VDT design factors

	Overview   |   AC component   |   Calculating the AC component   |   Flicker and VDT design   |   Image polarity   |   Display brightness/contrast measurements   |   Resolution   |   Color

Image polarity
Opinion is divided on the subject of image polarity, that is, whether there are significant visual advantages in displays of positive polarity (dark symbols on a light background) over displays of negative polarity (light symbols on a dark background.) Concerns include acuity, re-adaptation demands when looking between the display and a source document, stress of the pupillary response mechanism and the luminance balance of the visual field. When these polarities are tested, however, there is little human factors advantage of one polarity over the other for most applications and users (Cushman, 1983; Hedman and Briem, 1983; Taylor, McVey and Emmons 1983). There is some advantage for negative polarity, however, when images are under difficult viewing conditions or for users with marginal vision.

• With positive polarity, specular reflections are less perceptible, edges appear sharper, and luminance balance is easier to obtain.

• With negative polarity, flicker is less perceptable, legibility is superior for individuals with anomalous low acuity vision, and characters may be perceived as larger than they are."

Some proponents favor dark characters on a light background because there is a higher average light density. Under certain circumstances, there is indeed a relationship between light density and acuity. For that relationship to exist, however, the image would have to be very close to threshold, or just barely visible, and the light density would have to be less than about 35 cd/m² (Shurtleff, 1980). Otherwise, contrast is by far the more important factor.

For example, consider two displays, one of each polarity:

If, in this example, the contrast ratio is the same (specifically, 10:1) on both displays, the task of distinguishing a signal from its background will be the same. The light density, however, is considerably different, 91 cd/m² in the one instance and only 19 cd/m² in the other.

Now consider another example, one in which light density is equal, maintained at 19 cd/m².

As can be seen in this case, the equal average luminances result in a considerably different contrast ratio. The bright character display has five times greater contrast, which makes for a profound difference in the visual task.

In short, light density is not a good predictor of visual performance in supra-threshold conditions.

The ISO 9241, part 3, position on luminance is:

"From a perspective of the limits of visual acuity, the display shall be capable of a display luminance of at least 35 cd/m².......... If liminance coding is used, 35 cd/m² specifies the minimum for the lower luminance."

It should be noted that luminance coding is the situation where the display user must group data on the screen on the basis of luminance alone. Applications which group data by color and then map colors to luminance levels on monochromatic screens will comply only when the minimum luminance exceeds 1.5^(n-1) x 35 cd/m², where n is the number of code points on a monochromatic display.

The task of the retina is to detect the edges between luminances of two different levels. The edge is essentially the same regardless of the polarity. A higher cerebral process, and not the retina, distinguishes between figure and background (Frisby, 1980).

Adaptation was once thought to be a function of the bleaching of the photo pigments of the retina. If that were true, then the average amount of light would influence adaptation. Adaptation at light levels of concern here, however, is controlled by a higher order neural process. That process causes a spread of adaptation which, while biased toward the higher luminance, approaches uniformity. An area of the retina that is not stimulated will show the same change in adaptation level as the stimulated area (Barlow and Andrews, 1973). There should be no particular retinal adaptation problems when looking between a source document and a display if the higher luminances of each are not significantly different. For example, consider reading this printed page. Each time the point of fixation changes, the image on the retina changes. Some receptors that had been stimulated by the white area of the paper would then have the dark image of the symbol over them and vice versa. This does not cause any retinal adaptation problems.

The pupillary response might be stressed if there were frequent refixations between a negative image display and a source document. The size of the pupil, however, is a function of the total amount of light striking the retina, and at the light levels of concern here, average pupil size changes minimally. A study of Rupp and Clauer (1982) was conducted to observe the immediate response of the pupil when changing the point of fixation between a dark background display and a source document. Another study (Zwahlen, 1983) was conducted to determine average pupil size when viewing the display, the keyboard and the document. These studies did not show any more pupil activity than the normal pupil activity (hippus) when viewing a steady luminance. The pupil is in constant motion except during deep sleep; hence, it is unlikely that the pupillary response is a source of stress when frequently refixating between a negative image display and a source document.

There is some question as to whether acuity is limited by the background luminance level. It is true for images of dark characters on a light background, but it is not true for images of light characters on a dark background (Berger, 1944).

Back to top