.3 Illuminating Engineering Society (IES) visual performance curve: visibility of screens. SOURCE: Mayer and Barlier (1981~. REVIEW OF VDT STUDIES Field Surveys of VDT Workers Several field surveys have attempted to determine the opinions of VDT operators regarding problems caused by lighting and reflec- tions in the workplace. In some of these surveys, measurements of various aspects of the lighting conditions have also been made, and some investigators have attempted to relate those measurements to visual symptoms and complaints reported by VDT operators. Most of the surveys suffer from the kinds of limitations in method discussed in Chapter 2. Consequently, while the results of these surveys reveal that many VDT operators have complaints and symptoms related to workplace lighting conditions, they do not establish whether such complaints and symptoms are more frequent, more severe, or of a different nature than those that may be associated with non-YDT near-visual work. Our purpose in briefly reviewing several of these surveys is simply to provide an
123 overview of the kinds of studies that have been conducted and the kinds of problems that have been reported. There have been several published studies that used various types of questionnaires and interviews to determine the opinions of VDT operators about their work. In a study conducted by Ghiringelli (1980), 63 percent of operators reported that badly working equipment was a problem, 43 percent reported problems with reflections, and 43 percent also reported problems with "luminance." Using an unstructured interview technique, Grieco and coworkers (1980) interviewed an unspecified number of selected newspaper photocomposition VDT workers and reported that operators had problems with lighting, particularly with screen reflections. Dainoff (1980) and Dainoff and coworkers (1981) also used an unstructured interview technique in a study that covered 90 clerical workers, who reported that they worked with VTDs from 0-100 percent of the time (median 47 percent), and 31 data entry workers, who reported that they worked with VDTs 75 percent of the time. Complaints about lighting related specifically to the VDT were significantly higher in the data entry group; complaints about general workplace lighting, however, were made by the same percentage (37 percent) of workers in both groups. No attempt was made to relate complaints to specific aspects of the physical environment of the workplace. The designs of these studies do not permit an analysis of possible causal factors in reported complaints. Several published studies present data on the types and frequency of complaints, accompanied by measurements of certain aspects of the lighting and reflection conditions. Hultgren and Knave (1974) studied an insurance office that had 17 VDTs at various locations in a room. A questionnaire was used to deter- mine operator feelings about discomfort glare, eyestrain, and specific discomfort associated with reading the text on the screen. The luminance of the screen, illuminance of the work area, and angles of incident light and reflection were measured for 6 representative terminals. The operators reported problems with 13 of the 17 terminals in at least one of the three areas covered by the questionnaire. The luminance ratio of the screen to the brightest region in the immediate vicinity of the screen was a maximum of 1:500. In the worst case, reflected images had double or triple the luminance of an area of the screen containing a character. The character luminance was not measured directly. This study involved such a small sample that firm conclusions about the relationship between complaints and the VDT environment cannot be drawn. It did indicate, however, that operators in this one location had problems with glare.
124 Cakir and coworkers (1978) conducted a field study of 30 companies in which more than 1,000 VDT operators participated. Statements from operators about lighting problems were obtained by means of a questionnaire, and various aspects of the lighted environment were measured. The authors reported that the luminaires providing the general room lighting were a source of operator complaints. Luminaires of different construction elicited significant response differences to a question concerning direct glare. Bare fluorescent lamps were complained about more frequently as a source of direct glare than were luminaires constructed with some type of diffusing cover. Operators also rated bare fluorescent lamps as poorest with regard to the visibility of screen characters. More than 50 percent of the workers reported seeing reflections in their screens and, again, bare fluorescent lamps were singled out by the operators as producing more problems with screen reflections than luminaires with diffusing covers. At selected sites and on selected VDTs, Cakir and coworkers measured the luminances of ceilings, luminaires, display docu- ments, keyboards, and screen backgrounds. Ceiling luminances ranged between 15 and 35 cd/m2, while luminaires in the ceilings had luminances over 1,000 cd/m . Luminance ratios between screen background and display documents were found to be 1:6 if no antireflection filter was used over the VDT screen and 1:100 if a filter was used. The luminance ratio between the screen and the keyboard ranged from 2:1 to 1:70. Problems with beat frequencies, which might arise from the flicker of fluorescent lights and the refresh rate of the VDT screen, were not found. On the basis of the responses to a questionnaire, Elias and coworkers (1979, 1980) determined that 70 percent of operators who worked in an office that had windows almost all the way around had complaints about general lighting, while 45 percent of those who worked in an office with fewer windows had such com- plaints. The tasks differed in the two offices--in the windowed office the task was predominantly data entry; in the other office the task was interactiveand so did complaints about reflections on the screens: 45 percent of the data entry group and 65 percent of the interactive group had such complaints. The authors ascribe the higher frequency of complaints about screen reflections in the interactive group to the greater period of time these operators viewed the screen. The two groups also differed in complaints 1Elias and coworkers (1980) define the task as Data acquisitions; however, the description of the task (Elias et al.' 1979) corre- sponds more closely to what is referred to in this report as data entry (see Chapter 1).
125 about glare: 80 percent of the data entry operators and 52 percent of the interactive operators reported complaints related to glare that occurred, "sometimes" or "often" (four responses were available: "often," "sometimes," "rarely," "never"~. The authors ascribe the greater number of glare complaints in the data entry group to various factors, including range of contrast. Luminance levels were measured at a typical worksite in the data entry (windowed) office. The luminance ratio between the screen and the document was 1:8.5; between the screen and its periphery, the maximum ratio was 1:500, with daylight from a window providing general illumination on a sunny day; when illumination was provided only by a luminaire, the luminance ratio was 1:108. A higher percentage of data entry than of interactive operators reported discomfort glare. The authors ascribed this finding to the higher frequency of measured saccadic eye movements between objects of disparate luminances made by the data entry operators. Stewart (1980a) assessed environmental problems at 80 VDT workplaces. The most frequently occuring environmental problem was thermal (100 percent of the workplaces), but the next most frequently occurring problems were glare (71 percent of the workplaces) and reflections from windows and luminaires (83 percent of the workplaces). Measured illumination levels at the workplaces ranged from 100 to 2,500 lux. A study conducted under the auspices of the New Zealand Department of Health at selected VDT workplaces showed that screen reflections occurred on 42 percent of the VDTs surveyed (Coe et al., 1980)e No statistical relationship between room lighting intensity and presence of screen reflections was found. Of the workplaces sampled, approximately one-third exceeded a 3:1 ratio of luminances between hard copy and screen, and more than half exceeded that ratio between the immediate background of the screen and the screen itself. For those VDT tasks requiring hard copy, 54 percent of the operators viewing hard copy illumi- nated with less than 250 lux reported asthenopia (sore eyes and visual discomfort), 55 percent of the operators viewing hard copy illuminated between 250 and 500 lux reported asthenopia, and 34 percent of the operators whose hard copy was illuminated with more than 500 lux reported asthenopia. A comparison group was used in this study, but no data are reported on lighting complaints of the non-VDT comparison group. Field Surveys Comparing VDT and Non-VDT Work The studies reviewed above indicate that lighting and reflections do cause problems for VDT workers; however, these problems may
126 not be unique to VDT use, but may be present in general office work. Without evaluating non-VDT work, no comparisons can be made regarding such complaints in VDT and non-VDT tasks. Two studies have attempted such comparisons. Laubli and coworkers (1980) conducted a field study in which illumination and luminance levels were measured at workstations used in four different types of office tasks: (1) data entry VDT work; (2) interactive VDT work; (3) traditional clerical work; and (43 typing. At 90 percent of the VDT workstations, the illumi- nation levels of source documents were between 100 and 1,900 lux; at workstations used for traditional clerical work or typing, the levels were between 100 and 3,200 lux. The luminance ratios between source document and screen background ranged from 7:1 to 87:1 at VDT workstations. The frequency of ocular and visual symptoms pains, burning, fatigue, shooting pain, red eyes, head- aches, blurring of near and far vision, flicker vision, and double images (all referred to as eye impairments in the study)-- increased as luminance ratios increased. The incidence of symptoms was highest in the interactive VDT operators; it was lower in data entry VDT operators and in typists (and approxi- mately the same for these two groups), and considerably lower for workers performing traditional clerical work. Although images reflected by the VDT screen had lower luminances than screen characters, the measured intensity of the reflections was correlated with reported annoyance; it was not, however, correlated with frequency of ocular or visual symptoms. Stammerjohn and coworkers (1981) conducted a study at four newspapers and one insurance company. At selected workstations, they measured illuminances and luminances in operators' general visual field, including the luminance of the screen background, but not that of the display characters. Subjective operator ratings on a five-point scale from "no bother or problem" to "constantly bothersome" were obtained on screen brightness, character brightness, readability, screen angle, keyboard angle, screen height, keyboard height, distance to the screen, distance to the keyboard, screen glare, keyboard glare, noise from the VDT, and screen flicker. The majority of workstations had illuminances between 500 and 700 lux, with a low of 300 and a high of 1,200 lux. The range of luminance ratios in the immediate visual field of the operators was from 1:2 to 1:60. Potential glare sources (windows, lumi- naires) were reported at 46 of the 5) workstations. Luminances of these sources were near 2,100 cd/m . Reflected glare from windows or overhead lights was present on most of the screens surveyed at one site. [uminances of the reflected images had maxima of 3~60 cd/m . Although character luminance was not measured, the investigators had difficulty in reading the screen
127 text from VDTs that had particularly high reflected glare leve (17 percent of those examined). Is Several aspects of the VDTs were reported as bothersome by operators: screen glare (85 percent), character brightness (70 percent), readability (69 percent), flicker (68 percent), and screen brightness (62 percent). A slight majority were satisfied with workstation and background illumination. However, 80 percent of the VDT operators reported trouble with glare from the work- station lighting. More than 60 percent of the non-VDT operators also reported problems with glare from workstation lighting. Because of employee anonymity involving the use of the questionnaires, specific complaints cannot be related to specific VDTs and therefore to specific design features. This study did, however, determine that a significant relationship exists between complaints regarding visual function and employee rating of workplace design parameters, including glare, screen angle, noise from the VDT, and screen flicker. Laboratory Studies In a laboratory investigation, Radl (1980) demonstrated the importance to operator comfort and to one measure of perfor- mance of graded luminance from the screen to its surround. The subjects had to transcribe letters from the screen to a paper sheet. They viewed a VDT screen with a back2ground luminance of 18 cd/m2 and symbol luminance of 120 cd/m with a surround of 4,200 cd/m2 (i.e., a glare source) encompassing about 75 degrees of the subjects' visual field. The initial experiments used a black frame of varying dimensions around the VDT screen. The results are summarized in Table 5.2. Radl also compared the effects of positive and negative contrast on performance of the same task and on rating of visual comfort, using 24 subjects. The subjects rated negative contrast (dark characters on a light background) as more comfortable (general illumination of 500 lux; screen surround not described), and performance was also greater with this pre- sentation. Both findings were reported as statistically significant. Bauer and Cavonius (1980) examined the differential effects of positive and negative contrast on performance of a letter identi- fication task. The display~resentations were (1) low-luminance, positive contrast (10 cd/m background), (2) high-luminance positive contrast (80 cd/m2 background), and (3) negative contrast (80 cd/m2 background). The error rate was lowest for the negative-contrast display and highest for the hig~luminance positive-contrast display.
128 TABLE 5.2 Graded Luminance and Operator Performance and Comfort Width of Frame Rated Visual Comfort Performance(%)a (Worst, 0; Best, 7) . No frame 42 0.32 7° 67 1.45 1 1° 69 1.80 16° 84 2.05 21° 63 3.07 11°b 85 3.15 a Relative to performance tested in the absence of glare. b Frame continuously shaded from black at the screen to white at the outer edge. In a second experiment, Bauer and Cavonius (1980) compared the speed and accuracy of subjects using positive- or negative- contrast displays in detecting discrepancies between a VDT presentation and a typewritten presentation. The two conditions compared were positive Contrast display (symbol luminance 40~5 cd/m2, screen background luminance < 10 cd/m2, ambient illumination 270 lux) and negative-contrast display (symbol luminance set as low as possible, screen background luminance 50-70 cd/m2, ambient illumination 550 lux). The negative- contrast display produced the fewest errors and fastest performance time, and 18 of the 19 subjects preferred the negative contrast; the one subject who preferred positive contrast actually performed better on the negative-contrast display. As noted in Chapter 4, both the Bauer and Cavonius (1980) study and the Radl (1980) study should be interpreted cautiously because changes in contrast polarity were combined with changes in ambient illumination and absolute contrast magnitude. In summary, the field surveys and a few limited laboratory studies of VDT-related work indicate that workers who use VDTs have problems caused by lighting and reflections. However, we found no studies that compared equivalent, appropriately illuminated and arranged VDT- and non-VDT tasks and working situations, and so we cannot draw conclusions about the relative number, types, and severity of complaints and problems related to lighting and reflections in VDT and non-VDT work.
