Angles Of View
The proliferation of projection devices which are being pointed at or through Da-Lite screens these days continues to be remarkable. Because, however, in virtually every case each projector type is hard pressed not to add something of its own special signature to the image it casts, sometime viewers can become confused by what they see on a screen that is extra to that image. Sometimes these extraneous phenomena are visually quite bothersome, sometimes they are barely noticeable. This article will seek to help distinguish among them by
Telling the Difference - Attributes and ArtifactsThe resolution of the real world, everybody knows, is effectively infinite. Especially by using instruments like telescopes and microscopes, there is almost no limit to the visual detail we can extract from it. When we make a visual representation of some aspect of the real, world, however, our capacity to duplicate the full detail of the original becomes limited and circumscribed. Yes, sometimes those limits cannot be discerned by our unaided eyes, but they are always there. Much more typically, and particularly when it comes to looking at electronic displays, our eyes are more than capable of perceiving that the image they are looking at includes extraneous elements beyond itself.
The origin of these non-image phenomena is sometimes not easy to pinpoint. Does they result from a defect in the screen material? Is there something not right about the way the projector has been set up? Could they be coming from the computer? Or are they something contained in the software making up the image?
Each of these components possesses some number of attributes, a term which we will here define as intrinsic characteristics of its design and function. When attributes intrude into the viewing experience, however, they become artifacts, a term which we will here define as visible phenomena that are extrinsic to image quality.
Let's start first by looking at screen surfaces. Whenever a front projection screen is illuminated by a projector, you should not be able to see any discontinuities across its surface. If you do see something suspicious, the easiest way to decide whether it's in the screen or not is to jiggle or move the screen surface slightly. If the aberration moves too, you've got a problem in the screen. If it doesn't, the problem is sited somewhere else.
In the case of rear projection screens, the same test can't be performed because rigid screens can't be jiggled so easily. What you can do easily, however, is go around to the back of the screen and place a piece of white paper over the area where you saw the aberration from the front. If you see the aberration on the paper, it's within the image, if you don't, it's within the screen.
In both screen cases, two additional standards must be mentioned. The first is that no screen should be judged defective by the way it appears under ambient light alone. Screens are designed to display light from a projector (which is to say, light that is emanating from the equivalent of a point source). Artifacts, discontinuities, or occlusions that do not appear under projection but which may otherwise be visible under ambient alone do not count and deserve to be ignored. Obviously, the reverse is not true.
The second standard has to do with the viewing distance from which screen artifacts, even under projection, may be discerned. If any such aberration can be seen only from a distance that is less than one diagonal, we would suggest that it may safely be ignored. If, however, the surface blemish is visible from a distance greater than one screen diagonal, then a responsible screen company should be willing to help you repair or replace your surface.
Now, what about projectors? Can we discern artifacts or attributes according to the type of projector we choose? Indeed we can, especially if the imagery we are considering is stationary and not kinetic. Moving imagery (entertainment, etc.) will not be considered in the descriptions which follow. Furthermore, only projectors capable of displaying information generated directly by a computer will be considered.
An artifact which everybody notices right away whenever it is present is aliasing. Aliasing is a phenomenon which occurs because the image being displayed is not an analog (which is to say, continuous) picture. Instead all computer generated imagery have been sampled some finite number of times at particular finite spatial and temporal frequencies. The coordinates of the spatial sampling of a display are, of course, its resolution and will thereby disclose into exactly how many pixels the image has been broken. Since these pixels are arranged in a neat, orderly X/Y grid, they seldom have any problem displaying realistic lines that are parallel to either of those axes. Aliasing enters the picture
when some of its elements include curved or diagonal lines. Unless you have a very large number of pixels (each of which is, therefore, very small) or unless you're looking from a large viewing distance (which is the geometric equivalent), its pretty easy to see that the roof lines are not smooth and that the circumferences of the trees are not truly circular. That's aliasing.
CRT projectors draw their pictures by scanning them in with sequential horizontal lines. Because the number of lines (top to bottom) doesn't change as the image size gets bigger, it's often easy to detect these rasters or, really, the increasingly large spaces between them. When those spaces become sufficiently noticeable, the CRT's raster attribute has become an artifact.
Another attribute of CRT projectors is that they almost always include three, spatially separated light sources - the R, G, and B guns which made them famous. If the size of the projected image is not large, the absolute distance between the centers of those guns (e.g. 6, 7, or 9 inches) can be trigonometrically significant relative to the device's requisite throw distance. When this happens the colors in the perceived image will tend to shift toward the red when looked from a points opposing the B gun and toward blue when seen from opposite the R. This 3-gun artifact turns back into an attribute as the throw distance (image size) is increased.
Then there are all the matrix display projectors which rely on some sort of chip or chips to create their imagery. All of these are alike in that whatever chip they contain will have one and only one "native" resolution. Thus, for example, to know that the matrix of a particular projector is made up of 1024 x 768 "elements" is to know that the best that projector can ever do is to divide its imagery up into a number of pixels equaling the product of those two numbers. That's one of such a projector's attributes. An artifact often appears not from the limited number of pixels but from the spaces between them.
Whether the chips in matrix projectors are transmissive or reflective there must always be some degree of separation between one pixel and another. Using LCD panels as an example, if each liquid crystal pixel is a little window whose shade may be positioned at any instant anywhere between full open or full shut (and through which, therefore, varying amounts of light are permitted to pass), the mechanism to "motor" that shade is contained in the brickwork of the wall surrounding it. If we consider the entire wall, we can determine exactly what percentage of it is made up of windows (and, hence, is "not brick"). This percentage is called the chips' aperture ratio. If it's high (like 90%) it won't be easy to distinguish one pixel from another and we merely describing an attribute. If its low (like 40%) then the majority of the image's area will be made up of the visible "chicken wire" which prominently and discernibly outlines each and every pixel. That's an artifact.
In the case of dmd projectors, the "motors" which cause the mirrors to flicker back and forth are not adjacent to but below them. Even so there must still be space at all four sides of each mirror so that as it is rotated out of flat none of its edges touches those of its neighbors. Equally, reflective LCD projectors are able to produce very high aspect ratios because they, too, have their "motors" behind the pixels.
There are, of course, numerous other display attributes which can be artifacts and the cases mentioned above are meant merely to be illustrative. The point most worth making regarding artifacts, however, is that they tend to be required viewing. When they show up, we can't not look at them. Artifacts are not helpful; they obstruct our ability efficiently to assimilate projected information. Unfortunately, they often must be tolerated; but they should never be welcomed.