Angles Of View
Joe Kane is the founder and President of Joe Kane Productions, Inc. headquartered in North Hollywood, CA. Mr. Kane consults, lectures, and writes widely on defining, building, and maintaining high quality display devices. He can be reached at firstname.lastname@example.org and is interviewed here on the subject of
Display Screens - Revealing the StandardsDa-Lite: In a general way, what do we mean when we speak about a "display device"?
Kane: A display device is something that converts electronic signals carrying picture information into something that is intelligible to us as human beings. From a communications systems point of view, a display device has to be a fixed, well known pallet. Otherwise it's a useless piece of arbitrary hardware.
The purpose of a display device is to present visual information. In order to convey the intended information to a large audience, the medium has to be well defined. That fact was clearly recognized when our color television was first defined in 1953, but up until recently, the rules set out at that time have been difficult to implement. The computer industry is scattered enough so that there is not yet a recognition for display standards. Yet if they really want to move into the information age, there must be standards for displaying information. I've found myself in a position of demonstrating the need for such standards.
Da-Lite: If the purpose of the devices you're talking about is to project the display, may we say that the function of a screen is to display the display?
Da-Lite: And would you agree that the principal purpose of a screen is to deliver as much as possible of the energy from the projector into the eyes of the audience?
Kane: Well, yes. But the mention of energy makes me nervous. I have been fighting how much is made of that particular parameter above all others. Concentrating on light output over everything else is why most display devices produce such poor picture quality.
Da-Lite: So you believe that brightness should not be the foremost attribute of a display device?
Kane: I do. The first job of a display device is the accurate reproduction of the incoming signal.
Da-Lite: And the first job of a screen?
Kane: The screen should maintain the standards of the signal.
Da-Lite: If we're looking at an image displayed on a screen, what are the essential attributes of a good looking picture?
Kane: The ideal contribution of the screen to the display of information should be neutrality and uniformity.
Da-Lite: By uniformity do you mean that from any given position a viewer could not easily detect brightness differentials between one area of the image and another?
Kane: Uniformity of performance encompasses more parameters than just light output. There is, for instance, the ability of the display to handle color as well as light falloff.
Da-Lite: What sorts of properties should a screen have to enable it to handle color uniformly?
Kane: Any surface will somehow alter the characteristics of light hitting it. That's the nature of a surface; it's going to absorb some of the light and it's going to reflect or cast off part of it. It's important that a screen be able to reflect light off its surface equally across the entire light spectrum, at every point in the screen.
Another parameter of the screen surface is that it needs to be fine enough not to get in the way of light variations in the source signal. In other words, the screen shouldn't interfere with the detail in the picture.
Da-Lite: Are there other levels of uniformity?
Kane: Yes. There are actually three. First there's white field uniformity, or light output across the entire screen. This has a physical component, such as not having any waves.
Second its color characteristics have to be equal. That means that it has to provide the viewer with a flat spectral response.
And third there can't be anything, either fixed or variable, that would get in the way of light variations from the source which would degrade detail.
Da-Lite: What other characteristics should we think about? And, do they apply equally to front and rear projection screens?
Kane: Rear screens are slightly more complicated than front screens. A big issue is the thickness of the diffusing material. The light goes into a different medium (air to plastic or air to glass) and is bent. If the diffuser is too thick, this bending will defocus the information. Although they aren't Da-Lite's, there are some rear screen coatings being used today where a small dot of white on a black background will become an unrecognizable blur when it finally exits the screen. That's why it's so important for the diffuser to be as thin as possible.
Da-Lite: We absolutely agree with you.
Kane: Then, in rear screen in particular, there arises the necessity to help out the projector a little bit. This is so because the typical CRT projector is usually way too close to the screen and the rays of light coming from its three lenses are by no means parallel as they travel towards the screen. So we want to add a helper behind the screen called a Fresnel which helps bring the light straight out of the screen.
Da-Lite: Do you believe that a Fresnel lens would be a useful addition to any rear screen?
Kane: Actually there's a practical limit to a Fresnel's utility. If the focal length of projection lenses is greater than 1.5, the value of a Fresnel become questionable.
Da-Lite: But if the throw distance for a particular projector is, say, 1.2 times the screen's diagonal...?
Kane: Then you should include a Fresnel behind the diffuser screen.
Da-Lite: A display attribute you have yet to mention is contrast.
Kane: That brings us back to the discussion of diffusion thickness, which of course can destroy the contrast ratio.
Da-Lite: Why would it do that?
Kane: The real contrast of a picture is determined by looking at adjacent areas of an image. It is not determined from one corner of the picture to another or from a totally black picture to a 100% white picture. If you measure anything other than the adjacent area contrast it will have little or no meaning to the eye. What a thick diffuser does is kill adjacent area contrast by scattering the light from a bright area to an adjacent dark area. This will dump the contrast ratio something fierce and the picture will look washed out.
Da-Lite: At Da-Lite we believe that adding gray colorant to some of our Polacoat diffusers measurably improves screen contrast, particularly in the presence of ambient light. What is your opinion of that practice?
Kane: Unfortunately what I've found is that putting a colorant in the screen to absorb ambient light alters the spectral response of the screen. You see, when you put any kind of a coating on the screen the light passing through it will become attenuated. And once you start attenuating light, it's important to attenuate it equally clear across the spectrum. You can't favor one part of the spectrum versus another. However, that's what the majority of the screen darkening materials do.
This is why your Video Vision screen is a milky white color. That's what it takes to get a flat spectral response with a fairly high degree of efficiency in being able to pass the light from the source to the audience.
Da-Lite: Thus far you've been describing display devices which are 3-gun, CRT projectors. What happens if there is only a single exit pupil?
Kane: With three tube projectors the light sources are spatially separated and there will be differential gains to contend with. When all of the light comes from a single source spectral response is not as critical. Yet even here, white field uniformity remains important. The whole motion picture industry has begun to realize this, for example, in its attitude towards front screens. They're realizing that their movies look better on your Cinema Vision material than they do on the material they've always been using. They notice an improvement on Cinema Vision, because it's got such a flat spectral response, the blues and the reds are noticeably richer.
Da-Lite: Although much of your work focuses on kinetic imagery, how different do you think the standards are for static displays?
Kane: I think there are two important differences. In a static image we are able to produce more apparent resolution at the source and, when the image is not moving, we tend to study it more.
Da-Lite: In this context what then do you think about including lenticulations in the surface of a screen that, as you put it, has to be studied? Many people conclude that if you have twice as many lenticulations as the upper limit of the display device's horizontal resolution, you're OK. What do you conclude?
Kane: I think it was Nyquist who first theorized that all you need is two samples per element to tell what's there. I am much more conservative than that. I believe that screen resolution should be at least ten times greater than the source resolution.
Da-Lite: So you're saying you want as high a sampling rate as you can find?
Kane: Yes; that's what diffusers are all about. Some people in the industry have specified a good, thin diffuser as having a sampling rate of about 1400 lines per inch. There's not a display device in the world that could even approach resolution that fine.