Angles Of View
As manufacturers, designers, and end users have become increasingly sophisticated about the performance characteristics of their display systems, the correct interpretation of Gain and the quantification of Resolution now enjoy wide acceptance and consensus. There is, however, a third principal attribute whose standard has not, regrettably, been widely promulgated. The discussion which follows seeks to take a closer look at what might be revealed if we were to bring
Contrast - From Dark to LightDefining what we mean by contrast is easy, perhaps too easy. Contrast is the disparity between the dark and the bright portions of an image.
Calculating a contrast ratio is also perfectly straightforward. As we have noted elsewhere (Vol I,No.5), the formula has been
Max - Min
where Max and Min are the values we obtain for the brightest bright and the darkest dark.
Measuring those values or, more precisely, deciding where and how to measure them is not so easy and their definitions resultantly are sometimes ambiguous. Yet it is certain that a large contrast ratio is unambiguously better than a small one. As viewers, we can often detect changes in contrast that are as small as two or three percent. Perceptible changes in brightness, on the other hand, often entail variances greater than fifty percent.
To make all of this a little more concrete, let's enumerate some specimen contrast ratios. If you're reading these words, the contrast ratio of the page before you is about 80:1. If you're looking at them on your monitor, the ratio is closer to 50:1. If you go to the movies and watch a good, clean print, the ratio (given the right scene) might be 500:1.
Since each of these examples puts a value to what you will see, let's call this type of contrast Perceived Contrast. Another type might be what we measure directly on the surface of the projection screen. Let's call this Screen Contrast.
Now let's take the next step and look at our projection device. Is it possible that it may in turn exhibit a different ratio from our other two pairs of measurements? Very likely it will. Hence, Projector Contrast should be distinguished from the first two.
Finally, we are entitled to examine the contrast ratio of the image itself. We'll call this Image Contrast. If it is jet black print on a pure white page, surely its contrast ratio will be different from light gray characters written, shall we say, across a pale blue screen. Whatever this beginning ratio may be, it will be extremely difficult (but not, as we shall see, impossible) to enlarge through subsequent transitions.
Thus far we have kept things simple because our example has been text based. In terms of contrast, if there are only two candidate values to measure, it can't be too hard to decide which is which. (Furthermore, we can safely assume that the darkness of any character we measure is equivalent to the darkness of any other and equally that the field behind the letter W is no whiter than the field behind the letter M.)
Few of the projected images we routinely view are ever characterized by this monochromatic simplicity. Their grayscale gradations, the number of available colors they contain, instead of being a mere 2, will jump to something between 16 and 16 million. Still, in terms of brightness at least, that enormous palette must, nevertheless, contain just one blackest black and one whitest white. All we have to do is find them and we will have our Max and our Min.
So let's conceive a picture with lots of image contrast scattered through lots of image colors and send it along to a projection device for displaying. What happens to it next depends greatly on what sort of projector it is.
Looking at the descriptive literature which accompanies projectors these days, one is certain to find a luminous output spec (ANSI lumens only, please), one or more Resolution specs (VGA, SVGA, etc.) which are easy enough to interpret, and sometimes, but not always, an additional spec for Contrast. Here are some quoted examples: "100:1," "Greater than 100:1,""200:1," "250:1," "600:1," and "2000:1." It is as hard to find a projector admitting to a contrast ratio less than 100:1 as it is dumbfounding to find one professing 2000:1.
Because there is such a wide range of stated contrast ratios, we are entitled to wonder where, exactly, the various Max and Mins come from. Although we cannot answer that question with much surety here, we can state how a contrast ratio should be measured.
In 1992 a panel of experts convened under the auspices of the American National Standards Association to develop a document that is called Data Projection Equipment and Large Screen Data Displays -- Test Methods and Performance Characteristics (ANSI IT7.215-1992). Among the results of this effort was the establishment of the ANSI lumen standard for brightness. Less well known is the standard established for determining contrast ratios:
The contrast ratio shall be determined from illuminance values obtained from a black-and-white "chessboard" pattern consisting of sixteen equal rectangles.... The white rectangles shall be at full specified light output, as previously measured, with all controls at the same settings. Illuminance measurements in lux...shall be made at the center of each of the bright (white) rectangles and the dark (black) rectangles. The average illuminance value of the bright rectangles shall be divided by the average illuminance value of the dark rectangles. The contrast ratio is then expressed as (ratio):1 (e.g., 15 lux bright rectangles and 0.10 lux dark rectangles equal a 150:1 contrast ratio).
Notice that the projector is obliged by this procedure to project its brightest white and its darkest black simultaneously. Displaying one and then the other sequentially undoubtedly produces a more widely divergent pair of numbers, but their ratio is not descriptive of what a viewer actually would see.
Now that we know how to quantify projector contrast, let's turn our attention to screen surfaces and see how their design and selection can affect contrast.
Because front projection screens are obliged to reflect all sources of light incident to their surfaces, their perceived contrast is highly dependent on the presence of ambient light. The difficulty with non-projected light sources reaching a screen, of course, is not their effect on the Max portion of the image but the often disastrous effect on the Min. Generally speaking, front projection screens with higher gains will do the best job of preserving image contrast.
While it is true that front projection screens cannot increase the contrast of a display, rear projection screens can. They do this by a variety of techniques which exploit the opposite directions from which projected light and ambient light pass through them. Light from the former is transmitted toward the audience; light from the latter is transmitted into the projection booth.
Some rear projection screens, such as Da-Lite's Polacoat line, include spectrally corrective colorants which, by darkening the appearance of their front surfaces, render them less reflective of incident ambient light sources. (Again, see Vol I, No.5.)
For the maximal enhancement of Contrast, however, no screen is more effective than Da-Lite's new Black Stripe Screen from DNP. The design of this screen is elaborate and its structure intricate.
The drawing above well illustrates this complexity. Notice that the frontmost surface of the screen is comprised of a matrix of thin, opaque black stripes. Light incident to these stripes from ambient sources is not reflected (it is absorbed). Since the surface area of the stripes is nearly 40% of the total face of the screen, this resistence to ambient light is pronounced. Even in the presence of no ambient light, the black levels visible from this screen will be much darker than those obtainable from alternate surfaces.
The most intriguing aspect of the screen, of course, is what happens to the projected light which the projector aims at the back of the stripes. Because all of the projected rays are collimated by the rear element Fresnel, they all enter the back, lenticulated surface of the front element at effectively the same, 0ş, angle. This enables the curvature of the lenticulations to refract the incident rays so that they are neatly redirected out of the screen between the black stripes.
Complex as this arrangement is, its effect on image contrast is extremely beneficial. Even though this screen has an on-axis gain of 4 (an indication of its Max), its real achievement is the reduction of its Min. In fact, manipulating contrast is always an art of darkness.