Angles Of View

If the dominant topic discussed in this fourth series of articles has been Information, the essential lemma of their argument remains the importance of distinguishing between visual tasks. Projected displays which are intended, variously, to be looked at, read, or inspected are no longer interchangeable. The geometries, specification, and design are now defined not just by their source, but by their function. From, then, the vantage of a screen manufacturer, what can be done to enable and enhance

Reading by Da-Lite — Ne Plus Ultra

At the recent INFOCOMM show more than 20,000 visitors flocked to Dallas and spent three very full days looking at our industry’s myriad new wares. Many were kind enough to stop by the Da-Lite booth and cheerfully inquire, "What’s new?"

Answers to that most welcome question entailed a round-the-booth tour, with many stops along the way. Most (but not all) of those stops involved an opportunity to look at some new (or newly configurable) projection screen. One observer to these proceedings was struck by the number of people who, while examining a screen, would position themselves at truly enormous angles to the projection axis and then, and only then proceed with their appraisal.

Is it really reasonable, we want now to ask, that a screen and its image, should be judged by a viewer down on her knees, with her cheek pressed against the wall supporting it? Even if a screen could produce such extreme viewing angles (and some actually can, ask your Da-Lite Sales Consultant), why ever would anybody want to look at it from such an uncomfortably oblique angle?

Yet, time after time, seasoned A/V professionals thought themselves diligent by checking out each and every screen under their purview from viewing angles exactly as unreasonable as have been described. Why? What could be going on here that makes these intelligent and highly educated people insist on putting humble screen products to such a torture test?

The most articulate retort is likely to be founded on necessity. The designer asserts that she has no choice in the configuration of her audience. She can specify the kind of projector and (within limits) its potency. She can specify the kind of screen and (within limits) its size. But she cannot specify the configuration and size of the audience. Those, she will tell us, are established by the end user who, although sometimes susceptible to her advice, rarely requires her consent.

Thus, the reasoning continues, screens which enable extremely wide viewing angles are to be admired above screens which do not. Moreover, we happily repeat, there’s the convenient fact that some screens these days really do have half-angles of 60º or more. So, since such surfaces are now ubiquitously available, what’s the problem? Where, we might even ask, is the beef?

The suggestion we would like to make here is that everybody in the dialogue so far is forgetting (overlooking?) the change in the visual task. We’re no longer allowing that audience of end users to be able just to see the screen from some arbitrarily large viewing angle. We’re asking them to read it.

Can you recognize or see imagery on a screen from a 70º viewing angle? Yep; often even if the screen isn’t some high-tech rear projection lensing array. (Test a matte white front screen this way, for instance). Can you read alphanumeric data or discern high resolution graphics from a 70º viewing angle? Nope; can’t reliably be done.

Literature and research on maximum viewing angles for legibility abound. The consensus is overwhelming and unmistakable: 45º is the maximum viewing angle that can be tolerated by audiences required to read a display. Maximum.

Joy Ebben, Ph.D. and lots and lots of other experts all agree. Beyond 45º the eye is simply not a reliable data acquisition device and all kinds of characters and symbols will assuredly be mistaken for others.



Figures 1&2

In entertainment venues this distinction is largely meaningless. In commercial settings it is absolutely critical. Regardless of some screen’s ability to distribute light over large angles, any screen person who advocates reading from larger dispersion angles than 45º is either irresponsible or sadistic. Viewing angles are not reading angles.

After belaboring that point, what else might be said about screens whose manufacturers now understand that their surfaces will need more and more to be read rather than looked at? Does that single, enormous paradigm shift have impact on the design of screens themselves? Da-Lite Screen Company, Inc. has determined that the answer is an emphatic yes.

If we know that a screen needs to be read, there are a couple of immediate conclusions to be drawn. They are that, first, the screen should be a rear projection device and, second, that it should be as uniform as is optically possible.

The advantages of rear projection over front for non-entertainment displays have been throughly detailed in numerous prior articles in this series. They include more efficient projection geometry, greater tolerance of ambient light, and higher image contrast, to name just three.

The advantage of maximum uniformity, of course, results from the necessity of being able to read the whole "page" of each screenful of data. That "page" has four corners which will need to be as acutely discerned as its center. From any given reading angle, however, if one or more of those corners is perceived to be less than half as bright as any other portion of the text to be read, its data is likely to be literally illegible.

Keeping the edges and corners of an image from being less than 50% of the brightness visible from its center is no mean feat. Even if the output of the projector itself is extremely uniform, there remains a geometry problem that will never go away.

This, of course, is the necessity that all projectors must emit light rays through their lensing system which must diverge. It is this divergence which allows for the size of a projector’s image to get larger as its throw distance is increased. If the shape of a projector’s beam were, say, rectangular instead of conical all of the light rays comprising it would be parallel, but the image would never be larger than the surface of the lens transmitting it. Although such a projector might be interesting, it would doubtfully be popular.

The drawings above illustrate the paths of three principal light rays projected through a rear projection screen. To reach the figure in Figure 1, light from the right-hand ray must be bent through an angle of 50º. Conversely, light from the centermost ray might need to be bent only five or six degrees to reach his eyes. Because the second job is so much easier than the first, any viewer will always receive more light from the screen in front of him than from the opposite side. People who notice this phenomenon call it a hot spot. People who have to read through it call it a pain.

One constructive approach to minimizing this problem is to utilize Fresnel/Lenticular screens. Classically, this solution has been extremely effective. However, remembering what it takes to read a display, the periodic structure which all lenticulated arrays impose into the image plane is no longer an attribute which can safely be ignored. Anything which can degrade resolution should always be avoided if reading is the visual task.

Thinking hard about all this, Da-Lite engineers asked themselves the following question, "What would happen if we decoupled a Fresnel from its lenticulations and diffused just that?" The result is Da-Lite’s reading screen: the Polacoat® Ultra.

Put a Fresnel behind the screen in Figure 1 and you get Figure 2. Because a Fresnel lens will collimate light rays from the projector such that they all travel parallel to the centermost light ray, all bend angles become significantly smaller. By reducing 50º to 30º, Figure 2 puts a whopping 40% more light from the corners into the viewer’s eyes than can Figure 1. This huge improvement in overall uniformity cannot be overemphasized. And, since the grooves of the Fresnel (like lenticulations, also a periodic structure, after all) are behind the image plane rather than in it, the effect on resolution is imperceptible.

Thus, if you want to make data you wish to project maximally legible, there is no rear projection screen superior to the Ultra. It is, word-for-word, ne plus.