Angles Of View

When we look at data displayed for us on a projection screen, it is only in the rarest case that all we see is just the information. Unfortunately, projected imagery contains the potential for numerous visual artifacts unrelated to content that can compete for our attention. And, although the eye:brain interface is extremely good at filtering out extraneous input, it is even better at detecting aberrations and discontinuities. In order to minimize the effect of these distractions, it is important that we understand the differences between them. This article will take a closer look at some attributes of visual displays whose causes, by virtue of being often confused, sometimes make it difficult to distinguish between

Medium and Message - the Screen and the Image

All of us are well trained in the smooth assimilation of visual data if their source is the conventional printed page. The eyes skim along in practiced skips, line after line, neither disturbed nor even interrupted should some character be, for whatever reason, misprint ed. Yet, if we take a sample word and, instead of misprinting it, misspekl it, the eye suddenly stumbles and our attention must return to reexamine and verify the error. Such a defect, of course, is a blemish in the data. It is not a fault in the delivery system. What is more important, telling the two apart is not difficult.

The overall appearance, for instance, of nearly all printed pages is wonderfully uniform. The printing press impresses the letters and words at the corners of our book with exactly as much ink, sharpness, and density as it imparts to the characters in the center. In addition, the white of the paper itself is just as white at its edges as it is in its middle. These are consistencies we are accustomed to taking for granted in our reading materials. Not one of them, however, can be counted on if the text is to be projected.

A projector is not a printing press. A screen is not a blank piece of paper. The misapprehension of these two points regularly gives rise to bafflement and confusion. Each seems perfectly obvious. Both are actually subtle.

Let us first consider the projector and ask just what it is that it projects. The correct answer is the page, not the text. The text is a part of the page, but only a part. The whole page, pixel by pixel, is what the projector is obliged to display. That some pixels are instructed to be dark while most are told to be white is arbitrary and, in this sense, irrelevant.

Now let us consider the screen and ask what it displays. The answer again is the page, not the text. To see this, we need merely imagine that our projector is jostled or in some other fashion perturbed such that an edge of the image slides off the screen and onto, let us say, the adjacent wall. Obviously it will not only be the text that has moved but the "page" beneath it as well.

As emissive devices, projectors don't deposit jet black letters onto a snow white screen. In fact, while they try hard to project a snow white page, they simultaneously try very hard not to project any light into the text areas which, by virtue of their being projected, can never be jet black.

Seen in this light, we may rightly ask, well, if a screen is not the page beneath the text, just what is it? The answer is that a screen is not the page as long as what we mean by page is the background. But if what we mean by page is the surface on which is displayed, for example, a replica of a favorite painting, then, yes, that ' s exactly what a screen is: the high quality paper on which that detailed reproduction has been printed. Thus, although screens are indeed the paper, we need strongly to emphasize that they are not the page.

With these distinctions in mind, we can now turn our attention to the single greatest defect commonly attributed to projection screens: the Hot Spot. This unpleasantry barely needs defining. It's that nasty and distracting bright blob that seems spitefully to follow us no matter where we position ourselves in front of any display unfortunate enough to exhibit it. Confronted by such an eyesore, who in this industry has not heard a colleague point out, "Golly, that screen is hotspotting terribly, isn't it?"

And what really will they be looking at that causes this complaint? Clearly, they will be seeing an image which appears excessively brighter at its center than it is at its corners. Unmistakably, there is a hot spot.

Furthermore, if we leave everything else exactly in place and change out only the surface of the screen we are looking at, the hot spot can be diminished greatly. This, of course, convincingly demonstrates that hot spots are in fact a regrettable property of projection screens. Or does it?

Let us agree not to change anything about projector or image and let us agree to leave in place the very screen surface which produces the most fearsome and unpleasant hot spot we can find. But let us imagine only that the screen has magically been doubled in its width. This suddenly greater latitude will enable us to drag the screen to its left or to its right with no risk that the projected image will slip off its surface. Naturally we will expect to see the hot spot move right along with the screen. Or won't we?

Of course the hot spot doesn't move; it is a property of the image, not of the screen. No front projection screen is manufactured such that its center is intentionally differentiable from its edges. Most screen manufacturers make front projection surfaces with original dimensions that are something like 8 by 3,000 feet; certainly Da-Lite does. Then, from this very long roll, are cut hundreds of screens with hundreds of centers, in dozens of sizes, in numbers of aspect ratios, but without even one, single, solitary hot spot. At least that's the way it's done at Da-Lite.

" But wait! " you may say; "What about those screen surfaces we were changing out just a moment ago? Didn't one of those hotspot more than another? Exactly what's going on here?"

Good question. To answer it, let's put a projector behind a screen surface which is initially completely transparent and clear. (For simplicity, this example assumes a rear projection set up.) When we look at such a screen, all we will see through it is the blinding, white light made by the projector's lamp and directed by its lens. This, surely, is the mother of all hot spots.

When we start to add screen diffusion material to this substrate and as, therefore, we render it less and less transparent, the shape of the hot spot begins to enlarge and its core brightness begins to diminish. As we add yet more diffuser, vague outlines of the image contained within the beam begin to appear, although we can still see the light source beneath them.

Add still more diffuser, however, and it is the content of the image that will become more prominent to our eyes. When we reach the end of our diffusion process, the image will be fully resolvable and the hot spot (the light source) will no longer be resolvable at all. That, at least, would be ideal.

If we stop just short of putting on this optimal amount of diffuser, we'll still be able to see traces of the light source even though we can fully discern the contents of the image it's projecting. When we do that, we produce a diffusion screen that has gain. If we put on even less diffusion, we will produce a screen that's got even more gain. In this way, by allowing more of the light from the light source to be undisturbed by the diffusion at its surface, screens can be made varyingly to reveal or disguise a hot spot. But there is no way they can create one.

The really interesting question, then, is not why some screens sometimes hotspot. It's why most screens most times don't. To discover the answer, let's randomly choose a half dozen screens of the same size and of the same gain and place them before any six disparate projectors.

If the intensity of the light emanating from these projectors' lenses is carefully measured across the diameter of their lenses, it is extremely unlikely that the reading we take at the center of each lens will equal the readings we take from its edges. Some projector types will do a lot better in this regard than others, but all will fail to be perfectly uniform.

When we come, then, to point each of these projectors at our six screens (all of which, by definition, are uniform), it should in no way surprise us that the displays which exhibit the most egregious hot spots are the ones illuminated by the least uniform light sources. It is important to notice that we have said nothing whatever in this discussion about brightness. That's because, paradoxical as it may seem, varying the luminance of each (or any) of these projectors will neither erase nor exacerbate its hot spot.

The hot spot phenomenon, then, is correctly understood as being a function of uniformity and of uniformity alone. Moreover, when we see a hot spot, our eyes are actually disturbed more by the dimness at the corners of the display (the areas outside the hot spot) then they are by the brightness at the center. If there were some way to brighten those corners up such that they would become equal in illuminance to the center, the perceived hot spot would instantly vanish, even though that area's brightness would in no way be changed.

If we now leave all six projectors in place and swap out all six screens for new ones with a higher gain, the corners of all six will appear even dimmer relative to their centers and those that exhibited hot spots initially will now appear even worse. If we introduce lower gain screens, of course, the tendency reverses.

Projected displays are tricky. They include many more components than printed ones. Certainly their appeal is overwhelming and their power undeniable. Yet understanding them is important because otherwise, looking at them, it's sometimes hard to know precisely what we're seeing.