17

Displays

You cannot see data. The information that your computer processes is nothing but ideas, and ideas are intangible no matter whether in your mind or your computer's. Whereas you can visualize your own ideas, you cannot peer directly into the pulsing digital thought patterns of your computer. You probably have no right to think that you could—if you can't read another person's thoughts, you should hardly expect to read the distinctly non-human circuit surges of your PC.

Although most people—at least those not trained in stage magic—cannot read thoughts per se, they can get a good idea of what's going on in another person's mind by carefully observing his external appearance. Eye movements, facial expressions, gestures, and sometimes even speech can give you a general idea about what that other person is thinking, although you will never be privy to his true thoughts. So it is with computers. You'll never be able to see electrons tripping through logical gates, but you can get a general idea of what's going on behind the screen by looking into the countenance of your computer—its display. What the display shows you is a manifestation of the results of the computer's thinking.

The display is your computer's line of communication to you, much as the keyboard enables you to communicate with it. Like even the best of friends, the display doesn't tell you everything; but it does give you a clear picture, one from which you can draw your own conclusions about what the computer is doing.

Because the display has no direct connection to the computer's thoughts, the same thoughts—the same programs—can generate entirely different onscreen images while working exactly the same way inside your computer. Just as you can't tell a book's contents from its cover, you cannot judge the quality of a computer from its display.

What you can see is important, however, because it influences how well you can work with your computer. A poor display can lead to eyestrain and headaches, making your computer literally a pain to work with. A top quality display means clearly defined characters, sharp graphics, and a system that's a pleasure to work with.

Background

Although the terms are often used interchangeably, a display and a monitor are distinctly different. A display is the image producing device itself, the screen that you see. The monitor is a complete box that adds support circuitry to the display. This circuitry converts the signals set by the computer (or some other device, such as a videocassette recorder) into the proper form for the display to use. Although most monitors operate under principles like those of the television set, displays can be made from a variety of technologies, including liquid crystals and the photon glow of some noble gases.

Because of their similar technological foundations, monitors to a great extent resemble the humble old television set. Just as a monitor is a display enhanced with extra circuitry, the television is a monitor with even more signal conversion electronics. The television incorporates into its design a tuner or demodulator that converts signals broadcast by television stations or a cable television company into about the same form as those signals used by monitors. Beyond the tuner, the television and monitor work in much the same way. Indeed, some old-fashioned computer monitors work as televisions as long as they are supplied the proper signals.

New monitors have developed far beyond their television roots, however. They have greater sharpness and purity of color. To achieve these ends, they operate at higher frequencies than television stations can broadcast.

Computer displays and monitors use a variety of technologies to create visible images. A basic bifurcation divides the displays of desktop computers and those of laptop machines. Most desktop computers use systems based on the same cathode ray tube technology akin to that used in the typical television set. Laptop and notebook computers chiefly use liquid crystal displays. Occasionally, some designers switch hit with technologies and stuff LCDs into desktop machines—something we're destined to see more in the future—while aged portable computers weighed themselves down with CRT displays.

Cathode Ray Tubes

The oldest electronic image generating system still in use is the cathode ray tube. The name is purely descriptive. The device is based on a special form of vacuum tube—a glass bottle that is partially evacuated and filled with an inert gas at very low pressure. The tube of the CRT is hardly a tube but more flask shaped with a thin neck that broadens like a funnel into a wide, nearly flat face. Although a CRT appears to be made like a simple bottle—in fact, people in the monitor business sometimes refer to CRTs as "bottles"—its construction is surprisingly complex and involves a variety of glasses of many thicknesses. The face of the typical CRT, for example, often is about an inch thick.

The cathode in the CRT name is a scientific term for a negatively charged electrode. In a CRT, a specially designed cathode beams a ray of electrons toward a positively charged electrode, the anode. (Electrons, having a negative charge, are naturally attracted to positive potentials.) Because it works like a howitzer for electrons, the cathode of a CRT is often called an electron gun.

The electrons race on their ways at a substantial fraction of the speed of light, driven by the high voltage potential difference between the cathode and anode, sometimes as much as 25,000 volts.

At the end of their flight to the anode, the electrons crash into a coating made from phosphor compounds that has the amazing ability to convert the kinetic energy of the electrons into visible light.

Physical Characteristics

The CRT is a physical entity that you can hold in your hand, drop on the floor, and watch shatter. Little of its design is by chance—nearly every design choice in making the CRT has an effect on the image that you see.

Four elements of the CRT exert the greatest influence on the kind and quality of image made by a monitor. The phosphors chosen for the tube affect the color and persistence of the display. The electron guns actually paint the image, and how well they work is a major factor in determining image sharpness. In color CRTs, the shadow mask or aperture grille limit the ultimate resolution of the screen. The face of the screen and the glare reflected from it affect both image contrast and how happy you will be in working with a monitor.

Phosphors

At the end of the electrons' short flight from the gun in the neck of a CRT to the inside of its wide, flat face lies a layer of a phosphor-based compound with a wonderful property—it glows when struck by an electron beam. The image you see in a CRT is the glow of the electrically-stimulated phosphor compounds, simply termed phosphors in the industry. Not all the phosphorous compounds used in CRTs are the same. Different compounds and mixtures glow various colors and for various lengths of time after being struck by the electron beam.

A number of different phosphors are used by PC-compatible monitors. Table 17.1 lists some of these phosphors and their characteristics.

The type of phosphor determines the color of the image on the screen. Several varieties of amber, green, and whitish phosphors are commonly used in monochrome displays. Color CRT displays use three different phosphors painted in fine patterns across the inner surface of the tube. The patterns are made from dots or stripes of the three additive primary colors—red, green, and blue—arrayed next to one another. A group of three dots is called a color triad or color triplet.

One triad of dots makes up a picture element, often abbreviated as pixel (although some manufacturers prefer to shorten picture element to pel).

The makers of color monitors individually can choose each of the three colors used in forming the color triads on the screen. Most monitor makers have adopted the same phosphor family, which is called P22 (or B22, which is the same thing with a different nomenclature), so the basic color capabilities of most multi-hued monitors are the same.

The color monitor screen can be illuminated in any of its three primary colors by individually hitting the phosphor dots associated with that color with the electron beam. Other colors can be made by illuminating combinations of the primary colors. By varying the intensity of each primary color, an infinite spectrum can be generated.

Monochrome displays have their CRTs evenly coated with a single, homogenous phosphor so that wherever the electron beam strikes, the tube glows in the same color. The color of the phosphors determines the overall color that the screen glows.

Three colors remain popular for monochrome computer displays—amber, green, and white. Which is best is a matter of both preference and prejudice. Various studies support the superiority of each of these colors:

Green. Green screens got a head start as PC displays because they were IBM's choice for most of its terminals and the first PC display—as well as classic radar displays and oscilloscopes. It is a good selection for use where ambient light levels are low; part of its heritage is from the days of oscilloscope and radar screens (most of which remain stubbornly green). Over the last few years, however, green has fallen from favor as the screen of choice.

Amber. In the 1980s, amber-colored screens rose in popularity because they are, according to some studies, easier on the eyes and more readable when the surrounding environmental light level is bright. Yellow against black yields one of the best perceived contrast combinations, making the displays somewhat easier on your eyes. Amber also got a push as being a de facto European monitor standard.

White. Once white screens were something to be avoided, if just from their association with black and white televisions. A chief reason was that most early monochrome displays used a composite interface and gave low onscreen quality.

Apple's Macintosh and desktop publishing forced the world to re-evaluate white. White is the color of paper that executives have been shuffling through offices over the ages. White and black also happen to be among the most readable of all color combinations. In 1987, IBM added impetus to the conversion of the entire world to white with the introduction of the VGA and its white screen monochrome display.

If you look closely, you might see fine specs of colors, such as a bright yellow dappled into so-called "white" phosphors. Manufacturers mix together several different phosphors to fine tune the color of the monochrome display—to make it a cool, blue television "white" or a warm, yellowish paper "white."

No good dividing line exists between ordinary white and paper white displays. In theory, paper white means the color of the typical bond paper you type on, a slightly warmer white than the blue tinged glow of most "white" monitors. But "paper whiteness" varies with who is giving the name.

Often ignored yet just as important to screen readability as the phosphor colors is the background color of the display tube. Monochrome screen backgrounds run the full range from light gray to nearly black. Darker screens give more contrast between the foreground text and the tube background, making the display more readable, particularly in high ambient light conditions.

The background area on a color screen—that is, the space between the phosphor dots—is called the matrix, and it is not illuminated by the electron beam. The color of the matrix determines what the screen looks like when the power is off—pale gray, dark green-gray, or nearly black. Darker and black matrices give an impression of higher contrast to the displayed images. Lighter gray matrices make for purer white. The distinctions are subtle, however, and unless you put two tubes side by side, you're unlikely to be able to judge the difference.

Color Temperature

If your work involves critical color matching, the color temperature of your monitor can be an important issue. White light is not white, of course, but a mixture of all colors. Alas, all whites are not the same. Some are richer in blue, some in yellow. The different colors of white are described in their color temperature, the number of Kelvins (degrees Celsius above absolute zero) that a perfect luminescent body would need to be to emit that color.

Like the incandescence of a hot iron horseshoe in the blacksmith's forge, as its temperature gets higher the hue of a glowing object shifts from red to orange to yellow and on to blue white. Color temperature simply assigns an absolute temperature rating to these colors. Figure 17.1 illustrates the range of color temperatures.

Figure 17.1 The color temperatures associated with various conditions.

For example, ordinary light bulbs range from 2,700 to 3,400 Kelvins. Most fluorescent lights have non-continuous color spectra rich in certain hues (notably green) while lacking others that makes assigning a true color temperature impossible. Other fluorescent lamps are designed to approximate daylight with color temperatures of about 5,000 Kelvins.

The problem with color matching arises because pigments and paper only reflect light, so their actual color depends on the temperature of the light illuminating them. Your monitor screen emits light, so its color is independent of illumination—it has its own color temperature that may be (and is likely) from that lighting the rest of your work. Monitors are designed to glow with the approximate color temperature of daylight rather than incandescent or fluorescent light.

Alas, not everyone has the same definition of daylight. Noonday sun, for instance, ranges from 5,500 to 6,000 Kelvins. Overcast days may achieve a color temperature of 10,000 Kelvins because the scattered blue glow of the sky (higher color temperature) dominates the yellowish radiation from the sun. The colors and blend of the phosphors used to make the picture tube screen and the relative strengths of the electron beams illuminating those phosphors determine the color temperature of a monitor. Some engineers believe the perfect day is a soggy, overcast afternoon suited only to ducks and Englishmen and opt to run their monitors with a color temperatures as high as 10,000 Kelvins. Others, however, live in a Kodachrome world where the color temperature is the same 5,300 Kelvins as a spring day with tulips in the park.

Persistence

CRT phosphors also differ in persistence, which describes how long the phosphor glows after being struck by the electron beam. Most monitors use medium persistence phosphors.

Persistence becomes obvious when it is long. Images take on a ghostly appearance, lingering for a few seconds and slowly fading away. Although the effect may be bothersome, particularly in a darkened room, it's meant to offset the effect of another headache producer: flicker.

Exactly what it sounds like, flicker is the quick flashing of the screen image caused by the image decaying before it gets re-scanned by the electron beam. The persistence of vision (a quality of the human visual system) makes rapidly flashing light sources appear continuously lit. Fluorescent lights, for example, seem to glow uninterruptedly even though they switch on and off 120 times a second (twice the nominal frequency of utility supplied electricity).

The lingering glow of long persistence phosphors bridges over the periods between passes of electron beams when they stretch out too long for human eyes to blend them together. Long persistence phosphors are thus often used in display systems scanned more slowly than usual, such as interlaced monitors. The IBM Monochrome display, perhaps the most notorious user of long persistence green phosphors, is scanned 50 times a second instead of the more normal (and eye-pleasing) 60 or higher.

Long persistence phosphors need not be green, however. Long persistence color systems also are available for use in applications where flicker is bothersome. Most often, long persistence color phosphors are used in interlaced systems that are scanned more slowly than non-interlaced displays.

Long persistence phosphors also frustrate light pens, which depend on detecting the exact instant a dot of phosphor lights up. Because of the lingering glow, most light pens perceive several dots to be lit simultaneously. The pen cannot zero in on a particular dot position on the screen.

Electron Guns

To generate the beams that light the phosphors on the screen, a CRT uses one or more electron guns. An electron gun is an electron emitter (a cathode) in an assembly that draws the electrons into a sharp, high speed beam. To move the beam across the breadth of the tube face (so that the beam doesn't light just a tiny dot in the center of the screen), a group of powerful electromagnets arranged around the tube, collectively called the yoke, bend the electron beam in the course of its flight. The magnetic field set up by the yoke is carefully controlled and causes the beam to sweep each individual display line down the face of the tube.

Monochrome CRTs have a single electron gun that continuously sweeps across the screen. Most color tubes have three guns, although some color televisions and monitors boast "one gun" tubes, which more correctly might be called "three guns in one." The gun count depends on the definition of a gun. Like all color CRTs, the one gun tubes have three distinct electron emitting cathodes that can be individually controlled. The three cathodes are fabricated into a single assembly that allows them to be controlled as if they were generating only a single beam.

In a three gun tube, the trio of guns is arranged in a triangle. One gun tubes arrange their cathodes in a straight line, often earning the epithet inline guns. In theory inline guns should be easier to set up, but as a practical matter, excellent performance can be derived from either arrangement.

The three guns in a color CRT emit their electrons simultaneously, and the three resulting beams are steered together by the yoke. Individual adjustments are provided for each of the three beams, however, to ensure that each beam falls exactly on the same triplet of color dots on the screen as the others. Because these controls help the three beams converge on the same triad, they are called convergence controls. The process of adjusting them is usually termed alignment.

Convergence

The three electron beams inside any color monitor must converge on exactly the right point on the screen to illuminate a single triad of phosphor dots. If a monitor is not adjusted properly—or if it is not designed or made properly—the three beams cannot converge properly to one point. Poor convergence results in images with rainbow-like shadows and a loss of sharpness and detail, as illustrated in Figure 17.2. Individual text characters no longer appear sharply defined but become two- or three-color blurs. Monochrome monitors are inherently free from such convergence problems because they have but one electron beam.

Figure 17.2 Excellent (left) and poor convergence on a monitor screen.

Convergence problems are a symptom rather than a cause of monitor deficiencies. Convergence problems arise not only from the design of the display, but also from the construction and setup of each individual monitor. It can vary widely from one display to the next and may be aggravated by damage during shipping.

The result of convergence problems is most noticeable at the screen periphery because that's where the electron beams are the most difficult to control. When bad, convergence problems can be the primary limit on the sharpness of a given display, having a greater negative effect than wide dot pitch or low bandwidth ( both of which are discussed later in this chapter).

Many monitor makers claim that their convergence is a given fraction of a millimeter at a particular place on the screen. If a figure is given for more than one screen location, the center of the screen invariably has a lower figure—tighter, better convergence—than a corner of the screen.

The number given is how far one color may spread from another at that location. Lower numbers are better. Typical monitors may claim convergence of about 0.5 (one-half) millimeter at one of the corners of the screen. That figure often rises 50 percent higher than the dot pitch of the tube, making the convergence the limit on sharpness for that particular monitor.

Misconvergence problems often can be corrected by adjustment of the monitor. Many monitors have internal convergence controls. A few, high resolution (and high cost) monitors even have external convergence adjustments. But adjusting monitor convergence is a job for the specialist—and that means getting a monitor converged can be expensive, as is any computer service call.

Many monitor makers now claim that their products are converged for life. Although this strategy should eliminate the need to adjust them (which should only be done by a skilled technician with the correct test equipment), it also makes it mandatory to test your display before you buy it. You don't want a display that's been badly converged for life.

Purity

The ability of a monitor to show you an evenly lit screen that does not vary in color across its width is termed purity. A monitor with good purity will be able to display a pure white screen without a hint of color appearing. A monitor with poor purity will be tinged with one color or another in large patches. Figure 17.3 illustrates the screens with good and poor purity.

Figure 17.3 Comparison of good and bad monitor purity.

Poor purity often results from the shadow mask or aperture grille of a cathode ray tube becoming magnetized. Degaussing the screen usually cures the problem. Most larger monitors have built-in automatic degaussers.

You can degauss your monitor with a degaussing loop designed for color televisions or even a bulk tape eraser. Energize the degausing coil or tape eraser in close proximity to the screen, then gradually remove the coil to a distance of three or more feet away before switching it off. The gradually declining alternating magnetic field will overpower the static field on the mask, and the gradual removal of the alternating field prevents the strong field from establishing itself on the mask.

Shadow Masks

Just pointing the electron beams at the right dots is not enough because part of the beam can spill over and hit the other dots in the triplet. The result of this spillover is a loss of color purity—bright hues become muddied. To prevent this effect and make images as sharp and colorful as possible, all color CRTs used in computer displays and televisions alike have a shadow mask—a metal sheet with fine perforations in it, located inside the display tube and a small distance behind the phosphor coating of the screen.

The shadow mask and the phosphor dot coating on the CRT screen are critically arranged so that the electron beam can only hit phosphor dots of one color. The other two colors of dots are in the "shadow" of the mask and cannot be seen by the electron beam.

The spacing of the holes in the shadow mask to a great degree determines the quality of the displayed image. For the geometry of the system to work, the phosphor dots on the CRT screen must be spaced at the same distance as the holes in the mask. Because the hole spacing determines the dot spacing, it is often termed the dot pitch of the CRT.

The dot pitch of a CRT is simply a measurement of the distance between dots of the same color. It is an absolute measurement, independent of the size of the tube or the size of the displayed image.

The shadow mask affects the brightness of a monitor's image in two ways. The size of the holes in the mask limits the size of the electron beam getting through to the phosphors. Off-axis from the guns—that is, toward the corners of the screen—the round holes appear oval to the gun and less of the beam can get through. As a result, the corners of a shadow mask screen are often dimmer than the center, although the brightness difference may not be distinguishable.

The mask also limits how high the electron beam intensity can be in a given CRT. A stronger beam—which makes a brighter image—holds more energy. When the beam strikes the mask, part of that energy is absorbed by the mask and becomes heat, which raises the temperature of the mask. In turn, this temperature rise makes the mask expand unpredictably, distorting it minutely and blurring the image. To minimize this heat induced blur, monitor makers are moving to making shadow masks from materials that have a low coefficient of thermal expansion. That is, they change size as little as possible with temperature. The alloy Invar is favored for shadow masks because of its capability to maintain a nearly constant size as it warms.

Aperture Grilles

With all the problems associated with shadow masks, you might expect someone to come up with a better idea. Sony Corporation did exactly that, inventing the Trinitron picture tube.

The Trinitron uses an aperture grille—slots between a vertical array of wires—instead of a mask. The phosphors are painted on the inner face of the tube as interleaved stripes of the three additive primary colors. The grille blocks the electron beam from the wrong stripes just as a shadow mask blocks it from the wrong dots. The distance between two sequential stripes of the same color is governed by the spacing between the slots between the wires—the slot pitch of the tube. Because the electron beam fans out as it travels away from the electron gun and stripes are farther from the gun than is the mask, the stripes are spaced a bit farther apart than the slot pitch. Their spacing is termed screen pitch. For example, a 0.25 millimeter slot pitch Trinitron might have a screen pitch of 0.26 millimeter. Figure 17.4 shows how slot pitch as well as dot pitch are measured.

Figure 17.4 Measuring dot pitch and slot pitch.

The wires of the aperture grille are quite thick, about two-thirds the width of the slot pitch. For example, in a Trinitron with a 0.25 slot pitch, the grille wires measure about 0.18 millimeters in diameter because each electron beam is supposed to illuminate only one-third of the screen. The wires shadow the other two-thirds from the beam to maintain the purity of the color.

The aperture grille wires are held taut, but they can vibrate. Consequently, Trinitron monitors have one or two thin tensioning wires running horizontally across the screen. Although quite fine, these wires cast a shadow on the screen that is most apparent on light-colored screen backgrounds. Some people find the tensioning wire shadows objectionable, so you should look closely at a Trinitron before buying.

Trinitrons hold a theoretical brightness advantage over shadow mask tubes. Because the slots allow more electrons to pass through to the screen than do the tiny holes of a shadow mask, a Trinitron can (in theory) create a brighter image. This added brightness is not borne out in practice. However, Trinitrons do excel in keeping their screens uniformly bright. The aperture grille wires of a Trinitron block the beam only in one dimension, and so don't impinge as much on the electron beam at the screen edges.

Thanks to basic patents, Sony had exclusive rights to the Trinitron design. However, those patents began expiring in 1991, and other manufactures were quick to begin working with the technology. Other patents, however, cover manufacturing and other aspects of building successful Trinitrons. Consequently, an expected flood of Trinitron clones never appeared. In fact, the only new alternative to the Trinitron was introduced by Mitsubishi in 1993. Called Diamondtron by its manufacturer, the new design is based on aperture grille technology, but uses a refined electron gun. Whereas the Trinitron combines three guns into a single focusing mechanism, the Diamondtron gives each gun its own control. According to Mitsubishi, this refinement allows more precise beam control and a more accurate and higher resolution image.

Required Dot Pitch

No matter whether a monitor uses a shadow mask with a dot pitch or an aperture grille with a slot pitch, the spacing of image triads on the screen is an important constituent in monitor quality. A monitor simply cannot put dots any closer together than the holes in mask or grille allow. It's easy to compute the pitch necessary for a resolution level in a computer system. Just divide the screen size by the number of dots required to be displayed.

For example, a VGA text display comprises 80 columns of characters each nine dots wide, for a total of 720 dots across the screen. The typical twelve-inch (diagonal) monitor screen is roughly 9.5 inches or 240 millimeters across. Hence, to properly display a VGA text image, the dot pitch must be smaller than .333 (or 240/720) millimeter, assuming the full width of the screen is used for display. Often a monitor's image is somewhat smaller than full screen width and such displays require even finer dot pitch. The larger the display, the coarser the dot pitch can be for a given level of resolution.

Line Width

Another factor limits the sharpness of monitor images, the width of the lines drawn on the screen. Ideally, any vertical or horizontal line on the screen will appear exactly one pixel wide, but in practical monitors the width of a line may not be so compliant. If lines are narrower than one pixel wide, thin black lines will separate adjacent white lines and wide white areas will be thinly striped in black. If the line width exceeds the size of a pixel, the display's ability to render fine detail will be lost.

The ideal line width for a monitor varies with the size of the screen and the resolution displayed on the screen. As resolution increases, lines must be narrower. As screen size goes up (with the resolution constant), line width must increase commensurately. For example, a monitor with an active image area that's ten inches wide (about what you'd expect from a 14-inch display) will require an ideal line width of 1/64th inch or about 0.4 millimeter.

Several factors influence the width of lines on the screen. The monitor must be able to focus its electron beam into a line of ideal width. However, width also varies with the brightness of the beam—brighter beams naturally tend to expand out of focus. Consequently, when you increase the brightness of your monitor, the sharpness of the image may decline. For this reason, test laboratories usually make monitor measurements at a standardized brightness level.

Screen Curvature

Most CRTs have a distinctive shape. At one end, a narrow neck contains the electron gun or guns. Around the neck fits the deflection yoke, an external assembly that generates the magnetic fields that bend the electron beams to sweep across the inner surface of the wide face of the tube. The tube emerges from the yoke as a funnel-like flaring, which enlarges to the rectangular face of the screen itself. This face often (but becoming much less common) is a spherically curving surface.

The spherical curve of the face makes sense for a couple of reasons. It makes the distance traveled by the electron beam more consistent at various points on the screen, edge to center to edge. A truly flat screen would require the beam to travel farther at the edges than at the center and would require the beam to strike the face of the screen obliquely, resulting in image distortion. Although this distortion can be compensated for electrically, the curving screen helps things along.

In addition, the CRT is partly evacuated, so normal atmospheric pressure is constantly trying to crush the tube. The spherical surface helps distribute this potentially destructive force more evenly, making the tube stronger.

Screen curvature has a negative side effect. Straight lines on the screen appear straight only from one observation point. Move your head closer, farther away, or to one side, and the supposedly straight lines of your graphics images will bow this way and that.

Technology has made the reasons underlying spherical curved screens less than compelling. The geometry of inline guns simplifies tube construction and alignment sufficiently that cylindrically curved screens are feasible. They have fewer curvilinear problems because they warp only one axis of the image. Trinitrons characteristically have faces with cylindrical curves. Most shadow mask tubes have spherical faces.

In the last few years, the technical obstacles to making genuinely flat screens have been surmounted. A number of manufacturers now offer flat screen monochrome displays, which are relatively simple because compensation for the odd geometry is required by only one electron beam.

The first color flat screen was Zenith's flat tension mask system. The tension mask solves the construction problems inherent in a flat screen color system by essentially stretching the shadow mask. Its flat face and black matrix make for very impressive images, only the case of the monitor itself is bulky and ugly to look at; and an internal fan made the first model as much a pain for the ears as the screen was a pleasure for the eyes. Since then, such monitors have become less power hungry, but they remain more costly than more conventional designs.

Today's so called flat square tubes are neither flat nor square. They are, however, flatter and squarer than the picture tubes of days gone by so they suffer less curvilinear distortion.

Resolution Versus Addressability

The resolution of a video system refers to the fineness of detail that it can display. It is a direct consequence of the number of individual dots that make up the screen image and thus is a function of both the screen size and the dot pitch.

Because the size and number of dots limit the image quality, the apparent sharpness of screen images can be described by the number of dots that can be displayed horizontally and vertically across the screen. For example, the resolution required by the Video Graphics Array system in its standard graphics mode is 640 dots horizontally by 480 vertically. Modern display systems may produce image with as many as 1600 by 1200 dots in their highest resolution mode.

Sometimes, however, the resolution available on the screen and that made by a computer's display adapter are not the same. For example, a video mode designed for the resolution capabilities of a color television set hardly taps the quality available from a computer monitor. On the other hard, the computer generated graphics may be designed for a display system that's sharper than the one being used. You might, for instance, try to use a television in lieu of a more expensive monitor. The sharpness you actually see would then be less than what the resolution of the video system would have you believe.

Actual resolution is a physical quality of the video display system—the monitor—that's actually being used. It sets the ultimate upper limit on the display quality. In color systems, the chief limit on resolution is purely physical—the convergence of the system and the dot pitch of the tube. In monochrome systems, which have no quality limiting shadow masks, the resolution is limited by the bandwidth of the monitor, the highest frequency signal with which it can deal. (Finer details pack more information into the signals sent from computer system to monitor. The more information in a given time, the higher the frequency of the signal.)

A few manufacturers persist in using the misleading term addressability to describe the quality of their monitors. Addressability is essentially a bandwidth measurement for color monitors. It indicates how many different dots on the screen the monitor can point its electron guns at. It ignores, however, the physical limit imposed by the shadow mask. In other words, addressability describes the highest quality signals the monitor can handle, but the full quality of those signals is not necessarily visible to you onscreen.

Anti-Glare Treatment

Most mirrors are made from glass, and glass tries to mimic the mirror whenever it can. Because of the difference between the index of refraction of air and that of glass, glass is naturally reflective. If you make mirrors, that's great. If you make monitors—or worse yet, use them—the reflectivity of glass can be a big headache. A reflection of a room light or window from the glass face of the CRT can easily be brighter than the glow of phosphors inside. As a result, the text or graphics on the display tends to "wash out" or be obscured by the brightness.

The greater the curvature of a monitor screen, the more apt it is to have a problem with reflections because more of the environment gets reflected by the screen. A spherical monitor face acts like one of those huge convex mirrors strategically hung to give a panoramic view of shoplifters or cars sneaking around an obscured hairpin turn. The flatter the face of the monitor, the less of a worry reflections are. With an absolutely flat face, a slight turn of the monitor's face can eliminate all glare and reflections.

You can't change the curve of your monitor's face. However, help is available. Anti-glare treatments can reduce or eliminate reflections from the face of most CRTs. Several glare reduction technologies are available, and each varies somewhat in its effectiveness.

Mesh

The lowest tech and least expensive anti-glare treatment is simply a fabric mesh, usually nylon. The mesh can either be placed directly atop the face of the screen or in a removable frame that fits about half an inch in front of the screen. Each hole in the mesh acts like a short tube, allowing you to see straight in at the tube, but cutting off light from the sides of the tube. Your straight-on vision gets through unimpeded, while glare that angles in doesn't make it to the screen.

As simple as this technique is, it works amazingly well. The least expensive after-market anti-glare system uses mesh suspiciously similar to pantyhose stretched across a frame. Unfortunately this mesh has an unwanted side effect. Besides blocking the glare, it also blocks some of the light from the screen and makes the image appear darker. You may have to turn the brightness control up to compensate, which may make the image bloom and lose sharpness.

Mechanical

Glare can be reduced by mechanical means—not a machine that automatically intercepts glare before it reaches the screen, but mechanical preparation of the screen surface. By lightly grinding the glass on the front of the CRT, the face of the screen can be made to scatter rather than reflect light. Each rough spot on the screen that results from the mechanical grinding process reflects light randomly, sending it every which direction. A smooth screen reflects a patch of light all together, like a mirror, reflecting any bright light source into your eyes. Because the light scattered by the ground glass is dispersed, less of it reaches your eyes and the glare is not as bright. However, because the coarse screen surface disperses the light coming from inside the tube as well as that reflected from the tube face, it also lessens the sharpness of the image. The mechanical treatment makes text appear slightly fuzzier and out of focus, which to some manufacturers is a worse problem than glare.

Coating

Glare can be reduced by applying coatings to the face of the CRT. Two different kinds of coatings can be used. One forms a rough film on the face of the CRT. This rough surface acts in the same way as a ground glass screen would, scattering light.

The screen also can be coated with a special compound like magnesium fluoride. By precisely controlling the thickness of this coating, the reflectivity of the surface of the screen can be reduced. The fluoride coating is made to be a quarter the wavelength of light (usually of light at the middle of the spectrum). Light going through the fluoride and reflecting from the screen thus emerges from the coating out of phase with the light striking the fluoride surface, visually canceling out the glare. Camera lenses are coated to achieve exactly the same purpose, the elimination of reflections. A proper coating can minimize glare without affecting image sharpness or brightness.

Polarization

Light can be polarized; that is, its photons can be restricted to a single plane of oscillation. A polarizing filter allows light of only one polarization to pass. Two polarizing filters in a row can be arranged to allow light of only one plane of polarization to pass (by making the planes of polarization of the filters parallel), or the two filters can stop light entirely when their planes of polarization are perpendicular.

The first filter lets only one kind of light pass; the second filter lets only another kind of light pass. Because none of the second kind of light reaches the second filter, no light gets by.

When light is reflected from a surface, its polarization is shifted by 90 degrees. This physical principle makes polarizing filters excellent reducers of glare.

A sheet of polarizing material is merely placed a short space in front of a display screen. Light from a potential source of glare goes through the screen and is polarized. When it strikes the display and is reflected, its polarization is shifted 90 degrees. When it again reaches the filter, it is out of phase with the filter and cannot get through. Light from the display, however, only needs to go through the filter once. Although this glow is polarized, there is no second screen to impede its flow to your eyes.

Every anti-glare treatment has its disadvantage. Mesh makes an otherwise sharp screen look fuzzy because smooth characters are broken up by the cell structure of the mesh. Mechanical treatments are expensive and tend to make the screen appear to be slightly "fuzzy" or out of focus. The same is true of coatings that rely on the dispersion principle. Optical coatings, Polaroid filters, and even mesh suffer from their own reflections. The anti-glare material itself may add its own bit of glare. In addition, all anti-glare treatments—polarizing filters in particular—tend to make displays dimmer. The polarizing filter actually reduces the brightness of a display to one-quarter its untreated value.

Even with their shortcomings, however, anti-glare treatments are amazingly effective. They can ease eyestrain and eliminate the headaches that come with extended computer use.

Image Characteristics

Physical aspects of a monitor and its electronics control the size, shape, and other aspects of the images it displays. These qualities are defined and characterized in a number of ways. The most rudimentary is screen size—the bigger your monitor screen, the larger the images it can make. Because of the underscanning common among computer monitors, however, the actual image size is almost always smaller than the screen. The aspect ratio of the image describes its shape independent of its size. Most monitors give you a variety of controls to alter the size and shape of the image, so you are the final arbiter of what things look like on your monitor screen.

Screen Size

The most significant measurement of a CRT-based monitor is the size of its screen. Although seemingly straightforward, screen size has been at best an ambiguous measurement and at worst downright misleading.

The confusion all started with television, where confusion often beings. The very first television sets had round CRTs, and their size was easy to measure—simply the diameter of the tube. When rectangular tubes became prevalent in the 1950s, the measurement shifted to the diagonal of the face of the tube. The diagonal was, of course, the closest equivalent to the diameter of an equivalent round tube. It was also the largest dimension that a television manufacturer could reasonably quote.

Unlike television images, which usually cover the entire face of the CRT, computer monitors limit their images to somewhat less. Because the image is most difficult to control at the edges of the screen, monitor makers maintain higher quality by restricting the size of the image. They mask off the far edges of the CRT with the bezel of the monitor case.

That bezel means that no image can fill the entire screen—at least no image that you can entirely see. The tube size becomes irrelevant to a realistic appraisal of the image. Some monitor makers persisted in using it to describe their products. Fortunately, most of the industry recognized this measuring system as optimistic exaggeration, and began using more realistic diagonal measurement of the actual maximum displayable image area.

VESA adopted the diagonal of the maximum image area as the measurement standard in its Video Image Area Definition standard, Version 1.1, which it published on October 26, 1995. This standard requires that screen image area be given as horizontal and vertical measurements of the actual active image area when the monitor is set up by the manufacturer using the manufacturer's test signals. The dimensions must be given in millimeters with an assumed maximum variance of error of plus and minus two percent. Wider tolerances are allowed but must be explicitly stated by the manufacturer. In no case can the expressed image dimensions exceed the area visible through the monitor bezel.

Because the aspect ratio of PC monitor displays is 4:3 (see the "Aspect Ratio" section that follows), computation of the horizontal and vertical screen dimensions from the diagonal is easy. The diagonal represents the hypotenuse of a 3-4-5 right triangle, and that ratio applies to all screen sizes. Table 17.2 lists the dimensions for the most common nominal screen sizes.

Portrait displays, which are designed to give you a view more like the printed sheets that roll out of your laser printer and into envelopes, merely take an ordinary CRT and turn it on its side. The market is not large enough to justify development of custom CRTs for portrait applications. Moreover, the 4:3 aspect ratio works fine because the "active" image on a sheet of letterhead—the space actually occupied by printing once you slice off the top, bottom, left, and right margins—is about eight by ten inches, a nearly perfect fit on a standard picture tube. When measuring the images on these portrait displays horizontal becomes vertical, and all measurements rotate 90 degrees.

Overscan and Underscan

Two monitors with the same size screens may have entirely different onscreen image sizes. Composite monitors are often afflicted by overscan; they attempt to generate images larger than their screen size, and the edges and corners of the active display area may be cut off. (The overscan is often designed so that as the components inside the monitor age and become weaker, the picture shrinks down to normal size—likely over a period of years.) Underscan is the opposite condition—the image is smaller than nominal screen size. For a given screen size, an overscanned image will appear larger at the expense of clipping off the corners and edges of the image as well as increasing distortion at the periphery of the image. Underscanning wastes some of the active area of the monitor screen. Figure 17.5 illustrates the effects of underscanning and overscanning on the same size screen.

Figure 17.5 Underscan and overscan compared.

Underscan is perfectly normal on computer displays and does not necessarily indicate any underlying problems unless it is severe—for example when it leaves a two-inch black band encircling the image. Underscanning helps keep quality high because image geometry is easier to control nearer the center of the screen than it is at the edges. Pulling in the reins on the image can ensure that straight lines actually are displayed straight. Moreover, if you extend the active image to the very edge of the bezel or if you change your viewing position so that you are not facing the screen straight on, the edge of the image may get hidden behind the bezel. The glass in the face of the screen is thicker than you might think, on the order of an inch (25 millimeters), enough that the third dimension will interfere with your most careful alignment.

On the other hand, while overscan gives you a larger image and is the common display mode for video systems, it is not a good idea for PC monitor images. Vital parts of the image may be lost behind the bezel. You may lose to overscan the first character or two from each line of type of one edge of a drafting display. With video, however, people prefer to see as big an image as possible and usually pay little attention to what goes on at the periphery. Broadcasters, in fact, restrict the important part of the images that they deal with to a safe area that will be completely viewable even on televisions with substantial overscan.

Aspect Ratio

The relationship between the width and height of a monitor screen is termed its aspect ratio. Today the shape of the screen of nearly every monitor is standardized, as is that of the underlying CRT that makes the image. The screen is 1.33 times wider than it is high, resulting in the same 4:3 aspect ratio used in television and motion pictures before the wide screen phenomenon took over. Modern engineers now prefer to put the vertical number first to produce aspect ratios that are less than one. Expressed in this way, video has a 3:4 aspect ratio, a value of 0.75.

The choice of aspect ratio is arbitrary and a matter of aesthetics. According to classical Greek aesthetics, the Golden Ratio with a value of about 0.618 is the most beautiful. This beauty is mathematical as well as aesthetic, the solution to the neat little equation x+1 = 1/x. The exact value of the Golden Ratio is irrational, (SQRT(5)-1)/2. Expressed as a ratio of horizontal to vertical, the Golden Ratio is roughly 1.618, the solution to x-1 = 1/x.)

Various display systems feature their own aspect ratios. The modern tendency is toward wider aspect ratios. For example, High Definition Television (HDTV) stretches its aspect ration from the 3:4 of normal video and television to 9:15. The normal negatives you make with your 35mm camera have a 4:6 aspect ratio. The reason video is so nearly square carries over from the early days of television when cathode ray tubes had circular faces. The squarer the image, the more of the circular screen was put to use. Figure 17.6 compares the aspect ratios of three common display systems.

Figure 17.6 Aspect ratios of display systems.

The image on your monitor screen need not have the same aspect ratio of the tube, however. The electronics of monitors separate the circuitry that generates the horizontal and vertical scanning signals and results in their independent control. As a result, the relationship between the two can be adjusted, and that adjustment results in an alteration of the aspect ratio of the actual displayed image. For example, by increasing the amplification of the horizontal signal, the width of the image is stretched, raising the aspect ratio.

Normally, you should expect that the relative gains of the horizontal and vertical signals will be adjusted so that your display shows the correct aspect ratio on its screen. A problem develops when a display tries to accommodate signals based on different standards. This mismatch is particularly troublesome with VGA displays because the VGA standard allows images made with three distinct line counts—350, 400, and 480.

All else being equal, an image made from 350 lines is less than three-quarters the height of a 480-line image. A graphic generated in an EGA-compatible mode shown on a VGA display would therefore look quite squashed. A circle drawn on the screen would look like an ellipse; a orange would more resemble a watermelon.

Image Sizing

Not all monitors take advantage of this sync signaling system. Shifting display modes with such a monitor can lead to graphics displays that look crushed. Others use a technique called autosizing that allows the monitor to maintain a consistent image size no matter what video signal your display adapter is sending it without regard to VGA sync coding. Monitor makers can achieve autosizing in several ways. True autosizing works, regardless of the signal going to the monitor, and scales the image to match the number of display lines. Mode sensitive autosizing works by determining the display mode used for an image from the frequency of the signal. It then switches the size to a pre-set standard to match the number of lines in the signal. Monitors often combine VGA sync-sensing with mode sensitive autosizing.

Image Distortion

Between the electron guns and the phosphors in a cathode ray tube, the electron beam passes through an electronic lens and deflection system that focuses and steers the beam to assure that its path across the screen is the proper size and in the proper place. The electronics of the monitor control the lens and deflection system, adjusting it throughout out the sweep of the electron beam across the screen. In addition to their other chores. the electronics must compensate for the difference in the path of the electron beam at different screen positions. Modern monitors do a magnificent job of controlling beam placement.

When the control system is not properly adjusted, however, the image may exhibit any of a number of defects. Because these defects distort the image from its desired form, they are collectively called image distortion.

The two most common forms of image distortion are barrel distortion and pincushion distortion. Barrel distortion causes vertical or horizontal lines in the image to bow outward so that the center of the lines lies closer to the nearest parallel edge of the screen. Pincushion distortion causes the vertical or horizontal lines in the image to bow inward so that the center of the lines is closer to the center of the screen. Figure 17.7 shows these two kinds of image distortion.

Figure 17.7 Barrel and pincushion distortion.

Barrel and pincushion distortion arise from the same cause, improper image compensation, and are essentially opposites of one another. Overcompensate for pincushion distortion and you get barrel distortion. Collectively the two are sometimes simply called pincushioning no matter which way the lines bow.

Pincushioning is always worse closer to the edges of the image. All monitors have adjustments to compensate for pincushioning, although these adjustments are not always available to you. They may be hidden inside the monitor. Other monitors may include pincushioning adjustments in their control menus. Technician usually use test patterns that display a regular grid on the screen to adjust monitors to minimize pincushioning. You can usually use a full screen image to adjust the pincushioning controls so that the edges of the desktop background color are parallel with the bezel of your monitor.

Less common is trapezoidal distortion that leaves lines at the outer edge of the screen straight but not parallel to the bezel. In other words, instead of your desktop being a rectangle, it is a trapezoid with one side shorter than its opposite side. As with pincushioning, all monitors have controls for trapezoidal distortion but not all make them available to you as the user of the monitor. If your monitor does have an external control for trapezoidal distortion, you adjust it as you do for pincushioning.

Image Controls

A few (far from a majority) monitors make coping with underscan, overscan, and odd aspect ratios simply a matter of twisting controls. These displays feature horizontal and vertical size (or gain) controls that enable you to adjust the size and shape of the image to suit your own tastes. With these controls—providing they have adequate range—you can make the active image touch the top, bottom, and sides of the screen bezel or you can shrink the bright area of your display to a tiny (but geometrically perfect) patch in the center of your screen.

Size and position controls give you command of how much screen the image on your monitor fills. With full range controls, you can expand the image to fill the screen from corner to corner or reduce it to a smaller size that minimizes the inevitable geometric distortion that occurs near the edges of the tube. A full complement of controls includes one each of the following: horizontal position (sometimes termed phase), vertical position, horizontal size (sometimes called width), and vertical size (or height).

A wide control range is better than a narrow one. Some monitors skimp on one or more controls and limit you in how large you can make the onscreen image. Worse, sometimes a monitor maker doesn't include a control at all. For example, some monitors have no horizontal size controls. As a result you cannot adjust both the size and aspect ratio of the image.

The optimum position for these controls is on the front panel where you can both adjust them and view the image at the same time. Controls on the rear panel require you to have gorilla-like arms to reach around the monitor to make adjustments while checking their effect.

Image controls come in two types, analog and digital.

Analog controls are the familiar old knobs like you find on vintage television sets. Twist one way and the image gets bigger; twist the other and it shrinks. Analog controls have one virtue—just by looking at the knob you know where they are set, whether at one or the other extreme of their travel. The control itself is a simple memory system; it stays put until you move it again. Analog controls, however, become dirty and wear out with age, and they usually enable you to set but one value per knob—one value that must cover all the monitor's operating modes.

Digital controls give you pushbutton control over image parameters. Press one button, and the image gets larger or moves to the left. Another compensates in the opposite direction. Usually digital controls are linked with a microprocessor, memory, and mode sensing circuitry so that you can pre-set different image heights and widths for every video standard your monitor can display.

Digital controls don't get noisy with age and are more reliable and repeatable, but you never know when you are approaching the limit of their travel. Most have two speed operation—hold them in momentarily and they make minute changes; keep holding down the button and they shift gears to make gross changes. Of course, if you don't anticipate the shift, you'll overshoot the setting you want and spend a few extra moments zeroing in on the exact setting.

Size and position controls are irrelevant to LCD and similar alternate display technologies. LCD panels are connected more directly to display memory so that memory locations correspond nearly exactly to every screen position. There's no need to move the image around or change its shape because it's forever fixed where it belongs.

Most CRT-based displays also carry over several controls from their television progenitors. Nearly every computer monitor has a brightness control, which adjusts the level of the scanning electron beam; this in turn makes the onscreen image glow brighter or dimmer. The contrast control adjusts the linearity of the relationship between the incoming signal and the onscreen image brightness. In other words, it controls the brightness relationship that results from different signal levels—how much brighter high intensity is. In a few displays, both the brightness and contrast function are combined into a single "picture" control. Although a godsend to those who might get confused by having to twiddle two knobs, the combined control also limits your flexibility in adjusting the image to best suit your liking.

Other controls ubiquitous to televisions usually are absent from better computer monitors because they are irrelevant. Vertical hold, color (saturation), and hue controls only have relevance to composite video signals, so they are likely only to be found on composite interfaced displays. The vertical hold control tunes the monitor to best decipher the vertical synchronizing signal from the ambiguities of composite video signals. The separate sync signals used by other display standards automatically remove any ambiguity. Color and hue only adjust the relationship of the color subcarrier to the rest of the composite video signal and have no relevance at all to non-composite systems.

Electronics

The image you see onscreen is only part of the story of a complete display system. The video signals from your PC must be amplified and processed by the electronics inside the monitor to achieve the right strength and timing relationships to put the proper image in view.

The basic electronic components inside a monitor are its video amplifiers. As the name implies, these circuits simply increase the strength (amplify) of the approximately one-volt signals they receive from your PC to the thousands of volts needed to drive the electron beam from cathode to phosphor. Monochrome monitors have a single video amplifier; color monitors, three (one for each primary color).

In an analog color monitor, these three amplifiers must be exactly matched and absolutely linear. That is, the input and output of each amplifier must be precisely proportional, and it must be the same as the other two amplifiers. The relationship between these amplifiers is called color tracking. If it varies, the color of the image on the screen won't be what your software had in mind.

The effects of such poor color tracking are all bad. You lose precision in your color control. This is especially important for desktop publishing and presentation applications. With poor color tracking, the screen can no longer hope to be an exact preview of what eventually appears on paper or film. You may even lose a good fraction of the colors displayable by your video system.

What happens is that differences between the amplifiers cause one of the three primary colors to be emphasized at times and de-emphasized at others, casting a subtle shade on the onscreen image. This shading effect is most pronounced in gray displays—the dominant color(s) tinge the gray.

Although you don't have to worry about color tracking in a monochrome display, the quality of the amplifier nevertheless determines the range of grays that can be displayed. Aberrations in the amplifier cause the monitor to lose some of its grayscale range.

The relationship between the input and output signals of video amplifiers is usually not linear. That is, a small change in the input signal may make a greater than corresponding change in the output. In other words, the monitor may exaggerate the color or grayscale range of the input signal—contrast increases. The relationship between input and output is referred to as the gamma of the amplifier. A gamma of one would result in an exact correspondence of the input and output signals. However, monitors with unity gammas tend to have washed out, pastel images. Most people prefer higher gammas, in the range 1.5 to 1.8, because of their contrastier images.

If you want a monitor that works both with today's VGA, SuperVGA, and sharper display systems as well as old graphics adapters like CGA and EGA, you need a monitor with TTL capability. Most modern computer monitors have only analog inputs, which cannot properly display the digital (TTL) signals used by these older standards. You also have to check the color support of such monitors. To be compatible with the CGA standard, a monitor must be capable of handling a 16-color digital input, sometimes called RGBI for the four (red, green, blue, and intensity) signals from which it is made. EGA compatibility requires 64-color capability.

Synchronizing Frequency Range

Today's variety of signal standards makes it almost mandatory that your monitor be able to synchronize to a wide range of synchronizing frequencies. You have two frequencies to worry about. Vertical frequency, sometimes called the refresh rate or frame rate, determines how often the complete screen is updated. The horizontal synchronizing frequency (or horizontal scan rate) indicates the rate at which the individual scan lines that make up the image are drawn.

These frequency ranges are important to you because they determine with which video standards the monitor can work. Most monitors start at the 31.5 KHz used by the VGA system. If you need a monitor compatible with older display standards, you'll need even lower ranges. The CGA system requires a horizontal frequency of 15.75 KHz; MDA, 18 KHz; and EGA, 22 KHz. Usually the high end of the horizontal frequency range is the more troublesome. In general the higher the resolution you want to display, the higher the horizontal frequency.

The lowest frame rate normally required is the 59 Hz used by some early VESA modes, although the old MDA standard requires a 50 Hz frame rate. The highest frame rates are the 85 Hz signals used by some new VESA standards. Table 17.3 lists the scanning frequencies of most common computer display systems.

Note that the 8514/A and XGA systems have very low frame rates, 44 Hz, because they are interlaced systems (see the following "Interlacing" section). To properly display these signals a monitor must have sufficient range to synchronize with the field rate of these standards, which is twice the frame rate.

Interlacing

Interlaced systems like the 8514/A and the first implementations of XGA used a trick developed for television to help put more information onscreen using a limited bandwidth signal. Instead of scanning the image from top to bottom, one line after another, each frame of the image is broken into two fields. One field consists of the odd numbered lines of the image; the other the even numbered lines. The electron beam sweeps across and down, illuminating every other line and then starts from the top again and finishes with the ones it missed on the first pass.

This technique achieves an apparent doubling of the frame rate. Instead of sweeping down the screen 30 times a second (the case of a normal television picture), the top to bottom sweep occurs 60 times a second. Whereas a 30 frame per second rate would noticeably flicker, the ersatz 60 frame per second rate does not—at least not to most people under most circumstances. Some folks' eyes are not fooled, however, so interlaced images have earned a reputation of being flickery. Figure 17.8 shows how an interlace monitor is scanned.

Figure 17.8 Progressive versus interlaced scanning.

Interlacing is used on computer display signals to keep the necessary bandwidth down. A lower frame rate lowers the required bandwidth of the transmission channel. Of all the prevailing standards, only the original high resolution operating mode of the 8514/A display adapter and the first generation of XGA use interlacing. Their frame rate, 44 Hz, would cause distinct flicker. Interlacing drives the field rate up to 88 Hz.

Bandwidth

Perhaps the most common specification usually listed for any sort of monitor is bandwidth, which is usually rated in megahertz. Common monitor bandwidths stretch across a wide range—figures from 12 to 100 megahertz are sometimes encountered.

In theory, the higher the bandwidth, the higher the resolution and sharper the image displayed. In the case of color displays, the dot pitch of the display tube is the biggest limit on performance. In a monochrome system, however, bandwidth is a determinant of overall sharpness. The PC display standards do not demand extremely wide bandwidths. Extremely large bandwidths are often superfluous.

The bandwidth necessary in a monitor is easy to compute. A system ordinarily requires a bandwidth wide enough to address each individual screen dot plus an extra margin to allow for retrace times. (Retrace times are those periods in which the electron beam moves but does not display, for instance at the end of each frame when the beam must move from the bottom of the screen at the end of the last line of one frame back up to the top of the screen for the first line of the next frame.)

A typical color display operating under the VGA standard shows 288,000 pixels (a 729 by 400 pixel image in text mode) 70 times per second, a total of 20.16 million pixels per second. An 800 by 600 pixel Super VGA display at 75 Hz must produce 36 million pixels per second.

Synchronizing signals require their own slice of monitor bandwidth. Allowing a wide margin of about 25 percent for retrace times, it can be seen that for most PC applications, a bandwidth of 16 megahertz is acceptable for TTL monitors, and 10 megahertz of bandwidth is sufficient for sharp composite video displays, figures well within the claims of most commercial products. For VGA, 25 megahertz is the necessary minimum.

Multiplying the dot clock by 25 percent yields an acceptable estimate of the bandwidth required in a monitor. For the standards IBM promulgated, the company listed actual bandwidth requirements. Table 17.4 summarizes these requirements and calculates estimates for various PC display standards.

Although the above estimates are calculated using the dot clock of the display, the relationship between dot clock and bandwidth is not as straightforward as the calculations imply. Because in real world applications the worst case display puts an illuminated pixel next to a dark one, the actual bandwidth required by a display system should be the dot clock plus system overhead. A pair of on/off pixels exactly corresponds to the up and down halves of a single cycle. The higher bandwidth you calculate from the dot clock allows extra bandwidth that gives greater image sharpness—sharply defining the edges of a pixel requires square waves, which contain high frequency components. Consequently the multiplication of the dot clock by display overhead offers a good practical approximation of the required bandwidth, even though the calculations are on shaky theoretical ground.

Energy Star

Compared to some of the things that you might connect to your PC, a monitor consumes a modest amount of electricity. A laser printer can draw as much as a kilowatt when its fuser is operating. A typical PC requires 100 to 200 watts. A typical monitor requires only about 30 watts. Unlike the laser fuser, however, your monitor may stay on all day long, and an office may have hundreds of them, each continually downing its energy dosage. Those watts add up, not just in their power consumption but also their heat production that adds to the load on the office air conditioning.

To help cut the power used by computer equipment, the Environmental Protection Agency started the Energy Star program, which seeks to conserve power while computers and their peripherals are in their idle states. For monitors, Energy Star means that the monitor powers down to one of two lower power conditions or shuts off entirely when its host computer has not been used for a while.

Energy Star compliant monitors have four operating modes: on, standby, suspend, and off. During normal operation when the monitor displays an active image generated by your PC, it is on. In standby mode, the monitor cuts off the electron beam in its CRT and powers down some of its electronics. It keeps the filament or heat of the CRT (the part that has to warm up to make the tube work) hot so that the monitor can instantly switch back to its on state. In suspend mode, the filament and most of the electronics of the monitor switch off. Only a small portion of the electronics of the monitor remain operational to sense the incoming signals, ready to switch the monitor back on when the need arises. This conserves most of the power that would be used by the monitor but requires the CRT to heat up before the monitor can resume normal operation. In other words, the monitor trades rapid availability for a reduction in power use. In off mode, the monitor uses no power but requires you to manually switch it on.

To enable your PC to control the operating mode of your monitor without additional electrical connections, VESA developed its Display Power Management Standard. This system uses the two synchronizing signals your video board supplies to your monitor to control its operating mode. To signal the monitor to switch to standby operation, your video card switches off only its horizontal sync signal. To signal the monitor to switch to suspend mode, your video board cuts both the video signal and the vertical sync but leaves the horizontal sync on. In off mode, all signals are cut off. Table 17.5 summarizes these modes.

Advanced operating systems monitor your system usage and send out the standby and/or suspend signals when you leave your system idle for a pre-determined time. Your video driver software controls the DPMS signals. Note that screen saver software defeats the purpose of DPMS by keeping your system active even when it is not in use. You can trim your power usage by relying on DPMS rather than a screen saver to shut down your monitor.

Many monitors made before the DPMS standard was created often incorporate their own power saving mode that's initiated by a loss of the video signal for a pre-determined time. In other words, they sense when your screen is blank and start waiting. If you don't do something for, say, five minutes, they switch the monitor to standby or suspend no matter the state of the synchronizing signals. The DPMS system is designed so that these monitors, too, will power down (although only after the conclusion of both the DPMS and their own internal timing cycles.)

To enable the DPMS system under Windows 95, you must tell the operating system that your monitor is Energy Star compliant. If your monitor is in the Windows 95 database, Windows will know automatically whether the monitor is compliant. You can also check the Energy Star compliance in the Change Display Settings screen. You must also activate the DPMS screen blanker from the Screen Saver tab of your Display Properties screen.

Monitor Types

VGA displays were introduced by necessity with the PS/2s. They use analog inputs and a 31 KHz horizontal synchronizing frequency to match with the VGA standard. VGA is now the minimum you should demand in a computer monitor.

Identification

The prevalence of multi-scanning monitors with wildly different capabilities makes getting the most from your investment a challenge. You must be able to identify not only the display abilities of the monitor but also those of your video board, then make the best possible match between them. If you're wrong, you won't get everything that you've paid for. Worse, you might not see anything intelligible on your screen. Worse still, you face a tiny chance of actually harming your monitor or video board.

Hard Wired Coding

This rudimentary standard was not up to the task of identifying the wide range of monitor capabilities that became available in the years after the introduction of VGA. Yet adding true Plug-and-Play capabilities to your PC requires automatically identifying the type of monitor connected to your PC so that the display adapter (and the rest of the system) can be properly configured. To meet this challenge, VESA developed the Display Data Channel, an elaborate monitor identification system based on the same connections as the early IBM system but with greatly enhanced signals and capabilities.

Through the DDC, the monitor sends an Extended Display Identification or EDID to your PC. In advanced form the DDC moves data both ways between your monitor and your PC using either the I²C or ACCESS.bus serial interfaces discussed in Chapter 21, "Serial Ports." The DDC2B standard uses I²C bus signaling on two of the wires of the monitor connection to transfer data both ways between the monitor and its attached video board. DDC2AB uses a full ACCESS.bus connection and protocol which allow you to connect other computer peripherals (for example, your keyboard) to your monitor rather than the system unit.

All levels of the DDC system gain information about your monitor in the same way. The monitor sends out the EDID as a string of serial bits on the monitor data line, pin 12. Depending on the level of DDC supported by the monitor, the EDID data stream is synchronized either to the vertical sync signal generated by the video board present on pin 14 of 15-pin video connectors or to a separate serial data clock (SCL) that's on pin 15 in DDC2 systems. One bit of data moves with each clock cycle. When the system uses vertical sync as the clock, the data rate will be in the range from 60 to 85 Hz. With DDC-compliant monitors, your video board can temporarily increase the vertical sync frequency to up to 25 KHz to speed the transmission of this data. Using the SCL signal when both video board and monitor support it, data rates as high as 100 KHz are possible.

The serial data takes the form of nine-bit sequences, one per byte. The first eight bits encode the data, most significant bit first. The last bit can be either a zero or one at the choice of the monitor manufacturer. The only restriction is that the ninth bit must have the same value for every byte.

The DDC system sends data from the monitor to your display adapter in 128 byte blocks. The first of these is the Extended Display Identification or EDID block. It is optionally followed by an Extended EDID block or additional proprietary manufacturer data blocks. Table 17.7 lists the structure of the basic EDID.

The header is always the same data pattern and serves to identify the EDID information stream. The vendor identification is based on EISA manufacturer identifications. The product identification is assigned by the manufacturer. It includes the month and year of manufacture of the monitor.

The basic display parameters that EDID relays to your system include the maximum size of your monitor's display expressed as the largest width and height of the image area. Your applications or operating system can use this information to automatically set the proper scaling for fonts displayed on the screen. The timing data includes a bit representing the ability to support each of the various VESA standards so that your system can determine the possible modes and frequencies your monitor can use.

In addition to basic DDC support, VESA provides for two higher levels of standardization. The DDC2B system uses the Philips I2C signaling system to transfer data bi-directionally across the interface. The DDCAB system includes full ACCESS.bus support that supplies a low speed serial interconnection bus suitable for linking such peripherals as keyboards and pointing devices through the monitor.

Because standard expansion buses do not provide suitable connections for routing ACCESS.bus signals, VESA has defined an auxiliary connector for the purpose, a five-pin "Berg" connector. Table 17.8 lists the signal assignments of this connector.

Monitors that are compliant with DDC use the same connectors as ordinary VGA displays. All the active video and synchronizing signals are located on the same pins of the connector no matter the DDC level the monitor uses, or even if it doesn't use DDC at all. The only difference is the definition of the monitor identification signals. DDC1 video boards sacrifice the Monitor ID Bit 1 pin, number 12, as a channel to receive identification data from your monitor. DDC2 systems make this signal bi-directional and take over pin 15 for use in carrying the clock signal. In any case, pin 9 may be used to supply five volts for running accessory devices. Table 17.9 lists the signal assignments of the VGA 15-pin connector under DDC.

Manual Configuration

If your monitor or video board does not support any level of the DDC specification, you will be left to configure your system on your own. In general, you won't need to perform any special configuration to a multi-scanning monitor to match it to your video board if the signals from the video board are within the range of the monitor. That's the whole point of the multi-scanning display—it adjusts itself to accommodate any video signal.

That said, you may not get the most from your monitor. You might slight on its refresh rate and quickly tire your eyes with a flickering display. Worse, you might exceed its capabilities and end up with a scrambled or blank screen.

Windows includes its own configuration process that attempts to optimize the signals of your compliant video adapter with your monitor type. Windows already knows what video board you have—you have to tell it when you install your video drivers. Without a DDC connection, however, Windows is cut off from your monitor, so you must manually indicate the brand and model of monitor you're using.

You make these settings by clicking on the Change Display Type button in the Setting tab of your Display properties menu. Click on the button, and you see a screen like that shown in Figure 17.9.

Figure 17.9 The Change Display Type screen in Windows 95.

Flat Panel Display Systems

CRTs are impractical for portable computers, as anyone who has toted a forty pound, first generation portable computer knows. The glass in the tube itself weighs more than most of today's portable machines, and running a CRT steals more power than most laptop or notebook machines budget for all their circuitry and peripherals.

LCD

The winner in the display technology competition was the Liquid Crystal Display, the infamous LCD. Unlike LED and gas-plasma displays, which glow on their own, emitting photons of visible light, LCDs don't waste energy by shining. Instead, they merely block light otherwise available. To make patterns visible, they either selectively block reflected light (reflective LCDs) or the light generated by a secondary source either behind the LCD panel (backlit LCDs) or adjacent to it (edgelit LCDs). The backlight source is typically an electroluminescent (EL) panel, although some laptops use Cold Cathode Fluorescent (CCF) for brighter, whiter displays with the penalty of higher cost, greater thickness, and increased complexity.

Nematic Technology

A number of different terms describe the technologies used in the LCD panels themselves, terms like supertwist, double supertwist, and triple supertwist. In effect, the twist of the crystals controls the contrast of the screen, so triple supertwist screens have more contrast than ordinary supertwist.

The history of laptop and notebook computer displays has been led by innovations in LCD technology. Invented by RCA in the 1960s (General Electric still receives royalties on RCA's basic patents), LCDs came into their own with laptop computers because of their low power requirements, light weight, and ruggedness.

An LCD display is actually a sandwich made from two plastic sheets with a very special liquid made from rod-shaped or nematic molecules. One important property of the nematic molecules of liquid crystals is that they can be aligned by grooves in the plastic to bend the polarity of light that passes through them. More importantly, the amount of bend the molecules of the liquid crystal give to the light can be altered by applying an electrical current through them.

Ordinary light has no particular orientation, so liquid crystals don't visibly alter it. But polarized light aligns all the oscillations of its photons in a single direction. A polarizing filter creates polarized light by allowing light of a particular polarity (or axis of oscillation) to pass through. Polarization is key to the function of LCDs.

To make an LCD, light is first passed through one polarizing filter to polarize it. A second polarizing filter, set to pass light at right angles to the polarity of the first, is put on the other side of the liquid crystal. Normally, this second polarizing filter stops all light from passing. However, the liquid crystal bends the polarity of light emerging from the first filter so that it lines up with the second filter. Pass a current through the liquid crystal and the amount of bending changes, which alters in turn the amount of light passing through the second polarizer.

To make an LCD display, you need only selectively apply current to small areas of the liquid crystal. The areas to which you apply current are dark; those that you don't, are light. A light behind the LCD makes the changes more visible.

Over the past few years, engineers have made several changes to this basic LCD design to improve its contrast and color. The basic LCD design outlined above is technically termed twisted nematic technology or TN. In their resting state, the liquid molecules of the TN display always bend light by 90 degrees, exactly counteracting the relationship between the two polarizing panels that make up the display.

By increasing the bending of light by the nematic molecules, the contrast between light and dark can be increased. An LCD design that bends light by 180 to 270 degrees is termed a supertwist nematic or simply supertwist display. One side effect of the added twist is that the appearance of color artifacts results in the yellowish green and bright blue hues of many familiar LCD displays.

This tinge of color can be canceled simply by mounting two supertwist liquid crystals back to back so that one bends the light in the opposite direction of the other. This design is logically termed a double supertwist nematic or simply double supertwist) display. This LCD design is currently popular among laptop PCs with black and white VGA-quality displays. It does have a drawback, however. Because two layers of LCD are between you and the light source, double supertwist panels appear darker or require brighter backlights for adequate visibility.

Triple supertwist nematic displays instead compensate for color shifts in the supertwist design by layering both sides of the liquid crystal with thin polymer films. Because the films absorb less light than the twin panels of double supertwist screens, less backlight—and less backlight power—is required for the same screen brightness.

Cholesteric Technology

In 1996, the Liquid Crystal Institute of Kent State University developed another liquid crystal technology into a workable display system and began its commercial manufacture. Termed cholesteric LCDs, this design uses crystals that switch between transmissive and reflective states instead of twisting. These changes are more directly visible and require no polarizers to operate. In that polarizing panels reduce the brightness of nematic displays by as much as 75 percent, cholesteric LCDs can be brighter. Early screens are able to achieve high contrast ratios without backlights.

Cholesteric screens have a second advantage. They are bi-stable. That is, maintaining a given pixel in either the transmissive or reflective phase requires no energy input. Once switched on, a pixel stays on until switched off. The screen requires power only to change pixels. In fact, a cholesteric screen will retain its last image even after it is switched off. Power usage in notebook PC applications is likely to be as low as 10 percent that of nematic panels.

The fabrication technologies used to make the cholestric displays also allow for final detail. Kent State has already demonstrated grayscale panels with resolutions as high as 200 pixels per inch. Although initial production was limited to grayscale displays, color cholestric panels are currently under development.

Passive Matrix

Nematic LCDs also come in two styles based on how the current that aligns their nematic molecules is applied. Most LCD panels have a grid of horizontal and vertical conductors, and each pixel is located at the intersection of these conductors. The pixel is darkened simply by sending current through the conductors to the liquid crystal. This kind of display is called a passive matrix.

Active Matrix

The alternate design, the active matrix, is more commonly referred to as Thin Film Transistor (TFT) technology. This style of LCD puts a transistor at every pixel. The transistor acts as a relay. A small current is sent to it through the horizontal and vertical grid, and in response the transistor switches on a much higher current to activate the LCD pixel.

The advantage of the active matrix design is that a smaller current needs to traverse the grid, so the pixel can be switched on and off faster. Whereas passive LCD screens may update only about half a dozen times per second, TFT designs can operate at ordinary monitor speeds—ten times faster. That increased speed equates to faster response—for example, your mouse won't disappear as you move it across the screen.

The disadvantage of the TFT design is that it requires the fabrication of one transistor for each screen pixel. Putting those transistors there requires combining the LCD and semiconductor manufacturing processes. That's sort of like getting bricklayers and carpenters to work together.

To achieve the quality of active matrix displays without paying the price, engineers have upped the scan on passive panels. Double scanned passive works exactly like the name says: they scan their screens twice in the period that a normal screen is scanned only once. Rather than go over each pixel two times, the electronics of a double-scanned display divides the screen into two halves and scans both at the same time. The idea is something like the interlacing of CRT screens, lowering the required scanning frequency, but the arrangement and effect are different. Double scanned displays split the screen in the middle into upper and lower halves. The split means that each pixel gets twice as long for updates, as would be the case if the whole screen were scanned at the same frequency. As a result, double scanning can eke out extra brightness, contrast, and speed. They do not, however, reach the quality level set by active matrix screens.

Response Time

The LCD panel equivalent of persistence is response time. Charging and discharging individual pixels requires a finite period, and the response time measures this period. The time to charge and the time to discharge a given pixel can be, and often are, different, and are typically individually specified. For example, the off times of some active screens may be twice that of the on time.

The ambient temperature can have dramatic effects on the response time of an LCD panel. At freezing, the response time of a panel may be three times longer (slower) than at room temperature.

At room temperature, an active matrix display pixel has a response time on the order of 10-50 milliseconds.

Field Emission Displays

The chief challenger to the Liquid Crystal Display is the Field Emission Display or FED. Several manufacturers are actively developing FED technology. Commercial FED panels were offered in early 1996, although with smaller dimensions than required for PC displays. FED displays are nevertheless expected in new notebook PCs, with 10-inch panels entering commercial production in late 1997. Manufacturers are also developing small (2.5-inch) panels for hand held televisions and large panels (40 inches) for wall hung displays.

In a radical bit of retro design, the FED uses the same basic illumination principle as the cathode ray tube. A flux of electrons strikes phosphor dots and causes them to glow. As with the CRT, the electrons must flow through a vacuum, so the FED panel is essentially a flattened vacuum tube. Instead of a single electron gun for each color, however, the FED uses multiple, microscopic cones as electron emitters. Several hundred of these cathode emitters serve each image pixel. Each group of emitters has its own drive transistor, much like an active matrix LCD panel. Each emitter is cone shaped, a configuration which favors electron emission.

At the other side of the panel, each pixel has a conventional trio of phosphor dots, one for each primary color. Each dot has associated with it a separate transparent anode that attracts the electron flux. To separate the three colors of the pixel, the drive electronics activate each of the three anodes in sequence so the total display time of the pixel is split in thirds, one-third for each color. The whole assembly fits between two glass panels, which form the bottle of the vacuum tube. Figure 17.10 illustrates the construction used by one FED manufacturer.

Figure 17.10 Cross-section of a Field Emission Display.

The short travel of the electron flux in the FED dramatically reduces the voltage required for the operation of the device. The anode in a FED may be only about 200 microns from the nearest cathode. Instead of a potential of thousands of volts between the anode and cathode, the FED operates at about 350 volts. The FED is essentially a current device, relying on a high current in the form of a dense flux of electrons to provide sufficient illumination. This high current presents one of the major technological obstacles to designing successful FEDs. Conventional phosphors deteriorate rapidly with the onslaught of electrons, quickly and readily burning in.

A FED behaves more like a CRT than an LCD. Its electron currents respond quickly to changes in their drive voltages. The response time of a FED display is typically a few microseconds compared to the millisecond responses of LCD panels. FEDs also have wider viewing angles than LCD panels. The image on a FED screen is viewable over an angle of about 160 degrees, much the same as a CRT. In addition, FED technology promises to be more energy efficient than LCDs, capable of delivering about the same screen brightness while using half the power.

Electro-Luminescent Displays

The backlights used by many LCD displays use electro-luminescent technology. Some manufacturers are working to develop flat panel displays that do away with the LCD and instead use an EL panel segmented into individual pixels. Monochrome screens have already been developed, but color screens are problematic. Although green and blue EL elements have operating lives long enough for commercial applications, in excess of 5,000 hours, current red EL elements operate only about half as long. The longer wavelength of red light requires higher currents, which shorten the life of the materials. Manufacturers believe, however, that the technology will be successful in the next few years. Its primary application is expected to be wall hung displays.

Gas-Plasma

One alternative is the gas-plasma screen, which uses a high voltage to ionize a gas and cause it to emit light. Most gas plasma screens have the characteristic orange-red glow of neon because that's the gas they use inside. Gas-plasma displays are relatively easy to make in the moderately large sizes perfect for laptop computer screens and yield sharpness unrivaled by competing technologies. However, gas-plasma screens also need a great deal of power—several times the requirements of LCD technology—at high voltages, which must be synthesized from low voltage battery power. Consequently, gas-plasma displays are used primarily in AC-power portables. When used in laptops, the battery life of a gas-plasma equipped machine is quite brief, on the order of an hour.

LED

In lieu of the tube, laptop designers have tried just about every available alternate display technology. These include the panels packed with light-emitting diodes—the power-on indicators of the 1980s—that glow at you as red as the devil's eyes. But LEDs consume extraordinary amounts of power. Consider that a normal, full size LED can draw 10 to 100 milliwatts at full brilliance and that you need 100,000 or so individual elements in a display screen; you get an idea of the magnitude of the problem. Certainly the individual display elements of an LED screen would be smaller than a power-on indicator and consume less power, but the small LED displays created in the early days of portable PCs consumed many times the power required by today's technologies. LEDs also suffer the problem that they tend to wash out in bright light and are relatively expensive to fabricate in large arrays.

Practical Considerations

When you buy a notebook PC, you're usually stuck with the flat panel display system that the manufacturer chooses to install. Because everything you do with your PC depends on what you see on your screen, the display system is one of the most important factors in selecting a notebook PC. In addition to the underlying technology, two other characteristics distinguish flat panel display systems: resolution and size. Although the terminology is the same as for CRT-based monitors, the flat panel systems raise different issues.

Resolution

Onscreen resolution is an important issue with flat panel displays; it determines how sharp text characters and graphics will appear. Today, three resolution standards are dominant: CGA (640 x 200); double scanned CGA (640 x 400); and VGA (640 x 480).

Most people prefer the last because it's exactly equivalent to today's most popular desktop displays so it can use the same software and drivers.

CGA resolution is visibly inferior, producing blocky, hard to read characters, and remains used only in the least expensive laptops.

Double scanned CGA offers a good compromise between cost and resolution. It's actually as sharp as text mode CGA. Graphics pose a problem, as double scanned CGA mode is not supported by a wide base of software. Under Windows, however, many double scanned CGA systems are compatible with Toshiba and AT&T 640 x 400 pixel drivers.

VGA poses particular problems for LCD displays because it's really more three standards under a single name, operating with modes that put 350, 400, or 480 lines on the screen. Most VGA display panels have 480 rows of dots to accommodate these lines.

A problem develops with VGA images made with lower line counts. Many old notebook LCD screens displayed only the active lines and left the rest of the screen blank. For example, a panel would leave 80 lines blank during 400-line text mode displays on a 480-line LCD. The result was a black band at both the top and bottom of the screen.

Today's notebook PCs usually avoid this problem and fill the entire screen no matter what mode its video system operates in. Two different techniques adjust the image size, depending on whether the panel operates in text or graphics mode.

VGA text mode uses 400 lines and would otherwise leave 80 lines on the screen blank. To sidestep this problem, flat panel controllers either use a taller font or insert blank lines between lines of normal text fonts. Each technique has its drawbacks but overall either is more pleasing than the otherwise blank areas on the screen. Substituting fonts may result in compatibility problems with software that assumes a given font size (which, of course, it should not in text mode). Blank line insertion results in squat characters that appear widely spaced (in print, the effect is called leading). Block graphics may appear disjointed with vertical lines turning into dashes.

In 350-line graphics mode, flat panel controllers use line replication to duplicate a sufficient number of existing image lines to compensate for the blank area of the screen. In effect, line replication is an exotic form of the double scanning that expands 200-line graphics into 400-line displays. The math resulting from stretching a 350-line image to fill a 480-line screen requires a specific pattern of repeated and unrepeated lines. In text mode, this can make adjacent lines of text appear to be different sizes, so the technique is usually reserved for graphics.

With text mode and EGA-style graphics left far behind by operating systems with integrated VGA and better resolution, the need for both of these technologies is disappearing. However, if you ever need to step back and run older software, you'll appreciate a display system that can cope well with various image heights.

Size

Unlike with CRT-based monitors, the entire area of an LCD display is usable. Because the image is a perfect fit into the pixels of the LCD display, there are no issues of underscan and overscan. The size of the screen is the size of the image.

The VESA Video Image Area Definition standard applies to LCDs just as it does CRTs. Most LCD panel makers, however, specify the dimensions of their products with the same diagonal that served manufacturers so well for color televisions. Common sizes range from 10.4 inches to 13.3 inches, with even larger sizes under development.

Because most LCD panels are used on notebook PCs, their size is ultimately constrained by the dimensions of the computer itself. For notebook machines that approximate the size of a stack of true notepads, the maximum screen size is about 12.1 inches. Some manufacturers (notably NEC) have developed notebook PCs with larger screens, but the computers themselves are necessarily larger in length and width. Larger LCD panels are slated to replace CRTs as the basis for desktop and, in a new application, wall hung displays. Some manufacturers are experimenting with screens as large as 42 inches across for such applications.

Labeling

The final figures indicate the addressability of the screen listed as the number of horizontal pixels and vertical pixels. In that the addressability of most LCD screens extends only to all the visible pixels, this figure also represents the visible resolution of the screen. Note that this nomenclature describes only the electrical characteristics of the panel and does not extend to physical characteristics like size. Panels with identical designations will plug into the same electronics even if they are different physical sizes.

Connectors

Monitors can be grouped by the display standard they support, mostly based upon the display adapter card they are designed to plug into. One basic guide that helps you narrow down the compatibility of a display just by inspecting its rear panel is the input connector used by the monitor. After all, if you cannot plug a monitor into your computer, odds are it is not much good to you.

Three styles of connectors are shared by different PC video standards. By name, these three connectors are the RCA-style pin jack, the 9-pin D-shell, and the 15-pin "high-density" D-shell. In addition, some high resolution monitors use three or more BNC connectors for their input signals. In addition, VESA has created a new connector standard, the Enhanced Video Connector, that combines video with a host of other signals so that a single connection can link just about everything to your PC.

Video

The video connection carries the image between your video board and monitor. Several standards are used by computer monitors. These vary with the signals used by the monitor and the applications it serves.

Video systems based on composite video signals commonly use pin jacks. These signals allow you to connect multiple monitors together.

Computer displays most commonly use systems with D-shell connectors. Each display standard has its own distinct arrangement of signals. All computer display systems are designed to connect a single monitor to a given output.

Multimedia systems often combine the need for both as well as audio. A new connector system called the Enhanced Video Connector promises to put all the required multimedia signals in a single connector.

Pin Jacks

The bull's eye jack used on stereo and video equipment is used by most manufacturers for the composite video connections in PC display systems, although a wealth of monitors and television sets made by innumerable manufacturers also uses this connector. This connector does give you many choices for alternate displays—that is, if you don't mind marginal quality.

Composite monitors (those dealing with the composite video and NTSC color only) rank among the most widely available and least expensive in both color and monochrome. Even better quality television sets have such jacks available. Figure 17.12 illustrates a typical pin jack.

Figure 17.12 A jack for video pin connectors.

Although you can use any composite video display with a CGA or compatible color card, the signal itself limits the possible image quality to okay for monochrome, acceptable for 40-column color, and unintelligible for 80-column color. Nevertheless, a composite video display—already a multipurpose device—becomes even more versatile with a computer input.

Daisy Chaining

A side benefit of pin plug/composite video displays is that most have both input and output jacks. These paired jacks enable you to daisy chain multiple monitors to a single video output. For example, you can attach six composite video monitors to the output of your computer for presentations in classroom or boardroom.

In many cases, the jacks just loop through the display (that is, they connect together). The display merely bridges the input video signal and alters it in no other manner. You can connect a nearly unlimited number of monitors to these loop-through connections with no image degradation. Some monitors, however, buffer their outputs with a built-in video amplifier. Depending on the quality of the amplifier, daisy chaining several of these monitors can result in noticeable image degradation.

One way to tell the difference is by plugging the output of the display into the output of your computer. Most amplifiers don't work backwards, so if the display has a buffering amplifier nothing appears onscreen. If you do get an image comparable to the one you get when plugging into the input jack, the signal just loops through the display.

Analog Voltage Level

The specifications of composite monitors sometimes include a number describing the voltage level of the input signal. This voltage level can be important when selecting a composite display because all such monitors are essentially analog devices.

In analog monitors, the voltage level corresponds to the brightness the electron beam displays onscreen. A nominal one-volt peak to peak input signal is the standard in both the video and computer industries and should be expected from any composite monitor. The VGA system requires a slightly different level—0.7 volts.

Termination

For proper performance, a composite video signal line must be terminated by an impedance of 75 ohms. This termination ensures that the signal is at the proper level and that aberrations do not creep in because of an improperly matched line. Most composite input monitors (particularly those with separate inputs and outputs) feature a termination switch that connects a 75-ohm resistor across the video line when turned on. Only one termination resistor should be switched on in any daisy chain, and it should always be the last monitor in the chain.

If you watch a monitor when you switch the termination resistor on, you'll notice that the screen gets dimmer. That's because the resistor absorbs about half the video signal. Because composite video signals are analog, they are sensitive to voltage level. The termination cuts the voltage in half and, consequently, dims the screen by the same amount. Note that the dim image is the proper one. Although bright might seem better, it's not. It may overload the circuits of the monitor or otherwise cause erratic operation.

Composite monitors with a single video input jack and no video output usually have a termination resistor permanently installed. Although you might try to connect two or more such monitors to a single CGA composite output (with a wye cable or adapter), doing so is unwise. With each additional monitor, the image gets dimmer (the signal must be split among the various monitors) and the CGA adapter is required to send out increasing current. The latter could cause the CGA to fail.

Nine-Pin D-Shell Connectors

The first video connector used by PCs was the nine-pin D-shell connector. As with all D-shell connectors, its shape assures that all signal pins will be properly matched with their mates. Its nine pins provide enough signal positions for both monochrome and color systems. Figure 17.13 illustrates a 9-pin D-shell jack.

Figure 17.13 A nine-pin D-shell jack.

All PC digital display standards and one specialized analog standard used this connector design. These standards include: monochrome, standard RGB (CGA), and enhanced RGB (EGA), and professional RGB (PGA). In addition, some early multiscanning monitors used this connector for their input signals. All of these standards arrayed their signals differently on the connector pins.

Although this connector has fallen into disuse, it remains troublesome or troubling because of the huge potential for confusion it presents. You must follow one important rule when you approach this connector—know what kind of display adapter you are about to plug into. Making the wrong choice can be fatal to your display, particularly if you try to plug an IBM Monochrome Display into a CGA adapter. The mismatch of synchronizing frequencies leads to the internal components of the display overheating and failing.

A mismatch is easy to spot—you simply can't make sense of the image on the screen. You may see a Venetian-blind pattern of lines; the screen may flash; or it may look like the vertical hold failed in a dramatic way. Should you observe any of these patterns or hear a high-pitched squeal from your display and see nothing onscreen, immediately turn off your display. Hunt for the problem while the life of your monitor is not ticking away.

Monochrome Display Adapter

The video signal of the MDA system is digital. The only thing about it that matters to a monitor is whether voltage is present or not. MDA monitors ignore the voltage level of the video signal. This video signal will also drive analog monitors so that you can see monochrome images on them. However the signal strength is substantially higher than normal video signals (5-volt TTL level instead of 1-volt video). Moreover, because the intensity signal appears on a separate pin, highlighting will not appear in such a connection. Note, too, the MDA system uses separate sync with distinct pins for its horizontal and vertical synchronizing signals, and so requires a monitor that accommodates separate sync. Table 17.10 lists the signal assignments of an MDA connector.

Color Graphics Adapter

The CGA system uses three pins not used by the MDA system for its discrete color signals. The intensity signal remains at the same location, and the pin used by MDA for video is not used under CGA. Table 17.11 lists the signal assignments of a CGA connection. Although this system would seem to prevent any difficulties when mismatching a CGA monitor and MDA video board or vice versa, such mismatches can result in equipment damage. The synchronizing signals, which appear on the same pins in both connection systems, are sufficiently different to cause damage if they are applied to the wrong monitor type for a sustained period.

Enhanced Graphics Adapter

For compatibility with all IBM monitors, the EGA used the same connector as previous video boards, a nine-pin, female D-shell connector. The definitions of its signal pins were controlled by the setup DIP switches depending on the type of monitor that was to be connected to the board. Both the monochrome and CGA-compatible color schemes corresponded exactly with the MDA and CGA standards to allow complete compatibility. For EGA displays, the CGA intensity pin was used for the intensity signal of the green gun. Additional intensity signals were added for the red and blue guns. Table 17.12 lists the signal assignments.

Because EGA boards can supply signals complying with the EGA, CGA, or MDA standards, you must know which standard an EGA board is set to use before connecting a monitor. Otherwise it may cause damage just as it would for a CGA-MDA mismatch.

Professional Graphics Adapter

The first display system to use analog signals was IBM's specialized Professional Graphics Adapter. Designed for RISC-based workstations, the PGA system introduced many concepts that would find their way into VGA. Despite the similarity between its signals and those of VGA, the PGA system used a different connector, one based on the same nine-pin D-shell shared by all the other IBM display systems pre-dating VGA. Table 17.13 lists the signal assignment of this connector.

Fifteen-Pin High-Density D-Shell Connectors

The most common connector on PC monitors is the 15-pin high-density D-shell connector. Originally put in use by IBM for its first VGA monitors, it has been adopted as an industry standard for all but the highest performance computer displays.

Because the signals generated by the VGA are so different from those of previous IBM display systems, IBM finally elected to use a different, incompatible connector so the wrong monitor wouldn't be plugged in with disastrous results. Although only nine connections are actually needed by the VGA system (eleven if you give each of the three video signals its own ground return as IBM specifies), the new connector is equipped with 15 pins. It's roughly the same size and shape as a nine-pin D-shell connector but before IBM's adoption of it, this so-called high-density 15-pin connector, as shown in Figure 17.14, was not generally available. Nearly all of today's VGA-based display systems (including SuperVGA) use this connector.

Figure 17.14 A 15-pin high-density D-shell jack.

In addition to allowing for four video signals (three primary colors and separate sync) and their ground returns, the VGA connector provides a number of additional functions. In the original VGA design, it enabled the coding of both monitor type and the line count of the video signal leaving the display adapter. The modern adaptation of the connector to the VESA DDC standard redefines several pins for carrying data signals. Table 17.14 lists the signal assignments used by this connector for the basic VGA and SuperVGA systems.

IBM's 8514 and 8515 displays as well as 8514/A and XGA display adapters also use the same connector even though they at times use different signals. Again, however, IBM has incorporated coding in the signals to ensure that problems do not arise. The 8514/A and XGA adapters can sense the type of display connected to them, and do not send out conflicting signals. The 8514 and 8515 monitors operate happily with VGA signals, so problems do not occur if it is plugged into an ordinary VGA output.

The advantage of the 15-pin connector is convenience. One cable does everything. On the downside, the connector is not arranged for proper high speed operation and its deficiencies can limit high frequency performance, which in video terms equates to sharpness when operating at high resolutions and refresh rates. Consequently, the highest resolution systems often forego the 15-pin connector for separate BNC connectors for each video channel.

BNC Connectors

True high resolution systems use a separate coaxial cable for every signal they receive. Typically, they use BNC connectors to attach these to the monitor. They have one very good reason. Connectors differ in their frequency handling capabilities, and capacitance in the standard 15-pin high-density D-shell connector can limit bandwidth, particularly as signal frequencies climb into the range above 30 MHz. BNC connectors are designed for frequencies into the gigahertz range, so they impose few limits on ordinary video signals.

Monitors can use either three, four, or five BNC connectors for their inputs. A three-connector system integrates both horizontal and vertical synchronizing signals with the green signal. The resulting mix is called sync-on-green. Others use three connectors for red, green, and blue signals and a fourth for horizontal and vertical sync combined. This scheme is called composite sync. Five connector systems use three color signals: one for horizontal sync, and one for vertical sync. These are called separate sync systems.

Audio

Although the number of monitors (particularly those with composite inputs) with audio and video capabilities had been declining, they have enjoyed a resurgence over the last year. IBM, Apple, and many clone monitors are adding audio (both input and output). This can be useful in at least two cases—to take advantage of the new voice synthesis and voice digitization options now becoming available for PC systems, and to amplify the three-voice audio output of the PCjr. Most monitor audio amplifiers, even those with modest specifications (limited audio frequency bandwidth and output powers less than a watt), can handle either job adequately.

Only recently has music making become a passion among PC makers, and even the most favored systems still relegate audio to plug-in accessories. Although you can add accessories to transform the musical mission of your PC, you'll also want to add better quality audio circuitry than you'll get with any PC. A patch cord to connect the add on accessories to your stereo system will do just fine.

Enhanced Video Connector

Among the other problems with the standard VGA connector that's so popular on today's monitors, one stands in the way of getting truly high resolution images on your monitor—limited bandwidth. The VGA connector has a maximum bandwidth of about 150 MHz, so it already constrains the quality of some display systems. Conventional high frequency video connectors (primarily BNC) require a separate cable for each video signal and allow no provisions for auxiliary functions such as monitor identification.

The VESA Enhanced Video Connector is designed to solve these video and identification problems and incorporate sufficient additional signals to permit linking complete multimedia systems with a single plug. The final EVC standard allows for four wide bandwidth video signals along with 30 other data connections in a connector not must larger than today's VGA.

Unlike other connector standards, the EVC was not designed to accommodate any new kinds of signals. Instead it is a carrier for existing interconnection standards. It allows grouping together nearly all of the next generation of high speed computer connections in a single cable. The point is that you can put your PC on the floor, run an EVC connection to your monitor, and connect all your desktop accessories—keyboard, mouse, printer—to your monitor. The snarl of cables leading to the distant PC disappears thanks to the magic of the EVC.

In addition to RGB video with composite sync, the EVC standard includes provisions for composite and S-video signals as well as international video standards (PAL and SECAM) in systems that support DDC for monitor identification and negotiation. Connections are also provided for analog audio signals (both inputs and outputs). To connect additional peripherals, EVC also accommodates both Universal Serial Bus and IEEE 1394 ("FireWire") high speed serial interface signals. Other connections provide DC power for charging notebook computers.

The centerpiece of the EVC design is the Molex MicroCross connection system which uses four pins for video connections separated by a cross-shaped shield. The design links coaxial cables and allows for bandwidths of 500 MHz. The additional data signals are carried through 30 additional contacts arranged in a 3 by 10 matrix. Figure 17.15 shows the contact arrangement on an EVC connector.

Figure 17.15 The VESA Enhanced Video Connector.

Although the EVC design is not a true coaxial connector, it provides a good impedance match (within 5 percent of a nominal 75 ohms) and shielding that is 98 percent effective. The signal bandwidth of the connection is approximately 2 GHz.

All EVC connectors do not have to include all the signals specified by the standard. System designers are free to choose whatever signals they would like to include. That said, VESA recommends that manufacturers adopt one of three levels of support or signal sets for EVC for their products. These include Basic, Multimedia, and Full.

The Basic signal set is the minimum level of support required for devices using EVC. It includes only the video signal lines and DDC. At the Basic level, EVC operates as a standard video connector much like today's VGA connector but with greatly improved bandwidth. The DDC connection allows the monitor and its host to negotiate the use of higher resolution signals. VESA recommends that the Basic signal set be included in any subset of the EVC signals that a system designer chooses to support.

The Multimedia signal set adds audio support to the Basic signal set, allowing a single cable to carry video and audio to a suitable monitor. VESA foresees that the Multimedia signal set will usually be supplemented with USB signaling.

The Full configuration includes all of the signals provided under the EVC standard.

EVC connectors used for reduced signal sets such as Basic or Multimedia need not include physical pins or contacts for the unused connections. For example, a Basic EVC video connector may have only 2 of 30 subsidiary pins, those for the DCC link.