"Everything that can be invented has been invented." -- Charles Duell, head of the U.S. Patent Office, 1899. A New Frontier In the preceding chapters, you have learned much that there is to know about our present North American television system, affectionately known as NTSC video. All of that will be changing in the next few years. And the effects of that change will have far-reaching consequences whether you are a television producer or a television viewer. Sometime in the next ten years, the existing analog transmission of NTSC video will cease, replaced by a fully digital system, from origination to home receiver. In this chapter, we will explore what it means to you as a television producer, and what it means to the over thirty million Canadian viewers who will be affected. A Historical Review A brief synopsis of how our NTSC system came into being is valuable to shed some light on the more recent negotiations that culminated in our new digital TV universe. Black and White In 1936, the Radio Manufacturers Association (RMA), the forerunner of the Electronics Industries Association (EIA), set up a committee to recommend standards for commercial TV broadcasting. In December 1937, the committee advised the FCC to adopt the RCA 343-line, 30-frame system. In 1940, the National Television System Committee (NTSC) was established, which was placed under the sponsorship of the RMA. The FCC adopted the standard on April 30, 1941. Key elements of the standard included: the use of a 6MHz RF channel vestigial sideband modulation of the picture carrier frequency modulation of the sound carrier 525 scanning lines per frame 2:1 interlace at 30 frames (60 fields) per second and 4:3 aspect ratio. Although the FCC authorized the first two commercial TV stations to be constructed in July 1941, the growth of television was ended by the licensing freeze that accompanied World War II.

Colour

During its development, it was assumed that colour would be demanded by the public. Field sequential systems were demonstrated in 1929. CBS showed a field sequential (colour filter wheel) system in the early 1940s. The CBS mechanical system was adopted as the colour TV standard in October 1950! Monochrome television was about nine years old, but already had a base of 10 to 15 million receivers in homes. Broadcasters and the public were faced with having their new and expensive equipment quickly become obsolete.

The wisdom was that colour must be an adjunct to the 525/30 monochrome system so that existing equipment and receivers could accept colour transmissions. The proponents of compatible, all-electronic colour systems, were making advances. RCA had demonstrated a tricolour picture tube. Hazeltine demonstrated the constant luminance principle, and what was called the "shunted monochrome" idea. GE introduced the frequency interlaced colour system. Philco showed a colour signal composed of wideband luminance and two colour-difference signals encoded by a quadrature-modulated subcarrier. These were all principles that were eventually put into place with our present colour system.

The RMA's technical committee (NTSC) was reactivated in January 1950. By November, a system employing the basic concepts of today's NTSC colour system was demonstrated. Field tests showed defects, such as sound interference caused by the choice of colour subcarrier. This was corrected by the selection of a different frequency, but at the expense of lowering the frame rate to 29.97 Hz. Finally, RCA demonstrated unequal I and Q colour-difference bandwidths. The proposal was forwarded to the FCC on July 22, 1953. Demonstrations of the system were performed on Oct. 15, 1953, and on Dec. 17 of the same year the FCC approved the colour standard. Colour service began on Jan. 23, 1954.

It is worthwhile to consider the longevity of the NTSC colour standard. From 1953 to the present, the standard has endured.

So Now What?

Fast forward from 1953 to 1987, thirty-four years later. In the United States, a new committee was struck, called the Advisory Committee on Advanced Television Service (ACATS) to investigate the new high-definition television technology and recommend a broadcast standard to the FCC. Twenty-three different proposals were submitted to ACATS by September 1988. They mulled them over and decided to build six of them for testing - two analog, four digital.

The tests were conducted in 1991 and 1992, and in the process ACATS threw out the analog systems. They also recommended at this time that whatever standard they were going to decide upon should have data transmission capability and surround sound (AC-3, 5.1 channels.) While they were at it, they adopted a resolution that all the manufacturers should get together and choose one common system.

Goodbye, NTSC...Hello ATSC

And so, a group called the Grand Alliance was born: AT&T, David Sarnoff Research Center, General Instrument, MIT, Philips, Thomson and Zenith together in one room. They came up with a system, tested it in the summer of 1994 and showed that it would outperform the existing NTSC process by delivering HDTV-quality signals into fringe areas and do so with lower transmitter power. After a few more refinements, the FCC approved digital television on April 3, 1997. The ATSC (Advanced Television Systems Committee) was born.

The standard, called A-53, involves using the same 6 MHz of bandwidth for each television channel, but changes the whole works over to a digital transmission system. It's called 8VSB - the VSB stands for "vestigial sideband," which is not unlike the present analog transmission scheme in that sense. However, there is basically one large carrier with lots of digital capacity (19.39 MHz of compressed video) to stuff in one HDTV signal or a few standard definition (SDTV) streams.

In the United States, as of summer, 2003, there are approximately 1000 transmitters now on the air in more than 200 markets reaching about 99% of the nation�s population. Eighty percent of those viewers have a choice of five or more different digital TV stations in their market. Every station is expected to be transferred to digital transmission by the end of 2004. All analog transmission is scheduled to stop in the United States on December 31, 2006.

There are now over four million digital television sets in stores and homes in the United States. They are available at a cost of around $3,000 CDN. Not everyone has to purchase a new DTV set, though. If you want to keep your old analog set, you�ll be able to purchase a digital-to-analog set top converter for around $500 CDN. To date, there have been over six million digital television products (televisions, set-top boxes, etc.) sold in the U.S.

The Canadian Perspective

It appears that the distance between the United States' rollout and our own is broadening. Some consider that we're now about five years behind the U.S. There is only one commercial over-the-air digital television transmitter in this country - Citytv in Toronto. There are test transmitters in Ottawa, Toronto, Montreal and Vancouver. Mostly because of the ready availability of high definition DVDs and the fact that our DBS services and digital cable are offering HDTV channels on their channel lineups, there are about 600 thousand digital television viewers in this country - about two percent of our population!

After forming CDTV (Canadian Digital Television) in 1998 - an organization dedicated to the Canadian DTV rollout, comprised of two dozen broadcasters, specialty networks, cable operators, satellite distributors, manufacturers and others - we have allocated the digital frequencies for all Canadian television broadcasters. And wrung our collective hands a great deal about the distinctly Canadian issues.

What are those issues?

Because so much money is at stake, all of the organizations involved are playing wait and see. Consumer TV set manufacturers do not want to produce or sell DTVs in Canada, since there are no television transmitters to send signals. Cable companies don't want to get too involved in compatibility issues with DTV, as there is no need to alarm their subscribers. Broadcasters don't have the programming available in Canada to put on an HDTV transmitter. And programming producers won't invest in equipment to produce high definition television, since there are no broadcasters or specialty services willing to purchase the new product.

We may well witness a phenomenon which has happened time and time again in Canada: the United States will be sending us advanced TV signals across the border, long before we have our own system in place. All we need to do is purchase the hardware and point our antennas south. This has happened with television itself in the early 1950s; with colour TV in the early 1960s; with DBS transmissions in the 1990s, and may very well happen again with DTV.

Stay tuned...

From The Chrysalis to the Butterfly

So, you'll be hearing a lot about this digital television stuff from now on. Let�s get a few things figured out while there�s still time.

Transmission Formats and Hitches

Keep in mind that the DTV standard specifies details for transmission, not production, display or user-interface, which will all be affected by it anyway. Let's start with transmission, which will be the first change to confront broadcasters.

To accommodate the changeover from analog to digital, all broadcasters in Canada and the United States have been be assigned another channel on the television dial. For a few years, you will be able to receive your favourite programming in analog and digital, depending on what channel you select. Obviously, your old analog television set won't pick up the new digital signals (unless you have a set-top box that does the conversion), and your new digital television set won't understand the soon-to-be defunct analog system.

Let's look at the potential difficulties with this transition. At first, there seems to be a perfectly understandable solution to the transition problem. You put up your antenna, and receive what you need.

But, our country is one of the most "cabled" ones in the Western hemisphere - about 85% of us have the ubiquitous coax coming into our homes. And cable companies have clearly stated that they've run out of analog spectrum for additional channels. How can they accommodate twice the number of channels? The truth is that they can't with an analog distribution system. So maybe the answer is in the now available digital cable distribution system. But unfortunately, the system known as QAM (quadrature amplitude modulation) is incompatible with A53. And, of course, you need to have a cable set-top box to receive all the channels...

It doesn't stop there, however. The commonly used DBS satellite compression system is known as QPSK (quadrature phase shift keying) and it, too, is incompatible with the new A53 system. In addition, while several standard definition television (SDTV) signals can be put on one satellite transponder (as we do today), not even two full bandwidth HDTV (high definition) signals (each of which takes up about 19.5 MHz) can be uplinked to a single transponder (which has a total bandwidth of 27 MHz.)

One more thing to keep in mind. Digital transmission, as we know, requires a certain minimum signal strength. If there is sufficient signal strength for the set to decode digital data, a perfect picture will result. If the signal strength falls below that level, the reception stops completely. There is no gradual deterioration of the picture quality like there is with our analog NTSC system. It is either all or nothing. This shouldn't be a problem for viewers within the normally accepted coverage area of a television transmitter, but people who presently receive stations in "fringe" areas may find that they are no longer able to receive their favourite "long distance" programming.

There will be some bugs in the system for a while.

ATSC Formats

The digital television specification has more than one "screen size" and "resolution" format. In fact, there are eighteen of them! Fortunately, you don't have to worry about your new digital TV becoming obsolete - all sets will be "smart" in that they will feature a relatively inexpensive scan converter to transcode incoming signals if they are different from the "native" display format of the set.

So, what are these various formats? Here's a chart to help understand the complexity of your new digital signal:

Easy, right?

What does all of this mean for us as programmers and viewers? You can develop programming in any of the eighteen different formats, as the need suits you.

For example, you may want to send your viewer a "movie of the week" in sparkling high definition at maximum resolution. You'd do your film transfer at 1920 x 1080 resolution, 16 x 9 aspect ratio, and 24 frames per second rate (major motion pictures are shot at this rate on 35 or 70 mm film) and send it to your viewing audience. Their television sets would decode it properly in HDTV, or in a lower resolution if that's all that their set is capable of displaying.

An Aesthetic Consideration

You're out on location with your new digital 16:9 camera and your director has told you to shoot your program so that it can be displayed in a pleasing manner on both their existing NTSC transmitter and their new digital HDTV transmitter simultaneously. What do you do? How do you frame the shots?

You have some choices:

Converting to and from 16:9 and 4:3 (courtesy Broadcast Engineering)

	You can shoot a proper shot for the 16:9 audience, and have the footage converted using a panning process for 4:3, thus destroying all of your wide screen aesthetic work.
	You can frame for 4:3, but the 16:9 version will have all of its action centred in the middle of the screen (creating rather boring shots and not utilizing the capability of this new format).
	Or you can try and compromise between the two screen formats, neither one being totally satisfactory, but not looking all that bad either.

The choice is yours. You will have to take into account your intended audience, the importance of archiving the material for future broadcasts, time, money and effort.

Squeezing It All In

Major broadcasters in the United States have already chosen their preferred formats:

But wait, there's more you can do with that 6 MHz of television bandwidth. You see, with the standardized MPEG-2 compression you have 19.39 Mbps available to you. You can send some regular SDTV signals down the bitstream, but also can mix into it interactive services such as banking, games, shopping or even Internet surfing. Subscription services such as audio or pay per view can be added in.

What Is This Going To Cost?

That depends on whether you're a viewer or a producer.

The Consumer Issues

The sets cost about $3,000 but like all mass-produced home entertainment products, that figure will probably drop rather dramatically over the phase-in years. The marketing people have their work cut out for them. It will be relatively easy to convince younger people to go for the latest and greatest in video reception. However, as you go up the age scale (all the way to your grandmother, who feels her present 20-year-old Zenith console set in the living room works just fine, thank you), there will understandably be some resistance. Studies show that the average television viewer expects to pay between $500-700 for their next television set. There's also the question of multiple televisions: will you be willing to pay several hundred dollars for a set-top box, for that cheap 14" TV in the den?

All persuading aside, there may be a question of availability of the sets themselves. We will need approximately 230 million television sets for Canada and the United States over, say, the next ten years. Can the manufacturers make almost 30 million TVs a year?

While you're at it, you can toss your VCR - it won't work, either. You will be able to purchase the new Digital VHS machine, which will record DTV, digital cable and DBS directly from these sources. And it will play back all your old VHS tapes. But it's expected to cost about $1000 U.S. Oh, and your camcorder (the one that plays back through your TV) must be either modified or pitched out. The consumer backlash on all of this could be considerable. But only time will tell.

Let's look at the viewing places - your living room, for example. Right now, you tend to sit comfortably on your sofa about ten feet away from your 27" TV set. NHK Broadcasting in Japan (one of the earliest developers of HDTV technology) conducted an extensive psychophysical research program in the early 1970s to see how people would literally view HDTV.

With HDTV, the information density is very high; the picture has a startling clarity. It is also in a new 16:9 format. Suppose you place your brand new digital television in the same spot as the old TV: about 10 feet away from the couch. To make the most of this high definition experience, you'd need a screen about 75 inches in diagonal! To keep this in perspective, CRTs can generally only be made to about 38 inches diagonal, since the mass of glass necessary to maintain the vacuum in the tube becomes unmanageable (and heavy!) One could imagine that we'll all be sitting much closer to the screen to get the "new viewing experience."

But, is all of this what consumers want anyway? A consulting group has found that there is a definite struggle between your average consumers, broadcasters, and television manufacturers. Consumers want simple, passive viewing, with long shows (movies, for example.) Broadcasters want shorter duration programming (more traditional television fare) and often lower resolution DTV formats. Manufacturers, for their part, want television viewing to be a cinematic experience, and would like the DTV machine to be as much a computer as a passive viewing device. Which, in turn, is not what the viewer wants. And so on.

The Broadcaster Issues

Estimates put the financial cost of upgrading a teleproduction facility at anywhere between $2 million and $15 million, per facility, to do a complete upgrade of all equipment in a television plant. Canadian broadcasters could easily spend $500 million over ten years. Cable companies could end up spending anywhere between 800 million to 1.5 billion to upgrade their systems. Clearly, this is something that will happen over several years in most facilities. The estimate above is only for the hardware:

Upconverters and Downconverters

No matter where a broadcaster is in the process of the grand conversion, there will be, for some time, lots of analog and 4:3 format video and film. This will often come from sometimes exhaustive archives and tape libraries. As well, there will still be a lot of "old format" programming produced every day, since the old equipment will be used to produce it. For example, the nightly newscast probably won't be produced in HDTV in the early stages (maybe not at all) because the importance of its immediacy will be overshadowed by the relative unimportance of super high resolution to cover tonight's top story.

With this constant flow of older formatted programming, conversions will have to be made from these formats to the new digital television format. Sometimes the conversions will have to be made simply to a higher scanning rate (480i to 720P, for example.) In other cases the formatting will have to change from 4:3 to 16:9.

In other situations, new digital television plants will be creating programming in the 16:9 high resolution formats, but this material will have to be "downconverted" to old 4:3 (and lower resolution) for broadcast on the analog television transmission channel.

Cameras

You will need cameras that shoot in a 16:9 format. Some newer models produced in the last few years are 4:3/16:9 switchable, which means that they will produce a 480i signal in a new wider format. This is a start, but if you intend to do HDTV production eventually, the old cameras, even in their wide format, will have to be upgraded or replaced.

VTRs

VTRs, some say, are format independent - especially digital VTRs, since all they're recording are binary data. To an extent, this is true, especially some later formats. However, all of the older analog machines (including the thousands of Betacam units out there) will eventually become obsolete.

Timecode Systems

So, eighteen different formats, you say? With different frame and field rates? What timecode system will work with them all? You will need a timecode generator that works at 24 fps, 30 fps, 60 fps. For the time being (until NTSC finally goes away), you'll also want that generator to function with their "drop frame" equivalents of 23.98, 29.97 and 59.94 fps.

Monitors

Every 4:3 monitor will be unable to display a 16:9 raster, except those designed to be switchable from one format to another. This means all control room monitors (including master control), external viewfinders, local VTR room monitoring, feedrooms, audio control rooms in video facilities, boardrooms, edit suites, screening areas - everywhere there is a monitor, it will have to be replaced. There are hundreds in every television station.

Switchers and Processing Devices

Present switchers (except some new digital ones) don't work in this format, either, so they will have to be exchanged. All DVE systems won't do their work in 16:9. Processing amplifiers, TBCs and frame synchronizers all need to be changed out.

Waveform Monitors and Vectorscopes

NTSC monitoring units are for just that, NTSC. They all have to be swapped out with new digital versions wherever they are found - control rooms, VTR banks, master control, feed control areas, transmitter rooms, and so on.

Transmitters

All new transmitters need to be purchased if for no other reason than to duplicate digital and analog service on two different television channels. This will be a very costly upgrade.

Non-Video Upgrades

Sets

With HDTV, you can see all kinds of small imperfections in cheap paper mach� set construction. They'll have to be upgraded or changed for things that look very real, even close up, because the home viewer will be able to tell that it's "fake."

Makeup

HDTV shows every pore and wrinkle in the talents' faces. New techniques will have to be developed for these obstacles (if we still believe that they're impediments).

Personnel

Engineering staff will all have to be trained to install, repair, upgrade and manage this new television plant. Operations personnel (technical directors, camerapeople, VTR operators, editors, master control operators, character generator operators, graphics artists, set designers, to name but a few) will require instruction in various elements of high definition and/or wide screen formatting. Production personnel (directors, and others working in format design) will need new aesthetic paradigms to deal with the new medium.

Some Other Technical Considerations

New Year's Resolution

NTSC

Back in the Analog Video chapter, we discussed the resolution of our 1950s-era NTSC television signal. Expressed as "TV lines per picture height", it came out to about 330 TVL/PH.

Notice that this is limited by the transmitter - our broadcast equipment (cameras, character generators, etc.) may have much higher resolution when we make our pictures in the studio. It's just that this higher resolution doesn't make it to the home viewer.

Digital Resolution

Digital video's horizontal resolution is determined by the sampling frequency (see the Digital Video chapter.) A physical reality called the Nyquist theorem states that the highest frequency you can reproduce is one which is one-half the sampling rate. Let's look at composite and component video to see how the resolutions come out.

Composite Digital Video

The facts are as follows. Composite digital video's sampling rate is 14.3 MHz. The maximum frequency that can be reproduced is half of this, or 7.15 MHz. Our scanning line is still 52.86 Fs. long. But we can put black and white vertical stripes in each cycle of our maximum frequency. Finally we multiply the whole thing by .75 (we're dealing with a 4:3 format) to give us our new comparative number TVL/PH. The math goes like this:

52.86 Fs x 2 x (14.3 MHz/2) x .75 = 567 TVL/PH for composite digital video

If you sample in composite (14.3 MHz), you'll get 567 TVL/PH.

Component Digital Video

The formula works the same way, except that component's luminance sampling rate is 13.5 MHz, so you get a little less resolution:

52.86 Fs x 2 x (13.5 MHz/2) x .75 = 535 TVL/PH for component digital video

If you sample in component (13.5 MHz), you'll get 535 TVL/PH for a 4:3 aspect ratio system.

HDTV Analog Resolution

A longer scanning line (25.86 Fs of visible line time) and a higher video line count (1080 visible lines) will result in a higher resolution. In analog HDTV systems (yes, there are lots of HDTV systems that work in an analog system), the bandwidth limit is 30 MHz (which is a lot higher than our old NTSC transmitter limit). This is largely due to the fact that there are practically no HDTV analog transmitters out there and the ones that exist were designed to operate on special, non-NTSC channel allocations.

Also, we're now dealing with a 16:9 format so our "make it square" factor is no longer 3/4 (or .75) but is now 9/16 (.5625). The formula works out like this:

25.86 Fus x 2 x 30 MHz x .5625 = 873 TVL/PH.

HDTV Digital Resolution

The sampling frequency for HDTV digital is 74.25 MHz. This, of course, gets cut in half because of the Nyquist theorem. The resolution is limited only by that 30 MHz bandwidth we just mentioned in the HDTV analog world:

25.86 Fs x 2 x (74.25 MHz/2) x .5625 = 1080 TVL/PH

Here's a comparison chart of what we've learned.

It's a long road and the conversion will take decades, but it should all be worth it in the long run. We will, in fact, have much higher resolution pictures in our new system.

Integrating On Air Program Material

Because the existing bandwidth of our transmission systems (particularly landlines, satellites, cable and over-the-air) is limited, we have to compress our digital signals so that they can be sent from one place to another. MPEG-2 is the compression scheme used, and this process has been discussed in a previous chapter of this book.

One peculiarity of MPEG-2 is that our video signal is delivered in a kind of "shorthand" made up of Intra frames, Prediction frames and Bidirectional frames (the familiar I, P, and B). A series of these frames makes up a "group of pictures" (GOP). Because of the MPEG structural design you can't edit or switch in the middle of one of these groups. Therefore, when this signal comes to master control, it must be decoded back to full-stream serial digital so that commercial breaks, station IDs, "bugs" in the corner of the screen, and other related programming can be strung together to be delivered to the home viewer.

The problem comes in when you use a severely high compression ratio on material during original encoding, decode it again for master control integration, then re-encode it again for further transmission. The situation is compounded when the output of this master control is a "network feed" to be decoded, reintegrated with new material and re-encoded yet again for the home viewer. The process of compression and decompression can result in artifacts appearing in the picture - the more often you do this, the worse it gets. And the problem is even more noticeable in HDTV material where even the slightest flaw can show up in an otherwise beautiful highly detailed picture.

The partial solution to this is to use "network feeds" that are not as highly compressed as full MPEG-2 "for air" feeds. These feeds can be compressed only to about 30 to 50 Mbps, instead of the usual 19.39 Mbps used in normal transmission. The cost of this is the requirement for a higher bandwidth channel from the network to the local affiliate. However, this milder compression will reduce the chance of concatenated (linked, chain-reacted) errors "piling up" on your signal.

A Transmission Wrinkle

One rather unnerving lesson that's been discovered in the U.S. is that some of the new digital television channels are interfering with the existing analog ones. This won't be a problem once the analog transmitters are shut down, but for the time being, new channels are popping up every week, in previously unused spectrum space. What do you do when your analog signal, previously crystal clear to your viewers, ends up clobbered by a new digital TV station on an adjacent channel? Legally, both stations are not doing anything wrong, but in the end it's the viewer that suffers. So, the solution is cooperation between the engineering teams of the two services. With the use of directional antennas at your viewers' homes and cable company head-ends, problems such as these may be reduced somewhat.

Interactive Television

Interactive TV is the ability to send lower bandwidth information to the home viewer along with the usual television pictures. The viewer, in turn, has the opportunity to interact with this material to get additional information on programming, products and services. While not strictly a digital television technology, this new element of our broadcasting environment will blossom with the appearance of digital carriers.

There are three ways that interactive TV can be accomplished.

TV Crossover Links

This technology uses the existing closed captioning area (line 21 of the NTSC video signal) to send an Internet URL (Uniform Resource Locator, or web page address) directly to a set top box, along with a "trigger" (usually appearing as a small "i" watermark overlaying the program). The home viewer can click on the bug, and be sent to the Internet address for more details on programming, products and services. Boxes such as those distributed by WebTV Networks allow the viewer to watch regular television programming and interact with the Internet simultaneously, on the same screen.

Low Bandwidth Analog

This system relies on the space in the vertical blanking interval (VBI) of existing NTSC transmissions. Data is sent during regular broadcasting and is stored and displayed on a local set-top box. It allows limited interactivity, due to the limited space in the VBI, but has potential to link information directly with the show being broadcast at the time.

High Bandwidth Digital

Data no longer has to be inserted in the VBI here, since the digital television specification allows for packeted data streams to reach the home receiver directly, where they can be decoded and displayed. This allows better synchronized audio and video along with this data. Digital television allows the viewer to see multiple video sources at once and this feature can be exploited by interactive television programmers.

Conclusions

Clearly, there aren't any just yet. The Future of Television is changing every month, it seems. The best thing to do is stay informed and keep up with issues as they unfold.

Stay tuned...there's much more to come!