Our mailbag lately contains a lot of questions about various aspects of video. There appears to be some confusion about RGB, composite, and monochrome monitors, how these differ from television, what is meant by “resolution,” and so on. In SWN Vol. 1 (“Journal #1”) we ran an in-depth article on monochrome (“black-and-white”) video, and I still recommend this Video Primer for anyone wishing to learn more about video systems. However, this ZX81-oriented publication is not a prerequisite to this article. Since it was limited to monochrome, the previous article sheds little light on the various aspects of color video. I also recommend that you look over the back-issues of SWN Vol. 2, 3 and 4, since several specific video-related questions were answered in “Forum.”
Definitions
Let’s start by defining a few basic terms. It has been said that to understand all the words means to understand the subject.
CRT (Cathode Ray Tube)
This is the central element in all but the most recent video display systems (i.e. pocket TV’s based on LCD’s or “Liquid Crystal Displays.”). The term “CRT” derives from the very infancy of electronics, when it was discovered that certain rays were emitted by the cathode (negative terminal) in a tube from which most of the air was evacuated (i.e., a “vacuum tube.”) These particles were later named “electrons,” and are in fact what electronics is all about. Still, the original terminology has stuck to this day. A monochrome CRT has at its heart an “electron gun,” which is the heated cathode from which electrons are emitted. These are formed and focused into a narrow beam by various electrodes known as “grids” (a throwback to the nomenclature of vacuum tube technology). One of the grids is also used to vary the intensity of the beam. (This is important both for setting the average brightness with the Brightness control, and also the instantaneous brightness changes which make up the final picture.) The beam impinges on the front of the CRT (the part you look at), which is coated with a “phosphor” which emits light when struck by electrons. A color CRT is similar, except that it has three separate electron guns; one for red, one for blue, and one for green. (Using light sources in these three primary colors allows us to make any other color by appropriately blending them.) The face of the CRT is covered with dots of phosphor instead of a continuous layer. Right behind the face is a “shadow mask,” or plate with 100’s of thousands of tiny holes. It is carefully positioned such that the electron beams from the three different guns strike the appropriate color of phosphor dot. The key to operation is parallax; through any given hole, a red, green, or blue dot can be lit by “firing” the appropriate electron gun. Around the neck of the CRT are a set of “deflection coils,” which are electro-magnets that control the position of the electron beam(s). These coils are connected to the outputs of the vertical and horizontal “oscillators,” which cause the electron beam to sweep back and forth, and up and down.
Video
The term “video” is the first-person singular of the Latin verb videre, “to see.” (The title of the Vol 1 article thus translates as, “I came, I saw video, I conquered!”) In electronics, “video” means “relating to the electronic production of visual images.” More specifically, it is the name given to the signal causing the electron beam(s) in a CRT to vary in intensity, making the light and dark (or colored) areas in the final picture. Note that for monochrome, only one video signal is required, whereas for the simplest color systems, THREE separate signals are needed: one for each primary color.
Synchronization
Most of us know what it means to synchronize something. The word comes from Greek roots meaning “same time.” As relates to video systems, this refers to signals that mark exactly when each frame (picture) starts, and when each line within the frame begins. These “sync” (short for “synchronization”) pulses cause the vertical and horizontal oscillators in the destination device (your TV or monitor) to “lock step” with the corresponding signals in the source device (video camera, VCR, or computer).
Composite Video
Let’s say, in a monochrome video system, that we wish to run only one wire (or use only one band of radio frequencies) to convey all the video information. However, we have three discrete signals to deal with; vertical sync, horizontal sync, and video. These can be combined into a single signal, appropriately called a “composite” video signal. See the SWN Vol. 1 article for a more detailed description of how this is done. In the case of color, the problem is even hairier. We have not three, but FIVE different “primal” signals to deal with; the two sync signals, red, green, and blue. I used the term “primal” in the last sentence, because there are several more intermediate signals generated in the mix-down (and later, in the decoding) processes. I’ll state without proof that a composite color video signal is very complex. I’ll also make the somewhat philosophical statement, that it’s amazing that composite color works at all, truly miraculous that it works as well as it does.
Luminance & Chrominance
The video information in a composite color system is broken down into two basic elements, “luminance” (Latin, “lumen” = “light”) and “chrominance” (Latin, “chroma” = “color”). Luminance simply relates to “how much light,” whereas Chrominance means “how much color.” Luminance therefore describes the monochrome signal, whereas chrominance additionally specifies the colors of that signal. Why bother with luminance at all? After all, given the color information, we should be able to control intensity as well. The answer is akin to the reason that color photographs contain silver (black) in addition to the colored dyes. It results in better control of the finished picture. Your “contrast” control varies luminance, whereas “color” (chroma) varies the chrominance. How do we get three colors from one chrominance signal? The full answer is far beyond the scope of this article. Suffice it to say that this is done by varying the timing of the chroma signal, relative to a signal called the “color burst” or “chroma burst” sent by the source, and duplicated at the destination by an accurate crystal. The relative phase (timing) of the chroma signal determines the color. You can adjust the average (or quiescent) phase by adjusting your “tint” control.
Monitor
We all know what a TV is, so I won’t waste space on it here. But how does a “monitor” differ from a TV? The answer is similar to the difference between a radio and an audio amplifier/speaker. As I’ve said before, a TV is simply “a monitor with radio.” A television set contains a tuner (to receive radio signals encoded with video information), plus “intermediate frequency” amplifers and a “detector” to convert the radio signal into a composite video signal. (We’ll ignore the audio portion here.) Many TV’s are being sold these days with a “monitor input.” The manufacturers usually allude that this feature is what justifies a considerably larger price tag. In reality, they’ve had to do very little more than install a jack and a switch to a standard television. If you are looking for a monitor which doubles as a TV set, you might be better off buying a real composite monitor and a (perhaps used) VCR; to watch TV, simply use the tuner in the VCR. NOTE: most home VCR’s will NOT work with RGB monitors.
RGB
As you probably know, “RGB” simply means “Red, Blue, Green.” What you probably don’t know, is that an RGB monitor is the SIMPLEST color video system you can get. Three separate lines (R, G and B) are used to directly control the three electron guns, and an additional line or lines supply the sync signals. Again, manufacturers would like you to believe that there is something very special about RGB, justifying a larger price tag. Not so. Part of the explanation of higher cost has to do with production volumes, the other part is that RGB looks great, and therefore nobody complains. So why is RGB the simplest, and why does it look so good? There’s a lot less electronics that the signal has to go through before reaching the CRT, since the monitor doesn’t have to decode the composite video signal into sync, luminance, and chrominance elements. The computer doesn’t have to laboriously assemble the R, G and B signals into composite, only so that the composite monitor can take it all apart again before sending it to the electron guns.
Resolution
Resolution is a measure of how many discrete dots or lines can be displayed before they merge together. In photography, the term is very precise, and is usually expressed in “lines per millimeter.” In video, it’s usually much more qualitative. Monitors generally break down into “low,” “medium,” and “high” resolution, with corresponding quantum jumps in price.
Horizontal
Resolution also has different meanings when applied to monochrome or color monitors. In a monochrome unit, resolution depends almost completely on only two factors; dot size (diameter of the beam from the electron gun), and video bandwidth. The effect of dot size on resolution should be quite obvious; the smaller the dot, the higher the resolution. Bandwidth describes the maximum frequency (and therefore, discrete dots per line) that the video amplifier is capable of. Low resolution is anything up to around 4 mHz. (mega Hertz, formerly megacycles per second). Medium resolution is around 4-7 mHz., and high resolution is anything above that. Standard NTSC (American) television has a bandwidth just under 4 mHz., a limitation due to the restricted space allowed in our television channels. (This is one reason why a TV with “monitor” option is usually not as good as even an inexpensive composite “true” monitor; they are designed as TV’s first, monitors second.) In color monitors, a third factor affects resolution. This is that the phosphor is made of discrete dots corresponding to the shadow mask, instead of a continuous sheet. On the low-res CRT’s used on televisions, the number of dots (dot resolution) was chosen to be just within the bandwidth limitation. By minimizing the number of holes in the shadow mask, costs are kept down. So, even if “TV with monitor option” sets have video amps with a larger bandwidth, the CRT will itself impose a resolution restriction. The definition of high, medium, and low dot resolution depends on who you talk to. Perhaps the best definition is in the QL manual, and I quote: “The horizontal resolution of the QL is 512 pixels across the screen. This means that a monitor used with the QL should have a shadow mask resolution adequate to cope with this. For a 14-inch monitor, this resolution should be at least one dot every 0.43 mm. (known as “medium resolution” in England, and “high resolution” in the U.S.); one dot every 0.31 mm. (“high resolution” in England) is better still. Monitors with standard TV tubes (one dot every 0.85 mm.) are usable, but can suffer from poor distinction of vertical lines.” The above implies, that as regards our TS1000 and TS2068 machines with 256 horizontal dots, the low-res TV-type monitors are entirely adequate. And indeed they are, for the most part. When dealing with 64-column screens, however, they are woefully inadequate. For such text applications, you’re MUCH better off with even a cheap monochrome monitor.
Vertical
Vertically, resolution is less of a problem. This is because our machines (TS1000 family, and Spectrum/TS2068) use “non-interlaced scanning.” This requires an explanation of “interlacing.” In video as used by television, the horizontal scan rate (and therefore sync signal) is a non-integral multiple of the vertical rate. More precisely, it’s an integral multiple of HALF the frame rate, or 30 Hz. The result is that the two sets of lines from subsequent 1/60th second frames, will “interlace,” i.e. the second set of lines falls between the first set. This brings up a point of semantics. We computerists use the term “frame” to specify the 1/60 sec. interval during which an entire screen is displayed. This is more precisely termed a “field.” Two fields (1/30 sec.) make up one “frame,” or complete TV picture. (Comes from a “frame” of a motion picture film, but the analogy is not exact since motion pictures usually run at 24 frames per second). Some computers DO use interlaced scanning, but the TS family does not. You might think of this as using only the odd-numbered lines available on the TV or monitor. This is why, on monochrome monitors, horizontal lines appear as lines, whereas vertical lines appear as row of dots. In still other words, our “frame” (complete picture) is the same interval as our “field” (vertical scan interval), so it is correct to refer to the 1/60 second vertical interval as a “frame.”
R.G. Blues
The RGB system is not standardized from one manufacturer to another. Some RGB monitors have a single sync line carrying both horizontal and vertical pulses (composite sync), whereas others have separate sync lines. The QL gets around this rather cutely by supplying a composite sync and a vertical sync; monitors that require separate sync can usually live with the composite sync instead of horizontal. There isn’t perfect agreement about the phase of the sync signals. This is especially a problem in European monitors (and computers, as the QL or the Amstrads) vs. American ones. Frequently, one or the other or both sync signals will need to be inverted with a gate such as a 74LS04. If your RGB monitor simply won’t “lock in” your picture (rolls or horizontal tearing that can’t be stably adjusted with the “hold” controls), you will probably only have to invert the offending line to cure the problem. Another thing we have to cope with is “analog” vs. “TTL.” “Analog” means that the signals can be varied in intensity, just as in standard video. “TTL” means that any given dot is either on, or off; nothing in-between. For 16-color computers as the Sinclair family, the TTL type is superior (if your RGB interface supports it). However, other machines (as Amiga and ST) require the analog type, so that they can reproduce the nuances of 256 (or more) colors by varying the intensity of the dots. A feature on some TTL monitors is an “intensity” input. This basically allows the monitor to be used in an analog mode. Think of it as being a luminance input, separate from the “discrete chrominance” of the R, G, and B inputs. If your present computer doesn’t support this input, it generally does no harm to simply leave it open.
More Info
I’ll close with suggestions for buying the right monitor for the job, and a few other side notes. If you’re planning on staying with the TS machines, a “standard” (i.e. low-res) monitor is entirely adequate. However, if you’re later planning on getting a more advanced machine, such as QL, Atari ST, or Amiga, a higher resolution monitor bought now will save you money later when you really need it. If you’re mainly interested in text applications, by all means get a monochrome monitor. Not only is it a lot cheaper, it is also a lot cleaner. Some users may want a monochrome monitor for their day-to-day serious applications, but would like to play the occasional game or see graphics programs in color. You don’t HAVE to buy a color monitor! If your computer has a TV output (as 2068), use that for those “rare” occasions. If not (as Oliger TS1000, some Atari ST’s, probably others), get a modulator from Radio Shack. Alternately, run your composite video into the “Video IN” jack on the VCR. If you want the easiest-to-use color monitor, get one with composite video. It’ll plug right in, on the 2068 or “Oliger 1000.” You don’t have to worry about the interfacing hassles that can plague RGB systems. If you want the best possible picture, the ONLY way to go is with RGB. Though it may rub you the wrong way to pay more for less (it does me,) “facts is facts” and nobody guaranteed life to be fair. Beware of those “$19 monitor bargains.” Though they may be excellent monitors, these are usually designed for some specific application (such as surveillance). These are neither composite nor RGB, and won’t work with either type of system. Many such monochrome monitors have separate sync and video inputs. You might get these to work with a circuit as described in the first “Primer.” There are also color monitors with separate luminance and chrominance inputs (“separated video”); we haven’t found a simple way to deal with these, so they are best avoided. Sound is a nice feature on a monitor, even if your present system doesn’t support it. Who knows what you might buy later on? Note that the sound input on most monitors requires a line-level (about 2 volts) sound input. The 2068’s MIC output, of course, is much lower; around .0001 volts. The level at the edge connector 31A is not much higher. So to use this with your monitor, try tapping it into the output of the sound op amp output (actually part of the switching regulator), at pin 4 of U1, or at the speaker (junction of CR24 and R19). How big should your monitor be? Though it’s largely a matter of taste (and eyesight), a few subjective hints may help. My favourite monitor for word-processing is a 9″ green screen, positioned about two feet from my eyes. Provided that you don’t sit too far away, a smaller monitor actually gives a sharper look. I’ve done a lot of programming on my “compact 1500 system,” which has a 3″ screen. When working on this system, my eyes are about a foot from the screen; under such conditions, that tiny display is entirely adequate (after all, it’s about the same size as a 2040 listing). For playing games and goofing with graphics, I prefer my 13″ composite color set. Our 21″ TV is ok for games only if I’m about 10 feet away from the screen (and even then it’s dicey). Other features to look for: 1: Easy adjustment of horizontal width (almost crucial for TI video and QL). 2: DC Restoration switch (composite & Mono) 3: 75 ohm / Hi-Z switch (composite) 4: Desired monochrome color (amber, green or white) 5: Reversing switch (monochrome) 6: Front-panel controls (bright, contrast, v/h hold, color & tint if applicable). Hope this helps. If you just jumped out of your chair yelling “Nunc video!” (Now I see!), then my purpose is complete. If not, let us know. We also welcome reader suggestions on getting around RBG-related problems, adapting “bargain” monitors with unusual video/sync requirements, and anything else you may have discovered.