HDTV World Review Volume I Number 4: American Agenda, Intelsat, DBS, and Digital Cinema

Editorial Board
Executive Editors: Dale E. Cripps and Sam Bush

From The Editor

A Current History Of High Definition Television
Corey Carbonara here offers his third article in a series on HDTV history. Carbonara reviews HDTV development efforts of the 1960s and 1970s; discusses the standards issues of the 1980s, including the roles played by the U.S., Japan, Europe, industry groups, and government; and brings us to the present day with a synopsis of standardization and commercialization efforts of the 1990s. In a future issue of the HDTV World Review, we will present the author's views on how today's managers incorporate new technologies into their business and educational environments.

HDTV: The American Agenda
Zenith Electronics president Jerry Pearlman is well-placed, as the head of the only American television manufacturer, to present an American agenda for HDTV. Pearlman's opinion--in marked contrast to other industry leaders--is that wide-screen television is not as important an offering as some people may think, and that flat panel displays are still many years in the future. He believes that HDTV's high resolution will be seen and appreciated on small screens as well as large, and argues that because of this it makes economic sense to stay with CRT technology and 4:3 aspect ratio in the first phase. One thing is clear: the public will not move to advanced television in large numbers unless the cost increase over NTSC is small. Pearlman indicates that his company can provide receivers in their 787.5 progressive standard starting at a $500 premium, which should drop to $200 as production increases. Given that kind of set pricing, and assuming FCC action on a standards decision in 1993, Zenith predicts an upbeat future for HDTV.

International HDTV Service Via Intelsat
Comsat (Communications Satellite Corporation), the original American satellite operator, is an authoritative voice on satellite issues and is deeply involved internationally with satellite issues. Comsat also is a sophisticated player in the HDTV arena, and has invested heavily in testing both HDB-MAC and MUSE. Its subsidiary, Comsat Video Enterprises, is an entertainment delivery system that provides movies to over 1700 hotels nationwide. With the recent purchase of 20 decoders from Scientific Atlanta in their HDB-MAC transmission system, Comsat plans to test, deploy and ship HDTV to hotels on a large scale. In this article, Edward Faine and William Schnicke of Comsat describe the Intelsat satellite system and some of the testing configurations that have been examined thus far.

Direct Broadcast Satellite
Hubbard Broadcasting has been a pioneer in both radio and television and now, continuing the family tradition, the company is pioneering DBS and HDTV as well. Hubbard began developing its direct broadcast satellite strategy in the early 1980s (when DBS had its first big surge of enthusiasm), astutely snapping up an orbital slot from the FCC. Where others retreated, Hubbard continued planning, testing and promoting for a full decade, an effort that culminated in the company's recent ordering of a satellite. Hubbard believes strongly that small antenna size--18 inches or less--is the key to wide public acceptance, and is therefore committed to high-powered broadcasting. The author predicts that with full U.S. coverage (unlike a cable system or a broadcast tower) and a total potential market of 250 million people, vertical niche markets of one or two percent of the population will make for a hugely profitable business. Hubbard suggests that HDTV services can be introduced in the same way--and is fully prepared to do so.

It's About Quality: HDTV and the Movie Theater Business
John Johnston here presents Kodak's view of the electronic theater. Speaking from the perspective of the film business, he asks whether electronic HDTV distribution and display is an economically viable alternative to film, which today averages a distributor cost of only $6 per showing. Johnston argues that at current and foreseeable HDTV equipment prices, conventional theater exhibitors would be facing a ten-year amortization to make the switch. This reality--together with the ever-improving quality of film stock, the addition of digital audio, increased speed of film duplication, etc.--for Kodak add up to a value in film exhibition that will be hard to shake. However, the author does believe there is a place for HDTV display in special markets, live events, educational programs and business meetings that benefit from television's immediacy. This type of showing will appear in new "HDTV theaters" and also as a new service within conventional film outlets.

Field Emission Displays for HDTV
Chris Curtin's article was presented as a paper at the HDTV Newsletter-sponsored HDTV conference held in February of this year. (Future issues of the HDTV World Review will contain other articles on display technologies.) According to Curtin, the display target for high definition is the flat-panel, big-screen, wall-mounted display. Of the many labs now working to develop this product, most are focusing on new technologies. Curtin argues here that there is potential for current CRT technology to be modified for use in these displays--he wants to extend the current technology to create a flat panel display with known elements. The author believes this approach will bring about a faster flat panel solution, and that it will be commercially manufacturable with reasonable cost and high yield.

As always, we invite your comments on the views expressed by our authors. We look forward to continuing the dialogue in our coming second year of publication.

Editor

A Current History Of High Definition Television
By Corey P. Carbonara

Introduction: The 1960s

The introduction of television in Japan took place around 1960. During the mid-1960s Japan began its long-range planning for the requirements for an advanced television system that could meet the high definition demands of the twenty-first century. By 1968, within ten years of the introduction of television in Japan and only two years after Japan adopted NTSC as its color standard, NHK began research on a new HDTV system that would use projected 35mm film as a technical benchmark. Led by HDTV pioneer Dr. Takashi Fujio at NHK, the research team examined the psychophysical attributes of human vision as well as the technical foundation for such a system.

Fujio concluded that HDTV--called Hi Vision by NHK--needed a resolution of 1000 or more scan lines. NHK established 1125 lines for Hi Vision based on mathematical correlations to both the 525-line (U.S./Japan) and 625-line (Europe) color standards in use by those countries today.1 Much of this early work resulted in significant contributions to studies of visual and aural acuity as well as specific research on the movement and characteristic of color and the eye, psychological effects of the visual field, motion adaptive qualities of the human eye and the effect of collateral sound effects to the presentation of visual material.2

Decade of Development: The 1970s

In 1972, NHK drafted a program of study for HDTV to the International Radio Consultative Committee (CCIR). Based on the parameters of the NHK system that included 1125 lines, a 60Hz field rate and a 5:3 aspect ratio, the CCIR set up an internal committee in 1974 to study the possible standardization of HDTV.3 It was also in the early 1970s (around the same time period that RCA entered the satellite business) that RCA began to study a new satellite distribution technology that beamed powerful signals directly to very small receiving dishes at the home. This new distribution service is called Direct Broadcast Satellite (DBS).

By the late 1970s, HDTV research began to be transferred from Japan through many of the technical journals and conferences held by the Institute of Electrical and Electronics Engineers (IEEE) and the Society of Motion Picture and Television Engineers (SMPTE). In 1977, SMPTE formed its first study group on HDTV, under the leadership of Donald Fink. In 1978, the BBC entered into discussions on high definition by publishing a report on a system of satellite broadcasting for HDTV. That same year, reports on HDTV began to appear in various technical journals with more frequency. The following year, NHK began its first satellite transmission experiments in HDTV.

Searching For Standards: The 1980s

In February 1981, NHK demonstrated HDTV in North America for the first time at a SMPTE conference. As a result of these demonstrations, CBS began a more directed effort in HDTV. That same year, CBS petitioned the FCC for allocation of the 12 GHz DBS spectrum for an HDTV system. By March 1982, CBS initiated its first 12 GHz terrestrial broadcast experiments.4 In addition, 1982 marks the year that the Advanced Television Systems Committee (ATSC) was formed. The ATSC is based on the same NTSC that initiated standards for monochrome and color.5

In 1983, RCA began work on Improved Definition Television (IDTV). This research primarily involved merging RCA's vast knowledge of digital technology with its existing knowledge of color television. That same year, the CCIR (part of the International Telecommunications Union) set up an interim working party with the charter to produce a recommendation for a single worldwide standard for HDTV studio production and international program exchange.

The international focus on HDTV from the CCIR and SMPTE organizations drew attention from many companies. It was at this time that RCA began to focus on the utilization of HDTV as a studio production tool. After reviewing the initial findings of the SMPTE study group on HDTV from the late 1970s, Dr. Powers, through his role as chairman of the SMPTE Technology Group, set up a SMPTE working group on high definition electronic production (called WGHDEP) in 1983. Led by its chairman, Richard Stumpf, vice president of technology at MCA/Universal, this group of motion-picture cinematographers, technicians and television engineers has been responsible for the establishment of HDTV production standards that are currently in use today. The major focus of this group has been to develop standard parameters for the development of HDTV as an electronic production tool in both original production and post-production.6 Unlike the development of monochrome and color television, HDTV--with its equivalency to motion-picture quality--was getting participation from technical leaders in the motion picture industry.

In 1984, the Japanese government and industry, through the NHK, agreed on a 1125/60Hz emission standard called Multiple Sub-Nyquist Encoding (MUSE). MUSE is a bandwidth compression technique used for direct broadcasting via satellite transmission. Japan had already made the decision that DBS would be one of the first distribution media for HDTV in that country. Given the long-term commitment of Japan toward HDTV and the development of both transmission and production equipment, RCA began to increase its efforts in HDTV. And by 1984 RCA had developed a 750-line progressive scan 60Hz system that was in competition with the 1125-line, 60Hz system of NHK.

In January 1985, the principal parameters of the NHK 1125/60 system were adopted at the ITU interim working party meeting in Tokyo. And in February 1985, the SMPTE WGHDEP recommended to the ATSC and the U.S. Department of State HDTV production system parameters based upon the 1125/60 approach. Despite the recommendations of SMPTE and ATSC, RCA shifted its emphasis from HDTV production to IDTV and EDTV approaches for consumer reception. However, when RCA believed that improvements to the television signal--including the appeal of wider screens--had marketing potential, RCA combined its different technological approaches to IDTV and EDTV in its Advanced Compatible Television (ACTV) approach. ACTV was created as an alternative method of transmitting ATV signals. ACTV research began at the David Sarnoff Research Center in 1985.7

By the mid-1980s, CBS acquired a full 1125/60 HDTV system from Sony. CBS began to conduct numerous demonstrations of HDTV, some of which compared the effectiveness of the NHK MUSE8 bandwidth compression technique with the full bandwidth of the Sony high definition system. Around this same time, CBS began efforts to create an HDTV transmission system to be utilized exclusively for DBS. Ren McMann devised a two-channel approach in which one channel was NTSC compatible, while the other channel was high definition. The HDTV display in the home would output 1050-line/59.94Hz interlace signals.9

Other companies also began to investigate HDTV in the 1980s. For example, in Germany, Bosch began a project on HDTV under the direction of Dr. Ulrich Reimers. One of the earliest projects at the Bosch Laboratories was to measure the resolution capability of both television imaging pick-up tubes and display cathode ray tubes (CRT). In May 1986, the proposed 1125/60 HDTV production standard was submitted for adoption at the CCIR plenary meeting in Dubrovnik, Yugoslavia. By March 1986, the European Community organized enough support to have the process of standardization delayed at Dubrovnik for at least two years. Despite the lobbying efforts and free royalty considerations for HDTV equipment sold in parts of Europe to certain European consumer electonics companies--if they successfully lobbied for support from their country for 1125/60--the proposal was rejected at the CCIR meeting in May 1986.10 The impact of the 1986 CCIR plenary assembly was the postponement of a decision on the worldwide standardization of HDTV for studio production and program exchange, until the next CCIR cycle in May 1990.

The experience in Dubrovnik illustrated the politicization of the standards process--a lesson that had been learned once before with the development of the French SECAM standard as a non-tariff barrier to trade and the development of an indigenous industry for France. European countries felt they had an opportunity to prevent what they considered to be a "technology gap" in HDTV technology because of Japan's development of a working system. Just as SECAM was a method of building an indigenous industry for France, in a similar way, Europe hoped it could compete in HDTV on a global scale. However, a major difference between the example with SECAM and the situation with HDTV was that Europe realized that, unlike the development of color standards, no one country had the resources to develop the needed technology for an entire HDTV system. Considering the lesson of incompatible color standards that divided Europe, a strategic alliance was formed that would allow many Western European countries to compete in HDTV on a global scale.

In response to the outcome at Dubrovnik, in October 1986, many countries in Western Europe joined a Pan-European international strategic alliance called EUREKA. It included a project dedicated to the evolution of a European high definition system that could complement the Western European-based MAC System.11 The specific goals of EUREKA were to develop alternative standards for HDTV production that could be acceptable to both 60Hz (Japan, U.S., Canada) and 50Hz (Europe, Australia, Africa) countries and create an HD-MAC system that could compete with the Japanese HD-MUSE system.

In North America, the first announcement of the Sarnoff ACTV system was made at the Ottawa HDTV Conference in 1987. Also in 1987, the National Association of Broadcasters and Association of Maximum Service Telecasters petitioned the FCC to examine spectrum issues relating to HDTV and voiced concern over the possible reallocation of UHF spectrum to land mobile services (i.e., fire, police, industrial, CB two-way radio communications). As a direct result of the NAB/AMST petition, the FCC issued a notice of inquiry regarding Advanced Television Systems and also took the following actions:

• An FCC freeze on applications for new UHF stations in 30 of the top 40 markets.
• A freeze on the "UHF sharing" proceedings.
• Issuance of an order setting up a joint FCC-industry advisory committee on ATV to generate information and produce policy recommendations as a basis for future FCC rulemaking, with a September 1989 deadline for completion of its final report to the FCC.

This advisory committee represented a departure from the FCC's previous method of issuing hearings and demonstrations. In the area of HDTV, the impact of the Reagan Administration's deregulation was to make the FCC dependent on the private sector for assistance in framing the issues, planning and implementing testing procedures and offering recommendations on standards. The framework of the Advisory Committee established three subcommittees that are comprised of working parties. The subcommittees were placed into three categories consisting of planning, systems and implementation. These subcommittees coordinated their activities with other organizations such as the ATSC, Advanced Television Test Center (ATTC), Center for Advanced Television (CATS) and the National Cable Television Association-Cable Labs (NCTA).

In addition, by 1987 efforts were underway in the United States to standardize the 1125/60 system as an American HDTV studio production standard. Work by ATSC and WGHDEP culminated in a passed resolution on 1125/60 signal parameters, first by the SMPTE in 1987, and secondly, by the ATSC in 1988. However, two major broadcast organizations--the National Association of Broadcasters and The Association of Maximum Service Telecasters--voted against the 1125/60 proposal on the grounds that it diverted local broadcasters from the opportunity to participate in the full range of possible advanced television systems.12

In 1988, the FCC advisory committee on advanced television service (called ACATS) presented its interim report to the FCC. At that time, almost 20 proposed ATV systems had been under consideration by the FCC. This was considerably more systems than were presented to the FCC at any time in the standardization of either monochrome or color. However, on September 1, 1988, the FCC narrowed the field significantly by its tentative decision setting guidelines for an HDTV terrestrial broadcast transmission standard.13 In its "Tentative Decision and Further Notice of Inquiry" the FCC announced fundamental policies and defined the boundaries for the development of advanced television: HDTV must be compatible with existing NTSC service; it would have no additional spectrum allocations outside the VHF and UHF bands; and it would begin an inquiry on the relative advantage of a variety of allocation schemes for HDTV transmission, including: one 6 MHz channel; one additional 3 MHz augmentation channel, not necessarily contiguous to the main channel; and one additional 6 MHz channel, not necessarily contiguous, as either an augmentation channel or a simulcast channel for HDTV during a transition period.14

Through the tentative decisions, the FCC eliminated from consideration for standardization transmission systems that were incompatible with existing NTSC. This effectively ruled out the possibility of a terrestrial broadcast system based on the Japanese HDTV Multiple Sub-Nyquist Encoding technique called MUSE-E.15 In addition, the FCC decision ruled out the consideration of utilizing the spectrum band above 1 GHz as augmentation channels. 16

By the fall of 1988, HDTV-related activities had greatly intensified in the U.S. In October 1988, Secretary of Commerce William Verity appointed an advisory committee on advanced television to study the potential impact of HDTV on U.S. industries. That same month, the American National Standards Institute accepted SMPTE 240M (the HDTV 1125/60 production standard approved by SMPTE) as an American national standard for studio production. In December 1988, the U.S. Government once again played an active role in the technological development of television. That month, the Defense Advanced Research Projects Agency (DARPA) announced the availability of grants totalling $30 million for two HDTV projects. One project involved the development of HDTV display devices. The other project involved the development of HDTV video signal/computer processors.

In March 1989, North American countries established a strategy that appeared to be similar to that taken by the Europeans at Dubrovnik in 1986. The North American National Broadcasters Association recommended that the CCIR not adopt a worldwide HDTV studio standard at the conclusion of its present study period (in 1990). NANBA also made recommendations to encourage the identification of a common image format and the continuation of testing terrestrial formats, in order to establish potential relationships between emission standards and studio standards.17 Also in March 1989, The Sixth World Conference of Broadcasting Unions recommended studies based on two different approaches towards the achievement of a worldwide production standard. These two approaches involved systems with a common data rate and systems with a common image structure.18 On a similar note, in April 1989, an ANSI appeals board rescinded its earlier approval of the SMPTE 240M after a second appeal by Capital Cities/ABC. The 1125/60 system once given enormous support by the U.S. was increasingly challenged by broadcasting and cable industry groups.

A variety of U.S. Government and industry proposals were introduced in the first half of 1989. Secretary of Commerce Robert Mosbacher outlined a plan that would relax antitrust laws and change the capital gains tax to cultivate HDTV manufacturing efforts by American firms. By May 1989, eight Congressional bills were introduced on the topic of HDTV (six House bills, two Senate bills) advocating some form of government aid for strategic alliances, antitrust relaxation, and national cooperative research. Other interinstitutional players became more involved as the HDTV issues broadened to include other industries, such as the computer and semiconductor industries. By mid-1989, industry groups such as the American Electronics Association (AEA) introduced an HDTV business plan to the U.S. Congress. The AEA plan calls for Congress to provide $1.3 billion in grants, loans and credits for the development of a consortium providing a strategic alliance between government and American institutions to research, develop and manufacture HDTV products.19

In June 1989, DARPA awarded its first round of HDTV display technology development bids to Newco, Inc., Raychem Corporation, Texas Instruments, Inc., and Projectavision. Photonoics Technology, Inc. was awarded a DARPA contract on flat screen display technology.20 According to DARPA, more announcements regarding HDTV image processor awards were forthcoming. And an additional $50 million for DARPA's HDTV research project was recommended by a Congressional subcommittee working on the 1990 defense budget.21

In August 1989, a new hearing was set for a third appeal on the ANSI HDTV issue, this time validating the SMPTE 240M 1125/60 HDTV production standard. The ANSI/SMPTE standard stamp is important to the manufacturers and users of HDTV production equipment because of the need for common signal parameters for different pieces of the production process. Cameras need to be able to match with high definition videotape recorders, switchers and other signal processors. If standards do not exist, the risk of technical obsolescence or incompatibility becomes even more acute.22

Standardization And Commercialization: The 1990s

In March 1990, FCC Chairman Alfred Sikes stated the exact goals for the FCC in the implementation of an HDTV transmission standard. The FCC's intent is to select a simulcast HDTV standard that would be compatible with the current 6 MHz channelization plan, but utilizing design principles not limited to the constraints of existing NTSC technology and independent of NTSC. The target for the standardization selection by the FCC is the second quarter of 1993. With the intention of the FCC to standardize a simulcast system, the possibility of approving an augmentation system was eliminated. Augmentation systems represent substantial spectrum availability and utilization problems.23 The FCC formalized what the marketplace was leaning towards in the development of simulcast approaches to HDTV terrestrial broadcast. Developers of simulcast approaches include Zenith, the David Sarnoff Research Center/North American Philips consortium and NHK.24

The augmentation approach to terrestrial broadcasting of HDTV allowed for existing television channels to continue to broadcast NTSC signals on their 6 MHz allocations. However, an additional 3 MHz or 6 MHz would be assigned to each station for the transmission of additional information that served to augment the picture resolution of the main NTSC transmission. In addition, side-panel information could also be transmitted in the augmentation channel providing a 16:9 aspect ratio. Existing sets would continue to receive NTSC, while new HDTV sets would combine the NTSC channel with the augmentation channel to receive the full HDTV transmission.

The simulcast approach also allows existing television channels to keep their existing channel allocations. However, each station would also be given an additional 6 MHz channel to transmit a bandwidth compressed HDTV signal. Existing NTSC receivers would take the NTSC signal, while new HDTV sets would pick up the HDTV signal. Simulcast is deemed superior to augmentation by some experts due to the fact that simulcast makes more efficient utilization of RF spectrum. In addition, simulcast systems are not constrained by the limitations of various NTSC artifacts. However, if the testing of simulcast systems indicates that such a system is not feasible for broadcast, then the commission may still select an augmented version of an extended-definition system.

One important concern of an HDTV service is making sure that existing NTSC channels are not interfered with.25 The Commission also stated that it would not make a decision on extended definition television until it reached its primary goal of standardizing an HDTV standard based on a simulcast system approach. Therefore, the FCC decision did not stifle the continued development of EDTV systems. Proponents of EDTV systems are Sarnoff/Philips, NHK, Faroudja Laboratories, Production Services, and the Massachusetts Institute of Technology.

Regarding the distinction between HDTV, EDTV and IDTV, the FCC defined each technology in accordance with industry-wide accepted definitions of these terms that were developed by the U.S. Advanced Television Systems Committee in 1989. As stated earlier, HDTV represents a signal that provides at least double the horizontal and vertical resolution of existing 525 NTSC on a 16:9 aspect ratio, increased color fidelity and CD-equivalent sound quality. EDTV represents a signal that provides enhanced pictures to the home, including widescreen to special EDTV receivers, while at the same time providing a picture that could be decoded by existing NTSC receivers. IDTV represents a signal that improves NTSC without changing the transmitted NTSC signal. This includes improvements such as linedoubling and ghost-cancellation. According to FCC Chairman Sikes, IDTV is an area that does not require FCC action to implement improvements to NTSC.26

Conclusion

Given the history of the standardization of both monochrome and color television, the adoption and diffusion of HDTV for terrestrial use will be dependent on the development of standards set by the FCC. Historically, the potential for commercialization in the American television industry has been most realized after the presence of technological standards in the marketplace. If one considers the plethora of newer distribution technologies for HDTV on the horizon beyond broadcasting; the continued efforts of harmonization by international standards groups such as the CCIR; and the versatility of HDTV for a variety of business and industrial applications, the future for HDTV appears very promising indeed.

Notes

1. For more information on the early NHK research efforts see T. Fujio, "A Study of the High-Definition TV System in the Future," IEEE Transactions on Broadcasting BC-24.4 (December, 1978), pp. 92-100.

2. See Kenneth Donow, HDTV: Planning for Action (Washington D.C.: NAB, April, 1988), p. 18.

3. The original NHK HDTV system consists of the following parameters: 1125 lines, 60HZ field rate, 2:1 interlace, and 5:3 aspect ratio.

4. This is the bandwidth where direct broadcast satellite transmissions are located.

5. The charter members of the ATSC are the National Association of Broadcasters, Electronic Industries Association, the Institute of Electrical and Electronics Engineers, the National Cable Television Association and the Society of Motion Picture and Television Engineers; the chairman is James McKinney, and the executive director is Robert Hopkins.

6. This author has actively participated as a member in this SMPTE Working Group since 1984.

7. One of the key new developments associated with the ACTV system that distinguished it from all other previous work was the method of using the "Fukinuki Hole" to separate both high and low frequency components of the signal in the edges or "wings" of the image. The Fukinuki Hole was named after a Hitachi scientist named Takahiko Fukinuki who developed a scheme of utilizing the unused portion of the NTSC spatio-temporal spectrum signal with additional information that may be used for transmission of an ATV signal. For more information on the Fukinuki Hole see also Mark Schubin, High Definition Glossary (New York: HDTV Group/Videography, 1988).

8. MUSE was originally a term to describe NHK's DBS transmission scheme. Currently, MUSE refers to a family of ATV transmission proposals. The Sub-Nyquist portion of the name refers to the fact that MUSE is a sub-sampling system and as such, has been subject to motion artifacts.

9. The CBS HDTV DBS system utilized frame conversions that were initiated prior to the uplink from the source transmission.

10. Adam Watson Brown, "The Campaign for High Definition Television: A Case Study in Triad Power," Euro-Asia Business Review 6.2 (April, 1987); "NHK Offers Deal on HDTV," Broadcast 30 (August, 1985).

11. MAC signals are a family of television signal formats that separate color and luminance (monochrome) signals into component channels which are time compressed so that active line time remains constant. Most MAC signals contain digital audio and data channels. MAC has achieved most of its popularity as a satellite transmission system for Europe.

12. See E. Feldman, "Advanced Television Update," Via Satellite (March 1988); Donow, HDTV: Planning for Action.

13. "FCC Writes a First Draft for HDTV," Broadcasting 115.10 (September 5, 1988), pp. 32-34.

14. Federal Communications Commission, MM Docket 87-268, 88-288, 37462. As quoted in Federal Communications Commission, Economic Factors and Market Penetration: The Working Party 5 Report to the FCC Planning Subcommittee on Advanced Television Service (Washington, D.C.: FCC, 9 May 1988).

15. MUSE-E stands for a MUSE system optimized for broadcasting rather than satellite transmission. It is a non-compatible proposal that occupies 8.1 MHz base bandwidth, requiring four fields to build up a full resolution picture. The system requires motion compensation for display.

16. "High Definition TV," Broadcasting 115.13 (September 26, 1988), p. 20.

17. The NANBA members are ABC, CBS, CNN, NBC, PBS (U.S.); CBC and CTV (Canada); and Televisa (Mexico). See David Hack, "High Definition Television," Congressional Research Service, CRS Issue Brief (Washington, D.C.: U.S. Library of Congress, 5 June 1989).

18. The World Conference is made up of nine international regional broadcasting unions.

19. John Gatski, "Capitol Hill's HDTV Plans Draw Positive Reactions," TV Technology 7.7 (June, 1989), p. 1.

20. For more information on the DARPA HDTV initiative see "DARPA Allocates Money for HDTV," TV Technology 7.9 (August, 1989), p. 3.

21. Chip Cavanagh, "ATV Funding May Increase; DARPA selects 1st Companies," Television Broadcast (July, 1989), pp. 1, 23.

22. For more information the ANSI standards battle for SMPTE 240M see Alan Carter, "ANSI Sets Third Hearing Date for SMPTE 240M," TV Technology 7.9 (August, 1989), p. 1.

23."FCC to Take Simulcast Route to HDTV," Broadcasting (March 26, 1990), pp. 38-40.

24. Alan Carter, "FCC: Broadcasters Will Simulcast HDTV," TV Technology 8.4 (April, 1990), p. 6.

25. David Hughes, "FCC Decides to Push for 'Pure' HDTV," Television Broadcast 13.4 (April, 1990), pp. 1, 55.

26. "FCC to Take Simulcast Route to HDTV," Broadcasting (March 26, 1990), pp. 38-40.

Corey P. Carbonara holds BA and MA degrees from the University of Iowa, and a PhD from the University of Texas. He received his EEC from the Omega School of Communications. Carbonara currently is professor of telecommunication and new video technologies project director at Baylor University in Texas. He has held positions at Sony Broadcast Products Company, Caterpillar Tractor Company (video producer), and Columbia Pictures in Chicago.
Carbonara has represented the U.S. as a technical consultant on State Department and other committees. He is vice chairman of the systems subcommittee for the FCC advisory committee for advanced television. He is a member of the U.S. Advanced Television Systems Committee, and of SMPTE. He served as an HDTV technical consultant to the National Association of Broadcasters for its 1987, 1988 and 1989 conventions. He is a member of the International Television Association, the Institute of Electronic and Electrical Engineers, the Society of Broadcast Engineers, the American Society of Lighting Designers, the National Association of Broadcasters, and other professional groups. Carbonara has published articles in the IC/2 Working Paper, The Business of Film, Religious Broadcasting and the HDTV Newsletter. In addition, he has been a speaker on the subject of HDTV and other issues at conferences and conventions around the country.

Forecast Of The HDTV Market
By David E. Mentley and Joseph A. Castellano

Predicting the future market size of HDTV, or of any technology-driven product, is formidable and often inaccurate because of two tendencies: the first is to be far too optimistic in the near term (one to two years) and the second is to be too pessimistic in the long term (five to ten years). We will try to keep these tendencies in mind throughout this analysis.

It is very important when considering the future of HDTV to realize that it encompasses three domains: production, transmission and display. The display domain is the primary area of concern of this report and it refers to the equipment, format and technology (such as flat panel displays, projection systems or direct view CRTs) used to bring the images to the consumer. Television offers a unique opportunity for the market forecaster. The existing U.S. annual market of approximately 24 million color television receivers provides some sort of a baseline. The abundance of developing HDTV-related technologies provides an attractive means for substituting or supplementing the market. The inertia of the political, regulatory and international trade factors restricts the market. Any meaningful forecast must weigh all of these components.

Display standards are affected by the transmission and production specifications and also by the limits on display technology needed to put the receivers into millions of homes. In general, if we look at the projected manufacturing costs and assume that selling prices are a function of these costs, then we can get a perspective on the penetration of HDTV receivers into the American consumer market. We need not know definitively whether the U.S. standard will be 1250, 1125 or 762 lines in order to estimate manufacturing costs for HDTV monitors. Stanford Resources has done extensive cost modeling on displays of the size necessary to satisfy general high definition requirements.

Several other factors were considered in addition, in developing the market forecast for HDTV sales in the United States:

• Consumer acceptance: direct view vs. projection
• Selling price of the receiver
• Flat panel display availability
• Availability of programming

Full implementation of a widely accepted HDTV delivery system will require a number of advances in several different technologies: displays, signal processors, decoders and studio equipment. All of the basic technologies are currently available to demonstrate HDTV and to provide limited numbers of studio systems, but the full implementation using existing technology results in a configuration which is either too large, requires too much of the frequency spectrum or is too expensive for the market today.

There are two fundamental ways of displaying information in a large area high resolution format: projection and direct view. Within each category there are dozens of technologies and hundreds of possible variations on these technologies; each of these technologies has advantages and disadvantages. Projection technology relies on a high power light source, but can be used to make a display which weighs much less than a direct view CRT. The high power aspect of the projection technologies has an impact on the design of the equipment and the limitations on performance and lifetime. Mirrors, lenses, optics and cabinetry are non-display components needed by projectors and which comprise a larger fraction of the total cost than the display elements themselves. Direct view technologies may be active (light emitting) or passive (light reflecting), and may or may not involve high power, but they use a larger measure of direct materials relative to projection type displays, at least in their current embodiments.

Direct view CRTs suitable in size for usage in HDTV systems are now nearly at the upper limit for manufacturability, in terms of existing factory equipment, bulb weight, shadow mask handling ability, and related factors. This could change in the next few years but, currently, effort is being extended to push direct view CRT technology to 35-inch (but not much beyond 40-inch) diagonal tubes.

Direct view flat panel displays, on the other hand, are currently nowhere near the minimum requirements for size, pixel format, color, luminance, efficiency, speed and manufacturing cost to be used in an HDTV system. These features are changing, but the final result will only come after many more hundreds of millions of dollars are spent on research. This is where the tendency to be too pessimistic in viewing the future must be weighed.

Projection systems based on both CRTs and liquid crystal cells are currently positioned for acceptable application to high definition televisions, although there are shortcomings. Projection televisions have not been a growth segment during the past few years. Is there a case to be made that it will take a major stimulus, such as HDTV, to make this market take off? Not likely. According to Thomson Consumer Electronics, projection television has penetrated about two percent of households 12 years after introduction. Color televisions and camcorders both surpassed nine percent penetration at the 12-year mark. Why have projection televisions remained at a flat six to seven percent of total sales for the past ten years? Possibilities are:

• Picture quality does not measure up to direct view.
• Receivers are too expensive for the market to grow.
• There are not enough homes with the space to accommodate them.
• Consumers are irreversibly biased against projection technology.

Picture quality is improving rapidly as projection optics, screen optics and projection tube technology is pushing forward. The problem is that there is a ten year history of poor quality to overcome in the consumer's perception and with little high definition source material, there will be no incentive for the consumer to purchase a high definition projection receiver/monitor.

The price issue is one which does not get nearly enough attention. Almost seventy-five percent of the U.S. color television market is comprised of sets with 25-inch diagonal and smaller tubes. Can HDTV be expected to convert many of the 16 million consumers who pay $200 to $800 for a television into consumers who will pay $2000 to $5000?

The space question is one which affects both projection and direct view CRT. How many apartment dwellers have the space to devote to a 50-inch diagonal projection system? Direct view CRTs with a 110 degree deflection angle peak out at about 28 inches of cabinet depth for a 40-inch diagonal screen. There is also a weight of about 180 pounds for this 40-inch set.1 Both of these figures limit the salability of receivers.

The issue of inherent bias against projection is much harder to quantify. The direct view image is perceived as being more "tangible" than projected images. Just as computer users like to make a hard (paper) copy of important data, viewers of a direct image may feel more comfortable with an image which is generated directly. Other factors are room lighting, viewing angle and the issue of front vs. rear projection. This last issue is worthy of further note. The first television projection systems which came on the market were front projectors. But the need to set up a separate screen made the set a "two-component system" which was essentially unacceptable for the home. Eventually, the front projector was replaced by the single component rear projector.

Technology Assessment

Based on the current state of development of direct view flat panel displays, it appears that direct view flat panel based HDTV capable televisions are at least five years away from introduction and probably ten years from high volume markets. The three most suitable flat panel technologies based on the status quo are:

• Active Matrix Liquid Crystal Display (LCD)
• Plasma Display Panel (PDP)
• Electroluminescent (EL) Panel

Significant challenges remain in each of these relatively mature technologies before a full color video display greater than one meter in diagonal measurement can be economically produced.

Liquid crystal displays which can now be made with small size screens (two to five inches diagonal) for television are the current favorite for development efforts in Japan, to some degree in Europe and with a token effort in the U.S. The major obstacles center around the very capital-intensive nature of the production process. Additionally, no active matrix prototype larger than 16 inches has been built to date. Processes to step and repeat over several feet using a six-inch exposure pattern will mean that each layer of the required six will take 40 or 50 exposures. This translates to many hours of production per panel on a machine which costs over $1 million.

Superficially examining the economics, one can see that some breakthrough is required and even the direction of the breakthrough is not clear at this time. Our detailed cost model which analyzes the processes, material, labor and overhead necessary to build a factory to make 29 x 35-inch full color, active matrix LCDs indicates a factory cost of about $850. This optimistic model assumes a seventy percent yield and a capital expense of about $130 million. A television receiver based on such a display, using an allocation of twenty-five percent of the selling price for the display cost, would retail for around $3400.

A full color plasma display panel has been demonstrated in a 33-inch format, up from 20 inches only one year ago. The progress by just a single laboratory at NHK in Japan has been remarkable. Plasma displays may be manufacturable in large area format without the expensive equipment needed by active matrix LCDs. Much more investigation is essential. Dramatic improvements in phosphors are still required for plasma to be a success in the television market. Unit costs for the high voltage driver circuits must be brought down by nearly an order of magnitude for mass commercialization.

Electroluminescent display technology is still lacking in the presentation of full color, but resolution of this problem will open up the potential for medium sized, high resolution displays in the near future. As with plasma displays, driver costs must be reduced by nearly ten times for mass commercialization. Processing will still be capital intensive due to the large area, thin film deposition processes needed in production.

Projection technologies are the most likely to fulfill the market requirements for HDTV in the near term. Market acceptance of a projection display as a class is still not guaranteed. Performance improvements are coming rapidly from both projection CRTs and LCDs, so consumers may adopt the technology at a greater rate than the relatively flat growth of the past decade.

Projection CRTs are not capable of delivering a high brightness, high resolution image suitable for HDTV in their current state of engineering. Light valve systems are capable of both the resolution and brightness, but the high maintenance and acquisition costs put the price out of reach of nearly all segments of the consumer market.

Projection LCDs bring high resolution and medium brightness together. By using a 500 watt metal halide lamp, Sharp is able to achieve a 250 lumen image with nearly 800 lines of resolution in a relatively lightweight and compact front projection system. The compact optics and LCD cells will eventually allow for a low cost system, but the current retail price of $5000 limits market acceptance. Further increases in screen gain and optics may eventually merge projection and direct view, at least in terms of performance. Technology aside, front projection faces a great market risk in the home. While it has been acceptable in the professional market and in bars, particularly for sports, the home market has not yet embraced front projection despite several opportunities.

With dozens of display technologies vying for a share of the nearly 100 million color television receivers built every year, the progress in cost effective performance is sure to continue.

Market Forecasts: 1990 To 2005

Stanford Resources recently performed a detailed analysis of the worldwide television industry and prepared a forecast of the market for HDTV and other advanced television formats.2 The study began with a thorough search of the technical literature and with personal discussions among the knowledgeable experts in the field. This enabled us to develop a better understanding of the fundamentals behind operation of devices being developed or products being manufactured. It provided the basis for describing and explaining the technologies used in various advanced television systems. Information was also obtained from major domestic and international display conferences. The next step was to determine the number of units of conventional television product shipped into each of the major geographic market areas. This information was gathered through an exhaustive search of available data collection sources.3

For the major market regions such as the U.S., Japan, and Western Europe, the information on either shipments made (in U.S. dollars) or units sold was obtained by a combination of telephone interviews and statistical analysis. It was also possible to collect data on the average selling prices of the various products and product categories in these major geographic market areas. Generally, it is more difficult to obtain statistics on certain products or product categories in other geographic regions. In these cases, it was necessary to use a world demographic model which we developed previously. This model is based on the relationship between the number of television sets in use and the average income available to the consumers of each nation in the world with a population of ten million people or more.

Plots of the available income in billions of dollars (obtained by multiplying the per capita income by the population) vs. the number of television sets in use gave the predictable result; excellent correlation with a straight line which follows the equation: I = 11.36T + 70 where I is available income and T is the number of television sets in use. This really comes as no surprise because one would expect that those geographic areas with the highest available income should be in a position to buy the most consumer electronic products. This enables us to compute what we call a theoretical market share potential for the purchase of consumer electronic products. Thus, when little or no information was available on specific display based products in certain parts of the world, we used an estimate for the "rest of the world," based on this demographic model.

In determining the quantity and average selling prices of the various products sold into the particular geographic market areas, we not only used statistical sources but also conducted interviews with management and marketing personnel in organizations which make or use the systems.

Armed with the information and intelligence gathered through this research, we then proceeded to compute the degree of penetration of the advanced television sets into the various market segments. Projections to future years were then computed using econometric time-series modelling programs. Basically this involves establishing an average selling price (ASP) erosion relationship for each ATV technology type while at the same time determining the growth pattern of the various products. Thus, when a log-log plot of cumulative volume vs. average selling price in constant dollars is used, the resulting straight line has a slope which depends on the magnitude of the selling price.

A forecast of the worldwide market for all types and all formats of color television sets is given in Table 1. The table presents unit shipments and includes historical data for 1987 and 1988 as well as an estimate for 1989. The data is given for both direct view sets and projection television sets; annual growth data and compound annual growth rate (CAGR) figures also are shown. In 1990, some 86.2 million sets will be sold with only about 1.1 million of these being projection systems. The market is forecast to grow to 144.8 million sets in the year 2000 and 177 million sets in 2005. We predict that large screen, direct view sets will outpace the projection sets in the first five years of the next century. Direct view types are generally more compact (though heavier), brighter and have wider viewing angles than projection systems.

Based on the factors and methodology explained above, a worldwide market forecast of HDTV sets was prepared and the results shown in Figure 1 and Figure 2. Cathode ray tube-based sets will lead the way with growth to 938,000 units in the year 2000, from 1000 units in 1991 (see Figure 1). The market for direct view CRT sets will approach two million units in the year 2005. Flat panel-based HDTV is not expected to appear until after 1995; significant growth will not occur until after 2000. Both direct view and projection will coexist because projection systems will be less costly to manufacture than television-on-a-wall systems (Figure 2).

Table 1
Worldwide Color Television Market: All Types and Formats (Thousands of Units)

Annual Direct Annual Annual
Year All Sets Growth View Growth Projection Growth
1987 65,042 ---- 64,029 ---- 1,013 ----
1988 72,433 11 % 71,390 11 % 1,043 3 %
1989 79,963 10 % 78,892 11 % 1,071 3 %
1990 86,269 8 % 85,165 8 % 1,103 3 %
1991 90,960 5 % 89,818 5 % 1,142 3 %
1992 96,241 6 % 95,061 6 % 1,180 3 %
1993 101,841 6 % 100,623 6 % 1,218 3 %
1994 107,712 6 % 106,457 6 % 1,254 3 %
1995 114,081 6 % 112,790 6 % 1,292 3 %
1996 120,668 6 % 119,336 6 % 1,331 3 %
1997 126,372 5 % 125,008 5 % 1,364 2 %
1998 132,447 5 % 131,058 5 % 1,389 2 %
1999 138,565 5 % 137,159 5 % 1,406 1 %
2000 144,844 5 % 143,440 5 % 1,404 -0 %
2001 151,267 4 % 149,881 4 % 1,386 -1 %
2002 157,590 4 % 156,224 4 % 1,366 -1 %
2003 163,974 4 % 162,640 4 % 1,334 -2 %
2004 170,393 4 % 169,098 4 % 1,296 -3 %
2005 177,085 4 % 175,844 4 % 1,242 -4 %

Compound Annual Growth Rates
1990-1995 6 % 3 %
1995-2000 5 % 2 %
2000-2005 4 % -2 %

The value of HDTV shipments is computed by multiplying the ASPs by unit shipments. The world market for all types of HDTV sets will grow to $5.1 billion in 2005, from $2.6 billion in 2000 and $1.3 billion in 1995. This represents a CAGR of about fourteen percent, a reasonable figure which is consistent with historical patterns and the kind of price structure expected. Again, CRT-based direct view systems will lead the way.

During 1989, a number of other forecasts of the market for HDTV were made. Most tended to be wildly optimistic about the future size of the market. One such study was commissioned by the Electronic Industries Association in conjunction with Richard Nathan Associates. This study showed a growth from 700,000 units in 1993 to 12.1 million units in 2003, a compound annual growth rate of over twenty-five percent! Given the expected high price of HDTV sets in the early years of introduction (1992 to 1995), it is hard to imagine a 700,000 unit market just three years from now let alone the phenomenal growth forecast. Apparently, this group believes that HDTV will be such a dramatic improvement over conventional sets that consumers will flock to stores in an effort to purchase them.

Will HDTV be to conventional television as revolutionary as color was to black and white? To answer this question, one need only look at the history of color television market development. The first color television sets came on the market in 1956-57. The change from black and white (monochrome) to color was quite dramatic. And yet, the replacement of monochrome sets by color took many years to accomplish. A look at the statistics is quite revealing in this regard.

According to data compiled by the U.S. Department of Commerce, color television shipments on the U.S. market grew from 2.646 million units in 1965 to 4.822 million units in 1970 and 6.485 million units in 1975. This represents a compound annual growth rate of about ten percent over the ten year period. The reasons were: (l) the high price of the sets relative to monochrome sets, and (2) a lack of adequate color programming in the mid-1960s. It was not until the late 1960s that automated assembly of color television sets began. This resulted in a dramatic reduction in manufacturing cost and, ultimately, more affordable prices. Coupled with a major shift to color programming across all networks in the early 1970s, higher sales materialized. Still, it was not until 1978, over 20 years from product introduction, that the number of color sets in use exceeded the number of black and white sets.

Although HDTV will provide a greater number of pixels on the screen and a completely new format (a screen which is twice as wide as it is high), the change from conventional television will not be as great as the change from monochrome to color, at least in our opinion. Several focus group surveys have already shown that users do not see a revolutionary impact. The second problem faced by HDTV is that the price of such sets today and for the next few years will be well out of the reach of most consumers. And the third--and perhaps most important--problem is the lack of an international broadcast standard. The battle among the Japanese, Europeans and Americans over which format to adopt has already been raging for several years; no immediate resolution appears to be in sight.

All of these factors were taken into consideration in the preparation of the forecasts presented above. Although it is very difficult to predict the impact of future unknown world events on any market segment, it is possible to make reasonable and careful estimates of the future market based on a combination of historical patterns of market development and an evaluation of HDTV technology evolution.

Conclusions

Certain segments of the market, notably the videophiles and manufacturers facing narrow profit margins, are ready for HDTV. But it is clear that the display technology has a long way to go in order to provide the price points necessary to convert a significant portion of the existing color television market into HDTV.

Two other points must be made. First, the global television industry, including programming, manufacturing, service, delivery and peripherals is immense. The endeavor to improve and differentiate television products through development by hundreds of firms is so vigorous that progress is inevitable. So technological barriers will fall eventually. And second, even the best forecast can be invalidated overnight by a single FCC decision.

Notes

1. E. Yamazaki and K. Ando, Projection Display Technology, Systems and Applications, SPIE, Bellingham, Washington, Volume 1081, 1989, p. 30.

2. High Definition Television: Markets, Technologies, Strategies; Stanford Resources,
San Jose, California, 1990.

3. Some of these sources included: U.S. Department of Commerce, International Trade
Commission, Electronic Industries Association of both the U.S. and Japan, Electronics
Magazine, Electronic Business, Asian Sources (Electronics, Components, Timepieces),
Electronic News Financial Fact Book, International Marketing Handbook, Japan Economic
Yearbook, Japan Statistical Yearbook, United Nations Yearbook of Industrial Statistics,
Euromonitor Publications of European Marketing Data and Statistics, and the U.S.
Industrial Outlook Directory.

David E. Mentley has more than 13 years of experience in the electronics industry, and now specializes in strategic planning for high technology manufacturing companies. He holds a BS in ceramic science from Alfred University, and an MS in materials science and an MBA in marketing from the University of California, Berkeley. Mentley has worked as a product engineer at Texas Instruments and as a research engineer at AVX Materials; he also was product marketing manager for the EPID division of Exxon Enterprises, Inc. Mentley joined Stanford Resources in 1984 and currently is director of display industry research. He specializes in studies of flat, high information content displays. Mentley is co-author of the popular industry survey report series that includes Flat Information Displays: A Strategic Analysis; Enhanced LCD Flat Panel Displays; Large Screen Information Displays; High Definition Television; and The West European Display Market Report. He is a contributing editor to the monthlies Electronic Display World and EDIS Impact Reports. He is active the Society for Information Display.

Joseph A. Castellano has over 30 years of diversified industrial experience in research, market development and management. He holds a BS from the City University of New York and an MS and PhD from the Polytechnic Institute of New York. In 1965, Castellano joined RCA, where he performed and managed R&D programs in liquid crystal displays and was the recipient of several achievement awards. Castellano has held positions at Princeton Materials Science, Inc.; Fairchild Camera and Instrument Corp.; and Exxon Enterprises, Inc. Castellano is now president of Stanford Resources, Inc., a market research and management consulting firm he formed in 1978. He manages numerous major custom research projects in the electronics field and related industries. Castellano has published over 35 scientific and technical papers and holds 15 U.S. patents. He is editor-in-chief of Electronic Display World and The Electronic Display Industry Service, and is the founder of the annual Flat Information Display Conference.

HDTV: The American Agenda
By Jerry K. Pearlman

High-definition television...never before has there been a product that has received so much attention--an estimated five years before it is even produced. From the trade press to network news programs… from business and consumer magazines to virtually every daily newspaper in the country… HDTV continues to be widely covered. It has captured the imagination of editors and producers, all of whom seem to want to look at the “American Agenda” and make HDTV a broad proxy for American competitiveness. Understandably, that’s a tall order--and an especially difficult assignment for the poor reporter who covers pork belly futures one day and has to write a major HDTV feature the next. Many reporters do a solid job of covering basic HDTV news, and we all know there’s been plenty of it. But HDTV is a complex subject: a technology story, a public policy story, a lifestyle story--it's science, politics and marketing all rolled up together. And, of course, coverage of HDTV is only as good as a reporter’s sources.

Nevertheless, the wealth of bad information on the subject of HDTV really starts with the so-called experts, those who are trying to create an American Agenda out of the HDTV whole cloth. Two recent examples of such expert misinformation are the General Accounting Office’s HDTV report to Congressman Markey’s House Telecommunications Subcommittee (dated March 1990, but released publicly in late April); and an extensive Booz Allen study, also released in April, commissioned by the American Electronics Association’s high definition systems task force. I’ll be referring to these reports below as I address some of the myths surrounding the HDTV American Agenda.

There is an American Agenda for HDTV and it’s set by the Federal Communications Commission. In March 1990, FCC Chairman Al Sikes reaffirmed what we have believed all along: that true high-definition television, and not an “enhanced” television approach, is the kind of new broadcast system that will best serve American television viewers and broadcasters. Systems that offer modest improvements to current television, in the form of doubled lines and/or wider pictures, will be tested but will not be authorized for broadcast unless a true HDTV system does not pass the testing process.

Most important, he also said that the FCC will choose a “simulcast” transmission system when it selects an HDTV broadcast standard in the second quarter of 1993. “Simulcast,” as many experts know, means that conventional television sets will continue to receive conventional television signals on current channels, and new HDTV receivers will receive HDTV signals (often of the same program material--like movies or sports) on additional 6 MHz channels in the VHF and UHF television bands.

Of course, the new HDTV receivers will be dual-mode, also able to display current low-resolution NTSC television. Simulcast channels were used in Britain and France to introduce color television. Old channels continued to broadcast black and white television by different standards for 20 years.

Zenith is one of the three simulcast proponents to have demonstrated a system. The others are General Instrument and NHK (the government-owned Japanese broadcasting system). Three other proponents--including MIT and the consortium of Thomson, Philips, NBC and the Sarnoff Labs--are currently still in the race, but have not yet been able to demonstrate simulcast HDTV systems.

The FCC timetable calls for testing of the various systems beginning this fall (1990) and carrying through 1991 and into 1992, with over-the-air tests in the second half of 1992. The FCC’s Blue Ribbon Advisory Committee on Advanced Television Services is expected to make its recommendation to the FCC on the HDTV broadcast standard in the fourth quarter of 1992, followed by a final FCC decision in the second quarter of 1993. If the timetable holds, HDTV receivers should start hitting the market in the U.S. by late 1993, and many will be available by 1994 (more on this subject below).

In presenting Zenith's view of the American Agenda for HDTV, I'll touch on many of the myths and realities surrounding the subject--transmission system myths, display technology myths, and, finally, some of most widely misunderstood issues, or popular myths.

The First Transmission System Myth

The General Accounting Office report to Congressman Markey’s committee (on the effects of HDTV standards on U.S. entertainment industries) says HDTV has “about four to five times the information as current television and, as a result, transmitting an HDTV signal requires a wider bandwidth than is available for current television....Given current regulatory and technical constraints, even with the adoption of an over-the-air transmission standard, broadcasters will not be technically capable of delivering a picture of comparable quality to that of cable and satellite. As a result, broadcasters may be at a longterm competitive disadvantage with these other media.”

Most of this is simply not true. Satellite and cable are not going to broadcast a better HDTV signal than over-the-air, for two reasons. First, with a good compression algorithm, a 6 MHz signal looks almost exactly like 30 MHz. Zenith showed the two side-by-side at the National Association of Broadcasters’ show this spring and professional broadcasters could not pick the compressed version.

Second, the government report would lead one to conclude that satellite and cable can go it alone and use a different system than over-the-air broadcasters. This idea makes no marketing sense; it won’t happen in the U.S. If a U.S. cable or satellite system were to adopt a Japanese or European wide-bandwidth approach, they would need to have receivers in the marketplace--receivers costing over $4000, at that. This myth also presumes that U.S. satellite or cable people would be willing to tie up valuable channel capacity for a very small fringe market, and a market that uses incompatible equipment that wouldn’t be able to receive over-the-air HDTV.

The Transmission-Related “Chicken And Egg” Myth

A transmission-related myth is that the HDTV market will be held back because of the “chicken and egg” problem of broadcasters not having programming to support the new HDTV receivers.

In reality, there won’t be much locally-originated HDTV programming at first, nor will there be the need for it. There will be a lot of HDTV programming distributed nationally, almost immediately. Just think of all the movies already in the can, ripe for HD distribution. Networks will distribute HDTV programming via satellite to local affiliates. National distribution will happen quickly via satellite to local cable operations. All it will take is one or two stations or cable outlets in a major market equipped to pass through the HDTV signal, and HDTV will be off and running.

The Myth Of Expensive Conversion

This myth is about how expensive it’s going to be for broadcasters to convert to true high definition television. (The myth was probably started by those pushing “enhanced” television approaches as a way to woo broadcaster support, before the FCC sent them back to the drawing board.)

In fact, high definition does not have to be expensive for broadcasters. As was just pointed out, local broadcasters won’t have to convert their studios all at once to be able to offer HDTV. They will need a new transmitter for the HDTV simulcast signal, but it can be a cost-effective low-power transmitter. And, in many cases, they should be able to use the existing tower. For HDTV programming, initially at least, local broadcasters will simply re-broadcast network HDTV feeds. Low cost converters will be available to downconvert HDTV programming to NTSC for broadcast to conventional televisions. Likewise, locally-originated NTSC programs, like the six o’clock news, will be upconverted and transmitted (albeit in low-resolution) on HDTV channels.

Some transmission systems were actually designed from the outset to be cost-effective for broadcasters. Zenith’s, for example, will make it possible to broadcast HDTV signals with only one percent of the power needed to transmit NTSC signals. That means broadcasters with a $250,000 electric bill to transmit NTSC will spend only about $500 to transmit full HDTV on Zenith's system.

The Myth That Production Standards Are A Non-Issue

Linked to the myth about transmission systems is the Japanese myth that production standards aren’t really an issue, because converting between formats will be easy. (Remember, the Japanese already have designed cameras and other studio equipment based on inferior “interlaced” technology.)

The fact is, it’s a one-way street. Interlaced standards cannot be converted to true HDTV progressive scan without unpleasant motion artifacts in the picture. But progressive scanning can be converted easily into interlace. Conversions will be a fact of life forever, just as 24-cycle film into 50- and 60-cycle television is a way of life today. Film is very widespread. Eighty-five percent of non-live network prime time programming is produced on film. Just think of all those HDTV series reruns. Each market will produce video in its own standard and film will continue to be a common denominator for program exchange.

The Myth That Early HDTV Sets Will Become Obsolete

Finally, in the transmission system myth category, is the myth that an HDTV set made to receive over-the-air broadcasts of analog or hybrid signals will be obsolete when “full digital picture distribution over fiber” becomes a reality.

In reality, television manufacturers are very close to the market. When the market wants to pay for flexibility, manufacturers build special modes. (For example, televisions are now being made with built-in teletext and with jacks for direct computer and audio input.)

All that will be needed 20 years from now when full digital and fiber are ready to roll will be that the digital image has the same number of scan lines and the same frame per second rate as the HDTV transmission standard. By then, most HDTV sets already in homes will have jacks for video input to use the screen as a monitor. The consumer will only need to buy the digital-electronics black box. The digital signal will have to be converted back to analog anyway to drive the display.

Display Technology Myths

Cathode ray tube technology is “approaching basic limitations in … resolution … larger screen sizes mean exponentially increasing costs … and fabrication and alignment of CRTs becomes increasingly difficult as pixel densities and pixel counts continue to climb,” according to the Booz Allen study for the American Electronics Association.

This is wrong. CRTs are not yet approaching “basic limitations” in resolution, and costs should not increase with resolution as fast for CRTs as for flat panels. Zenith's flat tension mask (FTM) technology, for example, holds the promise of allowing for large sizes of HD tubes (up to over 30-inch diagonal screen size) to be produced at much lower cost and higher performance than is currently possible.

Zenith currently builds and markets 14-inch FTM monitors and picture tubes. In late 1991 and 1992, we plan to introduce 16-inch and 20-inch workstation displays. This spring, we showed publicly for the first time a prototype FTM display in the 20-inch family. Our plan includes even larger screen FTM tubes to dovetail with the start-up of HDTV, and should ultimately lead to low-cost, large-screen FTM tubes for the current television standard.

Flat tension mask CRTS built using new production processes will not “exponentially” increase costs with area. And fabrication of CRTs, in the case of the flat tension mask, does not become increasingly difficult as size grows. In fact, clean-room requirements for HDTV tubes decrease as with larger sizes. In many respects, it is easier to make a 30-inch FTM with 800,000 pixels for HDTV than a 14-inch FTM with 800,000 pixels such as we can make today.

The Myth That Flat Panel Competitiveness Is Crucial To The U.S.

The same study states that “Competitiveness in flat panel will be crucial if the U.S. is to retain a strong manufacturing base in large-screen receivers."

There is no support to the conclusion that flat panels will ever achieve CRT performance or cost. Look closely at the FTM; CRT performance is a moving target. The study also does not mention that flat panel displays may have exponential cost problems as they scale up in size and density unless seamless methods can be found to assemble flat panel displays from pieces. Finally, the connection between flat panel displays and competitiveness in large-screen receivers is nonsense. It flies in the face of the $700 million investment that foreign companies are currently making in large-screen CRT plants in the U.S.

The Myth That New Displays Will Change Television Viewing

Another display myth says that flat panel displays will revolutionize the way we watch television with a big “hang on the wall” television screen.

The fact is, while flat panel displays may ultimately revolutionize the form factor of televisions and will be important for portable applications, the larger television screens in a house are not constrained by depth. Flat panel displays may ultimately have a place in direct-view HDTV. But the consumer buying decision will continue to be influenced largely by the trade-off between cost and performance.

The First Popular Myth: The U.S. Won't Catch Up

No discussion of the American Agenda for HDTV would be complete without touching on many of the popular myths on the subject. The first popular myth says that the Japanese and Europeans are too far ahead and too heavily subsidized for U.S. industry to catch up.

This simply isn’t true, in terms of basic system design and key television components, although it may be true for key elements of studio equipment like cameras and recorders. Work in Japan and Europe has been focused on direct broadcast satellite delivery and does not address the critical issues of using the existing spectrum and compressing data into 6 MHz. Furthermore, the Japanese system design was frozen many years ago and includes “interlaced” scanning and display. So does the European system. This interlace choice for satellite delivery is a big problem for low-cost, compressed over-the-air HDTV.

The Japanese and Europeans are not protecting the terrestrial broadcasters, many of which they own. They are setting up competitive distribution. The Europeans at least are recognizing their installed base of television sets by using satellite transmission systems that will be down-convertible, using a set-top decoder, to play in low-resolution PAL and SECAM formats.

The U.S. is not lagging, despite the approximately $1 billion in R&D funding subsidies to foreign system developers, because the U.S. system must be capable of being broadcast over the air in the space of one channel. Knowing where the U.S. market was heading has allowed Americans--at least Zenith and our partner AT&T, and maybe others--to develop leading technology in two short years.

Second Myth: U.S. Government Funding Is Required

A second popular myth states that with government funding assistance, U.S. industry can be world-competitive in high-definition systems technologies and American companies will jump at the chance to be in the high-definition television business.

In truth, U.S. industry already has the world-class HD technologies necessary for success, plus the ability and desire to commercialize them on its own. The principal barrier that keeps American companies from making massive new commitments of resources to the consumer electronics business is the lack of opportunity for profitability that the market has demonstrated for the past two decades.

Television unit sales volume in the United States has set new records in each of the last ten years and for almost every one of the last twenty. Yet U.S. companies have, one by one, closed their doors or sold out at losses because our market became the world’s dumping ground for electronics products.

Loopholes in the dumping laws and a lack of aggressive enforcement have allowed our market to become an incremental volume opportunity for Far Eastern producers with closed home markets. They have used our market to build up the scale of their plants and push down their costs. These Far Eastern producers have been willing to operate their U.S. business at fully accounted losses for years (as they have regularly reported to the International Trade Commission). Their unrelenting pressure on pricing and profitability drove most of the U.S. companies not only out of R&D, but also out of business.

Sure, government co-funding of American technology in American labs would help accelerate the development and commercialization process. But the main thing the government should be doing is to level the playing field.

We must have fair and tough trade laws--and they must be vigorously enforced--to assure a strong American presence in the coming era of HDTV. Congress still needs to strengthen antidumping laws, making them as tough as those of the European Community. Much of what needs to be done requires no new legislation; it just needs changes in administration of the law.

The Myth Of Large Screens

A third popular myth is that high-definition television will be large-screen only.

Quite the contrary. Until you’ve seen HDTV on a 27-inch, 20-inch or even a 14-inch picture tube, you can’t appreciate how good it is in any screen size. We’ve shown all of these screen sizes, as well as ten-foot diagonal HDTV images, in our labs and at the National Association of Broadcasters’ show. Choosing the right HDTV screen size is simply related to how far the viewer sits from the screen. At four to five picture heights of viewing distance, HDTV is a knockout. From five feet away, it is truly stunning on a 20-inch screen for the kitchen or den. I strongly believe that there will be a substantial market in smaller-screen HDTVs.

The Myth Of Wide Aspect Ratio

Another popular myth is that high-definition, by definition, means wide aspect ratio.

On the contrary, it is clear that aspect ratio is completely unrelated to the subject of HDTV; 800,000 pixels vs. 150,000 pixels really is the issue. Of course, broadcasters selling commercials and television set manufacturers selling hardware always prefer to have something that looks new. That’s why many of them over-promote the widescreen concept.

Keep in mind that, for many years, all receivers that are manufactured to receive high-definition broadcasts will be dual-mode receivers--they will receive current broadcasts as well as the new high-definition broadcasts. And for many years, the new high definition receivers will be primarily used to receive current NTSC programming because most programming will continue to be NTSC. Broadcasters will be slow to convert. Consider how long it is taking for full conversion to stereo television… and it is inexpensive for broadcasters. If the screen is wide, the sides of the picture will be black for current broadcasts.

Remember, pleasing viewing distance will be related to picture height. So adding horizontal dimension to a picture tube does not add high definition viewing utility. But it sure does add to cost. Screen width is a determining factor in manufacturing capacity.

A large-screen picture tube plant that can build a million 27-inch 4:3 picture tubes annually (and there are seven such new plants announced or under construction in the U.S., at $100 million each) would have the capacity to build only about 300,000 widescreen tubes of the same height as conventional 27-inch tubes. It just doesn’t make sense to slow the growth of the HDTV market because of high-priced wide tubes. Even if 16:9 is chosen as the broadcast standard, some companies will market 4:3 sets with optional ways to view HDTV (letterboxed with black top and bottom borders, or zoomed up to full screen by chopping off the sides of the wide image). This will give a larger 4:3 conventional television picture to the viewer than a 16:9 set of the same cost. These are sets that, along with projectors, will have only the $500 premium initially.

The Myth Of HDTV Expense

A further popular myth is that HDTV will initially be terribly expensive, with a premium over existing televisions of thousands of dollars.

This is true for Japan, and false for the U.S. Cost is systems-dependent. The Japanese Direct Broadcast Satellite system, and their U.S. simulcast system, require multiple frames of RAM in each receiver and digital signal processing to manipulate it. Some Japanese HDTV sets have 30 megabytes of RAM. Japanese manufacturers have been quoted as saying that their home market sets may sell for as little as $4000 by the end of the 1990s. The Sony deputy president used $6000 as the future price in his speech at the Society for Information Display symposium last month.

Zenith's approach for the U.S. market, on the other hand, will make it possible to use only one to two megabytes of RAM in an HDTV receiver. That is why we think we’ll be able to make HDTV a true mass-market product at mass-market prices, with HDTV receivers requiring only a $500 premium over conventional televisions initially, and at even lower premiums later.

The Myth Of Slow Market Growth

The final popular myth I'll present here is that the HDTV market will grow slowly and parallel the conversion from black and white to color television, reaching a one percent saturation rate in ten years. (A number of “expert” economist studies have been this conservative.)

Think about this. Put yourself in the place of the person planning to spend $2000 or more for a projection television system, high definition or not. In our model of a volume forecast, within a year of introduction, 100 percent of all projection televisions sold in the U.S. will be high definition. At 300,000 units sold per year, the industry will arrive at one million HDTV projection sets (or about one percent of the overall 100-million-home television market) in just three years.

By factoring in large-screen direct-view sets, add another million units in the third year after introduction, and direct-view volume will reach three million units (or ten million cumulative sets sold) in the fifth year. We look for ten percent saturation only five years after sets enter the market. Volume will be directly related to selling prices. The learning and cumulative production curves will be critical to promoting strong growth of the HDTV receiver market.

With a cost-effective transmission system, such as Zenith's hybrid digital-analog approach, for example, we believe that the first HDTV receivers will hit the market in late 1993 or early 1994 at no more than a $500 premium over regular NTSC sets with the same screen area. As the curves progress for integrated circuits and display devices, we expect the price premium to be only about $200 within that first five years.

High definition display devices will be about 40 percent of the cost of an HDTV set. Manufacturing scale will help promote fast growth. By 1995, thanks to the expected strength of the high definition workstation display market and new CRT technology we are developing, we should have the necessary scale to make HDTV displays cost effective.

Jerry K. Pearlman is chairman, president and chief executive officer of Zenith Electronics Corporation. He joined the company in 1971 as controller and was named vice president of finance in 1974. Pearlman became senior vice president in 1978, and later directed the company's computer products business, which was sold in 1989. Pearlman serves on numerous boards, including those of the First Chicago Corporation, Evanston Hospital Corporation and Stone Container Corporation; the Chicago Museum of Science and Industry, Northwestern University, the Kellogg Graduate School of Management, the Evanston Research Park, and the Committee for Economic Development. From 1962 to 1970, Pearlman worked for Ford Motor Company in a variety of managerial posts. Just prior to joining Zenith, he was a director and vice president of finance of the Behring Corporation in Florida. He graduated cum laude in 1960 from the Woodrow Wilson School of Public Affairs at Princeton University, and earned an MBA in 1962 from the Harvard Business School, where he was named a Baker Scholar.

International HDTV Service Via Intelsat
By Edward A. Faine and William R. Schnicke

Abstract

The development of a commercially viable international HDTV service is underway. Several field trials have been conducted and experience has been gained in providing commercial HDTV applications over the Intelsat global satellite system.

This paper describes HDTV activities on Intelsat since 1987, when initial tests using the NHK MUSE analog system were conducted. Several demonstrations throughout the Pacific Basin are described, including the use of HDTV for the broadcast of the Seoul Olympic Games to Japan. In 1989, tests were made with the digital DITs system, developed by KDD and Canon taking advantage of a Comsat Laboratories-developed Coded Octal Phase Shift Keying (COPSK) modem providing digital transmission at 140 Mbps over a 72 MHz transponder. Finally, tests in 1990 using the Scientific-Atlanta HDB-MAC system, the first HDTV system which is backwards compatible with NTSC broadcast signals, are described.

The future prospects for commercial HDTV via Intelsat are summarized, noting the advanced capabilities provided by the coming generations of Intelsat spacecraft, including Intelsat VI, the Intelsat VII and the Intelsat K designs.

Introduction

While the worldwide debate on HDTV production, emission and transmission standards rages on, American, Canadian and Japanese equipment manufacturers and transmission facility providers have been using the Intelsat system for international transmission of HDTV programming on an experimental and commercial basis for over two years. Extensive HDTV field trials have been conducted on the Intelsat system, resulting in a wealth of data for dimensioning satellite configurations for future transmissions of HDTV programming. This paper provides a summary overview of the international field trials and commercial transmissions conducted to date using the three known commercially-available emission systems, namely: the NHK MUSE system, the KDD/Canon DITs system and the Scientific-Atlanta HDB-MAC system. Each of the above emission systems uses the NHK-developed 1125/60 production standard as its source. The paper also addresses the future potential of HDTV over the Intelsat system in the context of new generation Intelsat Satellites that will become available in the next few years, all of which will have higher power and will therefore be able to cater to the use of smaller-size earth stations than are used today for the transmission and reception of international HDTV programming.

MUSE

NHK developed the MUSE system, based on the 1125/60 production standard, to broadcast HDTV signals via satellites using frequency modulation with a carrier bandwidth of 27 MHz.1 Transmission experiments using the MUSE signal via Japanese domestic satellites began in 1986. In 1987, loopback transmission tests of the MUSE signal over an Intelsat V satellite were conducted in Japan using the Ibaraki Intelsat Standard A (32 meter) earth station.

Based on the results of these experiments, international transmissions from Japan to Australia and from Korea to Japan were successfully carried out in 19882, and from the U.S. to Japan in 19893. The first was a three-hop transmission using the Japanese domestic CS-2b communication satellite, the Intelsat V satellite, and Australia's domestic AUSSAT K2 satellite. The second was a two-hop transmission using an Intelsat V satellite and the Japanese BS-2b broadcast satellite. The third was a three-hop transmission using the ANIK C-2 satellite, the Intelsat V satellite, and the Japanese BS-2b broadcast satellite. The results of these transmissions, including the system configuration and received signal quality, are detailed in the following sections.

Transmission From Japan To Australia

In July 1988, the MUSE signal was transmitted from Nara, the ancient capital of Japan, to the Japanese Pavilion at World Expo 88 in Brisbane, Australia. The transmission configuration is shown in Figure 1.

An HDTV program produced in Nara was encoded into a MUSE signal and frequency-modulated. It was then fed through a coaxial cable to a transportable earth station with a 3.0 meter antenna, and then relayed by the Japanese CS-2b satellite at C-band to the Ibaraki Intelsat Standard A earth station, where it was relayed over an Intelsat V satellite to the Sydney Intelsat Standard A earth station. In Australia, the HDTV signal was relayed from Sydney to Brisbane via the AUSSAT K2 satellite in Ku-band using transportable earth stations with 4.6 meter antennas for both transmission and reception. In Brisbane, two terrestrial radio links in the 42 GHz band were used to provide the program to the Japanese Pavilion. Although the bandwidth of the transponders utilized on the three satellites and terrestrial links differed, the MUSE carrier was bandwith-restricted to 27 MHz and interconnected by an IF signal from one transmission facility to another.

A Carrier-to-Noise Ratio (CNR) of 17.5 dB, corresponding to a just perceptible level of impairment, was the preestablished target CNR for the MUSE transmission. The measured total CNR of all links was 17.6 dB and offered good picture quality from the standpoint of noise impairment.

Transmission From Korea To Japan

HDTV programs of the Seoul Olympic Games, held in September 1988, were transmitted from Korea to Japan and broadcast there. The configuration is shown in Figure 2.

In Seoul, HDTV program materials were gathered at the International Broadcasting Center from the Main Stadium through an optical fiber link. These signals were encoded into MUSE format, frequency modulated, then provided by coaxial cable to a transportable earth station with an 8.0 meter antenna. This signal was then transmitted to Japan using a Ku-band transponder onboard an Intelsat V satellite.

Two sites were available for reception in Japan: the Yamaguchi Intelsat Standard A earth station and the Yoyogi earth station, located in the NHK Broadcasting Center in Tokyo. These two diversity receiving sites were switched, depending on local weather conditions, to maintain high CNR.

The received signal at either station was retransmitted to the Japanese BS-2b broadcast satellite and received by 75cm antennas at 81 receiving points throughout Japan. For retransmission from Yamaguchi, a transportable earth station with 2.5 meter antenna was used. For retransmission from Yoyogi, the fixed main earth station in the NHK Broadcasting Center was used. In addition, the received signal at Yoyogi also was fed to the NTT earth station in Tokyo and distributed to NTT's seven earth stations via the Japanese CS-3 communication satellite. It was then provided to additional demonstration sites through terrestrial links in the 11 GHz band. As before, at each repeater point, the 27 MHz MUSE signal was interconnected by an IF signal.

The measured CNR in Tokyo for the two-hop transmission was 18.3 dB, slightly greater than in the transmission from Japan to Australia. Thus picture quality was satisfactory from the viewpoint of noise impairment.

Transmission From The U.S. To Japan

On June 3, 1989, a global broadcast was held; it was called Our Common Future and was meant to raise worldwide environmental awareness. The broadcast originated at Lincoln Center in New York City and featured numerous stars of rock, folk and classical music, as well as world leaders such as Prime Minister Brian Mulroney of Canada, U.S. President George Bush and Soviet President Mikhail Gorbachev. Our Common Future was seen via Intelsat in more than 100 countries, reaching one billion people, but only one country--Japan--(with the exception of two sites in the U.S.) saw the event in HDTV. The HDTV program produced at Lincoln Center was encoded into a MUSE signal, frequency-modulated, and fed through a coaxial cable to two transportable earth stations, one of which accessed a U.S. domestic satellite for delivery to the two HDTV viewing sites in the U.S., while the other accessed the Canadian ANIK-C2 satellite, which relayed the MUSE signal to the Lake Cowichan Intelsat Standard A earth station on the West Coast of Canada. That station then relayed the MUSE signal via an Intelsat V satellite to Tokyo, where, similar to the Seoul Olympic Games, it was uplinked to the Japanese BS-2b broadcast satellite for distribution to 75cm antennas at 89 viewing sites throughout Japan.

The above transmissions show that international and domestic transmission of HDTV using the MUSE signal can be done by existing satellites and earth stations. An HDTV signal can be transmitted from any one place to any other place in the world via a two- or three-hop satellite transmission using Intelsat satellites in conjunction with domestic communication or broadcast satellites.

DITs

KDD (Kokusai Denshin Deuwa Co., Ltd.) and Canon developed the DITs system, also based on the 1125/60 production standard, to broadcast HDTV signals via satellite using digital modulation and a carrier bandwidth of some 60 MHz, thereby requiring a 72 MHz satellite transponder.4 The DITs HDTV Codec takes a full-baseband HDTV signal and compresses the signal by a factor of roughly five down to 140 Mbps or 120 Mbps. The output rate is selectable by switch. The higher rate, 140 Mbps, is a standard rate in the international digital hierarchy, and 120 Mbps is the standard rate in the Intelsat TDMA system.

In the spring/summer of 1989, experimental transmissions using DITs were conducted over an Intelsat V satellite between the U.S. and Japan.5 The transmission tests were conducted in two phases. The objective of the first phase was to investigate the transmission performance in two modes at 140 Mbps:
1. One-way transmission from the U.S. to Japan in the cross-strap mode of operation, i.e., C-band uplink in the U.S. and Ku-band uplink in Japan; and
2. Loop-back transmission in the U.S. at C-band.
The objective of the second phase was to investigate two additional modes at both 140 and 120 Mbps, namely:
1. Two-way simultaneous transmission between U.S. and Japan, i.e., C-band uplink and downlink in the U.S. and Ku-band uplink and downlink in Japan; and
2. Loop-back transmission in Japan at Ku-band.

The test configuration for the first phase, shown in Figure 3, consisted of an HDTV video camera and videotape recorder which provided a full-baseband HDTV signal to the DITs Codec where it was compressed to 140 Mbps and phase modulated by the Comsat 140 Mbps COPSK modem and uplinked to the Intelsat V satellite by the Intelsat Standard A (30 meter) AT&T earth station at Triunfo Pass, California. The DITs signal was downlinked by the Intelsat V satellite at Ku-band and received by a fixed and a transportable earth station in Tokyo, with 5.5 and 2.6 meter antennas, respectively. The received DITs signal was received, demodulated, decoded and the full-baseband HDTV signal recovered. For completeness, C-band to C-band loopback tests using the Intelsat Standard A earth station also were conducted.

Besides the DITs Codec, the other key digital technology used in the transmission was the 140 Mbps COPSK modem developed by Comsat Laboratories. The digital modulation technique used today in many telecommunication satellite systems is QPSK or Quadrature Phase Shift Keying. The COPSK modems used in the field trial are rare and special and are in fact prototypes for units now being considered for satellite restoration of undersea fiber optic cables.

The test configuration for the second phase of the field trial is shown in Figure 4. The DITs signal flow was similar to that just described in the first phase. However, as shown in Figure 4, both 140 Mbps and 120 Mbps modems were used and the fixed earth station used in the second phase had an antenna diameter of 3.3 meters, rather than 5.5 meters.

Prior to the transmission testing, the threshold Bit Error Rate (BER) of the DITs Codec had been determined in the laboratory. It was found that the threshold BER at which the influence of transmission errors can be ignored is 1 x 10-4. The objective of the field trial was to determine, for each configuration tested, the CNR required to produce a BER of 1 x 10-4. With this information in hand, one can then dimension a transmission system-- determine earth station sizes and the like--for commercial delivery of HDTV via Intelsat satellites.

The CNR required to obtain a BER of 1 x 10-4 in each configuration tested, for transmission rates of 140 Mbps and 120 Mbps, is summarized in Table 1. First, the IF loopback test indicated that a CNR of 12.5 dB would be needed to obtain a BER of 1 x 10-4. This test included everything but the satellite link; the camera was connected to a Codec/modem pair back-to-back, so the full baseband HDTV signal was encoded, modulated, demodulated, decoded and the full baseband HDTV signal recovered. In the other configurations with the satellite link inserted, the C-C U.S. loopback, Ku-Ku Japan loopback and C-Ku U.S.-Japan Transpacific tests resulted in only a slight CNR degradation of about 1 dB. In the Ku-C Japan-U.S. Transpacific tests the CNR degradation was more than 2 dB, but that was due to inadequate uplink provisioning, which would be rectified in a commercial setting. It is therefore concluded from the results summarized in Table 1 that the CNR required to obtain the threshold BER of 1 x 10-4 are 13.5 dB for 140 Mbps COPSK DITs transmission and 13.0 dB for 120 Mbps QPSK DITs transmission.

With the above information, appropriate transmission configurations for DITs HDTV exchange between the U.S. and Japan were determined assuming the use of an Intelsat V, a 13.0 dB CNR for a threshold BER of 1 x 10-4, 2 dB of downlink margin and Intelsat Standard A (30 meter), E3 (8 meter), E2 (5.5 meter) and E1 (3.5 meter) earth stations. The results are shown in Table 2, which shows the resultant CNR and availability for various combinations of earth stations.

In a given row, the upper CNR and availability numbers are for the Japan-U.S. direction and the lower numbers for the U.S.-Japan direction. To interpret Table 2, take the case of an E1, or 3.5 meter, earth station transmitting to another E1 station. The availability would be 93.2 percent in the Japan-U.S. direction, or 99.28 percent in the U.S.-Japan direction. By 99.28 percent availability, it is meant that the BER would be better or equal to the threshold BER of 1 x 10-4, 99.28 percent of the time over the course of a year.

The following is concluded from Table 2. Setting aside combinations of E1 stations, DITs HDTV transmissions over Intelsat would have a 99.9 percent or greater availability for all other earth station groupings. Further, for E1 stations, it would be desirable that the interfacing station be an E3, or larger, station. For E2 stations, it is desirable that the interfacing station be an E2, or larger, station.

Based on the above, it is concluded that DITs HDTV transmission via Intelsat is commercially viable. The field trial results clearly demonstrate it would be possible to exchange DITs HDTV via Intelsat using small-size transportable earth stations in the E1 to E3 class (3.5 to 8 meter diameter range) and achieve a 99 percent or greater availability.

In fact, in October 1989, the transportable 2.6 meter diameter antenna earth station used in the field trial was used commercially to transmit DITs HDTV programming via Intelsat from Japan to the SMPTE conference in Los Angeles. The receiving earth station located outside the conference hall was an E2.

HDB-MAC

Scientific-Atlanta developed the HDB-MAC system, based on the 525-progressive and/or 1125/60 standard, to broadcast HDTV signals via satellites using frequency modulation with carrier bandwith of 36 MHz.6 HDB-MAC is the first emission system to be Intelsat-tested that is backward (NTSC)-compatible and security encrypted. In March 1990, the HDB-MAC system was used in the first international two-way HDTV video teleconference. 7 The HDTV videoconference linked the MAST Industries home office in Andover, Massachusetts with its regional office in Hong Kong. Mast Industries, the design and procurement arm of The Limited, Inc., a worldwide clothing retailer, used the videoconference to evaluate and select designs and materials for future fashions. High definition pictures of designs and materials developed by 1125/60 standard HDTV equipment installed in MAST offices were encoded into a HDB-MAC signal, frequency modulated and routed over fiber optic cables within Hong Kong to the Hong Kong Intelsat Standard A (30 meter) earth station and relayed by an Intelsat V satellite in C-band to an Intelsat Stand B (11 meter) earth station in Brewster, Washington. The HDB-MAC signal then was relayed via the GE Astro K1 satellite in Ku-band from Brewster to Andover using transportable earth stations with 5 meter antennas for both transmission and reception. A similar path carried the HDB-MAC signal back to Hong Kong, although different Intelsat V satellite with different accessing earth stations was used on the return path.

Good picture quality was obtained, prompting Martin Trust, president of MAST Industries, to remark: "In our business, color, texture, style, quality and timeliness are essential factors in making buying decisions. Now we have direct communications between Hong Kong and the home office with the clarity and intimacy of face-to-face meeting. With these (HDTV) technologies we can make decisions on the spot, expediting our buying process."

The Future Potential Of HDTV Via Intelsat

To date, all HDTV transmissions via Intelsat have been relayed by current generation Intelsat V satellites. In the next several years, Intelsat will be replacing its current fleet of Intelsat V satellites with new generation Intelsat VI and Intelsat VII satellites, as well as augmenting the system with a specialized all-Ku-band satellite known as Intelsat-K. These satellites have higher e.i.r.p. than the current Intelsat V satellites. At C-band, the Intelsat VI and Intelsat VII have 2 db and 4 db higher e.i.r.p. than the Intelsat V satellite, while at Ku-band they have 3.5 dB and 5.5 dB higher e.i.r.p. in the East Spot beam, with about the same e.i.r.p. in the West Spot beam as the Intelsat V satellite. The increase in e.i.r.p. can be used to cater to the use of smaller-size antennas and/or to improve picture quality. The most dramatic reduction in antenna size for future HDTV transmissions is expected to occur at C-band. For example, the reference antenna size for the Intelsat V satellite was a Standard A (30 meters), while for the Intelsat VI satellite it is a revised Standard A (18 meters) and for the Intelsat VII satellite it will be a Standard F3 (9 meters). It should be possible, for example, to receive high quality HDTV via an Intelsat VII satellite at C-band with 9 meter antennas; rather than 30 meter antennas used today. The first Intelsat VI satellite was launched successfully in 1989 for transatlantic service; the first Intelsat VII launch is planned for early 1992.

The most exciting prospect on the Intelsat horizon is the anticipated arrival of the Intelsat-K satellite in the Atlantic basin in early 1992. This Ku-band satellite with broadcast satellite parameters, 50 dBW e.i.r.p. and 54 MHz transponders, offers exciting possibilities for international HDTV delivery. Assuming full use of an Intelsat-K 54 MHz transponder, high quality HDTV could be delivered directly into antenna sizes as small as 1.8 meters. If the past two years of HDTV via Intelsat can be characterized as bright, the future looks even brighter.

Conclusions

HDTV via Intelsat is a commercial reality. Three different HDTV emission systems have been employed in commercial end-to-end settings delivering a wide range of HDTV programming in a variety of earth station configurations involving two- and three-hop satellite links. Moreover, as the Intelsat system is replenished with new generation Intelsat VI, VII and K satellites in the next several years, improvements are anticipated in the delivery of international HDTV through the use of smaller-size antennas coupled with the attainment of better picture quality.

Notes

1. Y. Ninomiya, et. al., "HDTV Broadcasting and Transmission System: MUSE," Proceedings of HDTV Colloquium 1987, Ottawa, Canada, (October, 1987).

2. Keiichi Kubota, et. al., "International Transmission of HDTV Signals," SMPTE Journal, (February, 1990).

3. "Our Common Future," HDTV Newsletter, (July, 1989).

4. S. Matsumoto, et.al., "120/140 Mbit/s Intrafield DPCM System for Digital Transmission of HDTV Programs," Second International Workshop on Signal Processing of HDTV, L'Aquila, Italy, (March, 1988).

5. S. Matsumoto, et. al., "First International HDTV Digital Transmission via Intelsat Satellite," Proceedings of PTC 1990, Honolulu, Hawaii, (January, 1990).

6. K. Lucas, "HDB-MAC, A Conditional-Access HDTV Transmission Format," Proceedings of PTC 1990, Honolulu, Hawaii, (January, 1990).

7. "First HDTV Videoconference conducted by SONY and S-A," Satellite News, (February, 1990).

Acknowledgements

The authors would like to acknowledge those who contributed to this paper, namely NHK, Kinji Ono of KDD, Barry Hobbs of Scientific-Atlanta, Peter Malcolm of Comsat World Systems Division (Intelsat Satellite Services) and Russel Fang of Comsat Laboratories; in addition to those individuals who contributed to pioneering HDTV via Intelsat.

Edward A. Faine is director of international systems planning of the Intelsat Satellite Services Division of Comsat's World Systems Division. He is responsible for current and future space segment planning relative to U.S. participation in the Intelsat system and serves as a U.S. delegate to the Intelsat advisory committee on planning. Prior to joining Comsat in 1973, Faine was a systems analyst at Computer Services Corporation, and before that he worked as a microwave engineer at Sylvania Amherst Laboratories. He holds a BS from Ohio State University and an MS from the State University of New York at Buffalo, both in electrical engineering.

William R. Schnicke has worked at Comsat since 1974. He has held the position of senior director, systems engineering and planning, since 1989. Schnicke's achievements at Comsat include the overall satellite concept and design for Intelsat VII, Intelsat's seventh generation spacecraft; as well as system concepts and satellite designs for Intelsat VI and Intelsat V. He conceived and led an international task force to develop and implement international television satellite to provide high power Ku-band service between North America and Europe in the 1990s. He also managed the first live international demonstration of high definition television from California to Tokyo over Intelsat satellites. Prior to joining Comsat, Schnicke founded and worked as a senior analyst at Ketron, Inc.; and held positions at Kappa Systems and Magnavox, all in Pennsylvania. Schnicke has written articles for many publications, and has been an active member of IEEE since 1967. He holds a BS from MIT and an MS from the University of Pennsylvania, both in electrical engineering, and an MBA in finance/accounting from the University of Maryland.

Direct Broadcast Satellite
By Robert W. Hubbard

Direct broadcast satellite (DBS) will be the greatest transmission mode for HDTV and could be the only means for a very fast mass market penetration.

The idea of DBS (the transmission of television via satellite directly to a consumer's home) has been around since the introduction of satellites. Today, there are somewhere between two and three million C-band satellite receivers installed for this purpose. Satellite transmission in the commercial band--both C and Ku--transmits with low power levels and the satellites are spaced two degrees apart. This causes interference between adjacent satellites and requires large receive antennas. A cable operator or a broadcaster can easily afford to invest in the large antennas needed to eliminate this adjacent satellite interference. It is very difficult for a consumer to use this large equipment. The K-Prime medium power venture is planning on using an antenna more than three feet in diameter, and the size will have to grow, maybe to five or six feet, when neighboring satellites are launched two degrees apart. These large antennas, because of their much greater weight and the tremendous force exerted by the wind, need a concrete foundation or significant structural strengthening of a roof when placing them on a home. Both the large size, and the high cost of the equipment and installation, prevent these systems from being effective choices for mass-market television service.

Compare this to true high power DBS, which is authorized for many times more power, and just as importantly, nine degrees spacing in the orbital arc. Current technology will allow us to reach every television household in American with a very small, inexpensive antenna. The majority of the population will be able to use an antenna less than 18 inches in diameter. An 18-inch antenna can be installed inside the home, looking out a window, strapped to a chimney or porch railing, or fastened directly to a roof. Also, as technology improves, the antennas needed for true high power DBS reception will drop to less than ten inches in diameter. The cost of these antennas and receivers initially will be around $1000. This will drop very quickly. In Japan, where DBS has been available for three years, receive sets cost less than $500. In the U.S., the cost should decline to the $300 level in jsut a few years.

True high power DBS will offer a new, previously unattainable capability to reach every household in the United States with television, audio, and data simultaneously, and from one source. What this means is that any programming concept that has interest to as few as one half of one percent of the United States population will be economically viable. Let’s just assume that two percent of the U.S. population has an interest in stamp collecting and is willing to invest $700 for the equipment to receive a one-hour weekly program on stamp collecting. Local television stations and local cable systems cannot afford to provide service to such a narrow audience. Today there is no way to reach these two million households efficiently. Because television stations and cable systems cannot afford to offer this service locally, the broadcast networks and cable networks cannot get nationwide carriage. This assures that there will be no television service for those people interested in stamp collecting. However, this two percent, when combined from all across the country, is a very large two million television household audience. This is larger than almost any individual market. It is easily economically viable. Again, just pulling numbers out of thin air, the one percent interested in woodworking, the two percent interested in stamp collecting, the one half percent interested in Japanese language programming, the four percent interested in water skiing, and so on, can now be collected together to start the penetration of DBS receiving equipment. Once a household has this equipment for their specialized interest they also become potential viewers for any other services offered.

This economy of scale also works in the consumer's favor when pricing a service. Today, in cable, over half the revenue stays with the local cable operator. After expenses, roughly 35 percent is left over for the program producer and the cable network to split. With DBS, more than 80 percent is left over to split between the program producer and the DBS operator. This means the identical service can be sold for half the price and the program producer will still make more money.

The same economies of scale are applicable for high definition television. The introduction of HDTV in the United States will have many obstacles to overcome. Television stations, video rental stores, and cable operators will often be unwilling and many times unable to invest the capital needed to provide a wide selection of HDTV programming. At the same time, consumers will be unwilling to invest significant dollars into equipment that will have very limited use. This slow growth in sales will keep prices high, further slowing growth. What this means is that the penetration will be very slow using terrestrial television outlets. Though there will be some markets where a television station, cable system, or video rental store will invest the dollars needed to make it worthwhile for a consumer to buy HDTV equipment, on a national scale, penetration will be slow and sporadic. This was the situation when color television began in the United States. In the early 1960s if you lived in Minneapolis/St. Paul, or some other market early into color, there was an incentive to buy a color television set. But if you lived in a market that had no color television programming there was no reason to buy a color set. This limited introduction kept prices high, even in markets where color was available. This is why the penetration of color television went very slowly.

Some have said that all markets will very quickly have an HDTV outlet. We believe this is very unlikely. Again, let us assume that there are two million television households that would spend the thousands of dollars necessary to buy an HDTV set. This translates into an average of 2500 viewers per television station, 250 subscribers per cable system, and only 50 customers per video store. If the cost to equip a television station or cable system runs into the tens of thousands, hundreds of thousands, or quite probably the millions of dollars, then the investment is too great for the small number of viewers that will be gained. Contrast this to a true high power DBS system where the cost to provide a nationwide HDTV service will be only slightly greater than the cost to provide a HDTV service to just one local market. Two million households is a huge business for a DBS operator. Because of the very large audience, a variety of high quality high definition programs will be available. This will be a nationwide incentive for consumers to purchase HDTV sets. The problems that the introduction of color television sets faced will be eliminated.

We do not mean to imply that terrestrial television stations, cable operators, and video stores will not play in the HDTV game. After HDTV sets reach a critical mass in a local market, then these providers may be able to offer an economically successful service. We believe very strongly that DBS will be the driving force behind the initial sales of HDTV sets. Without DBS, it will take a long time for HDTV to become a mass market service.

Technically, it will be much easier for DBS to send HDTV than either cable or terrestrial television. Television, through conventional outlets, relies on satellites to distribute the programs. This assures that any new system will always be satellite friendly. Satellite's wider bandwidth, pure ghost-free transmission path, and no maintenance make DBS the best transmission system ever designed. In the unlikely event that digital HDTV is unavailable in the near future, there are two existing and already commercially available and operating: Scientific Atlanta’s HDB-Mac and NHK’s MUSE-E. These two systems probably never will be economically available for terrestrial television stations or cable systems. However, they can very easily be used for a DBS service. In Japan, MUSE-E is in use every day and very shortly, service will be greatly expanded. In effect, DBS's worst case scenario will be using a system that has had all the experimenting and debugging already completed in a consumer application prior to our introduction of HDTV services.

We are confident that digital HDTV will be available for use when we start in 1993 or 1994. Digital technology offers tremendous advantages to DBS. These are advantages that will be much more difficult for television stations and cable systems to take advantage of. DBS is an entirely new service with no existing infrastructure. It will start with tomorrow’s technology. Terrestrial television and cable, on the other hand, are mature industries with very extensive and expensive existing infrastructures. In order for a television station to utilize these advantages it must replace a great deal of its transmission and service facilities and persuade consumers to purchase new equipment. It is unlikely whether regulators will allow the television industry to force consumers to upgrade or be left out. Cable systems will need, at the very minimum, to replace or add to the boxes in people's homes. This will be an unbelievable expense. Many systems will also have to rebuild portions of their plant, at costs higher than the cost for an entire DBS system, in order to take advantage of these digital capabilities. Obviously the cost advantage for a new technology always lies with the new service.

Fiber optics is just a new form of cable. If fiber is used only to replace existing cable installations, it brings no real advantage to the consumer. If, on the other hand, switched fiber is installed, the advantages are still only marginal. DBS with digital compression will be able to offer hundreds of channels to every home in America. This vast availability of channels will allow a video-on-demand capability at a fraction of the cost of fiber optics. The cost to install switched fiber to every home is too great for this country to pay.

DBS is the last opportunity for a gateway directly into people's homes and will be an unbelievable success, both financially and from its ability to offer new services to the American public. United States Satellite Broadcasting Company (USSB), as a broadcast licensee, intends to offer many types of programming: free advertiser-supported programming, subscription programming, pay-per-view programming, minority programming, HTV programming, and a great deal of public service. USSB will offer traditional television services and new services that are impossible to deliver today. These will be limited only by our imaginations. DBS will be the road that will allow HDTV a quick and easy journey into a large number of Americans' homes.

Robert W. Hubbard is vice president of Hubbard Broadcasting, and a third-generation member of this family business. He has experience in all areas of television, including news, sales and engineering. Hubbard works with the company's United States Satellite Broadcasting Company, and is well-versed in the promise and problems related to the development of HDTV for over-the-air broadcasting, both from satellites and terrestrially. He is now working with Hubbard Broadcasting and USSB's plans for implementation of DBS and HDTV. In this capacity, he coordinated the Minneapolis/St. Paul portion of Our Common Future. Hubbard is a graduate of the University of Minnesota.

It's About Quality: HDTV And The Movie Theater Business
By John W. Johnston

The excitement and anticipation surrounding HDTV has generated intense speculation about potential uses for the new technology. One set of proposals calls for the creation of HDTV movie theaters. The argument is that traditional film theaters will be replaced by some form of electronic theater, using HDTV video projection of programs recorded on HDTV tape or transmitted by satellite.

Those who champion HDTV theaters seem to have forgotten a basic question. Yet the answer to that question is the lens through which we should look at any discussion of HDTV. The question is: What is HDTV all about? Why are so many companies, in so many countries, dedicating so many people and investing so much money trying to develop it? The answer can be summed up in one word: quality! What drives the global pursuit of HDTV is the promise of a dramatic improvement in the quality of the television image.

This may seem obvious, as it should. Nevertheless, we often overlook the equally obvious implication: our goal is to improve quality, not to develop technology. If we become so involved with a particular technology that we advocate it even if it is the lesser quality alternative, then we're working against our own goals, against the reason we got involved with it in the first place.

Thus, when we focus on quality, we can see how illogical it is to "force fit" HDTV into situations where it does not improve quality. But this is exactly what the proponents of HDTV movie theaters seem to be doing.

There is no question that, no matter which format of HDTV is eventually adopted, it will offer the home consumer tremendously improved television picture and sound. But when we move from the home to the movie theater, we are moving to a different level of quality, a different set of needs.

Today, the vast majority of theatrical motion pictures are originated on 35mm film and the vast majority of movie theaters project 35mm or 70mm film prints. The superior picture and sound quality of this film system is recognized around the world. In fact, HDTV developers have always used 35mm film as the quality standard they want to equal. For "HDTV theaters" to make any sense, they would have to offer a quality improvement.

But none of the of the proposed HDTV formats can do that. None of them has been able to imitate the "film look." Like art, the film look can be hard to define. But we all know it--and like it--when we see it. We do know that several technical factors contribute to the film look. Resolution, of course, is one. Dynamic range--the gamut of colors and tones that can be captured and displayed--image structure, sensititivity and exposure latitude are among the others. In none of these areas does any HDTV format improve on, or even equal, film.

Naturally, these comparisons can be made only with today's films. But film is not a static technology; its quality and performance in all these technical areas is constantly being improved. So if HDTV developers are using 35mm film as their quality target, they are aiming at a constantly moving target. Moreover, 35mm film is not really the ultimate target. 70mm releases are becoming increasingly popular, especially for highly touted "big" films, like last year's Batman, this year's Dick Tracy, and the restored Lawrence of Arabia.

Except for Lawrence of Arabia, these have been 70mm prints from 35mm originals. But there is also a growing interest in originating on 65mm negative for 70mm release. Three of the major professional motion picture camera manufacturers--Arriflex, Cinema Products, and Panavision--have introduced brand-new, state-of-the-art 65mm motion picture cameras. And several 65mm productions have been announced. Other producers have been experimenting with advanced film systems, such as Showscan. This system uses 70mm prints projected (and shot) at 60 frames per second, to create incredibly realistic images; the effect must be seen to be appreciated.

During the summer of 1990, cinema cigital sound debuted in selected theaters across the country. This revolutionary step forward in theatrical sound quality offers movie theater audiences the superior quality of digital sound, the quality they have come to expect and enjoy on their home CD systems. In short, the motion picture industry is now concentrating efforts on further improving the quality of the theatrical experience, an experience that is already superior to that offered by any proposed HDTV theater.

The reason is simple: theatrical distribution remains the cornerstone of the industry. It is true that producers no longer expect to make a profit, or even to recoup their costs on many pictures, from domestic box office returns alone. Today, the so-called ancillary markets, including pay, cable and broadcast television, and, most importantly, videocassette sales and rentals, are what really make or break a movie.

But in terms of marketing strategy, the theatrical box office is the launch point for every major American motion picture. The ancillary markets depend on the theatrical release to give a film its initial exposure, to spread word-of-mouth advertising, and to build the momentum they will profit from. For example, distributors in export markets aren't interested in films that have failed at the U.S. box office. Just about every videocassette outlet says that its biggest hits were last year's biggest box-office hits. And producers who have tried making films directly for cassette admit they have not been very successful.

So the focus today is on luring audiences into the theater, by making the theatrical experience distinctly superior in audio/visual quality to anything available elsewhere. This interest in improving quality coincides with the changing demographics of movie audiences. Recent studies suggest that so-called aging baby boomers are becoming the largest movie-going public. As in other areas, when they choose a movie--and a theater--they look for quality and value.

Not only producers, but exhibitors as well, are responding to this call for quality. Theater owners and managers are working to upgrade their environments and several theater consulting programs have been set up to help them. Among them are George Lucas's TAP (Theater Alignment Program), the Peterson Consultants, Inc. seminars and Eastman Kodak's MBO (Management by Observation) program.

Exhibitors are also trying new ideas designed to attract older, upscale audiences. A number of theater chains, for instance, are turning one auditorium in their multiplex theaters into a deluxe screening room that offers amenities such as a smaller number of larger, plusher seats, reserved seating at slightly higher prices, and private lounges and concession areas.

The trend is international. In Belgium, the country's leading theater operators, the Bert family, recently built the magnificent Kinepolis in Brussels. This huge, ultra-luxurious multiplex has 22 screens for 35mm projection, two for 70mm for Showscan and one for IMAX (another, large format, advanced film system). During its first six months of operation, theater attendance in Belgium shot up by 38 percent. Now, the Berts are adding cinema digital sound to the theater, to make it even more attractive to audiences.

Quality improvements have been paying off here, too. Even though VCRs spread rapidly across the country during the 1980s, U.S. box office receipts rose every year but one during the past decade, and every year during the past four years, reaching an all-time high of $5.03 billion in 1989.

There is another way to look at the question of HDTV theaters. We could simply ignore the issue of quality and see if a business case can be made for them based on costs alone. Of course this is contradictory; as we have seen, the theater business is increasingly based on quality. The motion picture industry, like business in general, has learned that quality pays. But, for the sake of argument, let's try it.

In fact, the argument for HDTV theaters has always focused on costs, specifically the cost of making and distributing film prints. But no one really knows what the cost of duplicating and distributing high definition tapes would be, nor what the cost of satellite transmission time will be.

We do know that the cost to the distributor of a typical 35mm print for an average first-run engagement comes to between $5 and $6 per show. And we know that the present system of film print manufacture is highly efficient and cost-effective. Today's high speed film printers run at well over ten times real time and some labs have proprietary equipment that works much faster. So, even during peak seasons, labs can fill high-volume print orders in remarkably short time.

In contrast, high definition videotapes will probably be duplicated in real time, at least initially. Filling a 2000 video copy order would take a lot of time and a lot of equipment. And labs would have to pay for that new equipment. It would be a long time before there would be any possible cost savings to pass on to customers.

Investing in satellites and satellite time may be equally expensive, especially if a two-hour feature is being transmitted to a theater four or five times a day. It may be possible to transmit a program only once, and have a theater record it on its own equipment, then project its own video copy. But this raises the problem of security and piracy; what happens to those video copies the theaters create? Further, the cost of owning equipment for both satellite distribution and tape display makes this a very inefficent system.

Even if, somehow, one of these electronic distribution systems cost the distributor nothing, there's still no guarantee that the savings--which amount to only $6 per show--would be passed on to exhibitors. And even if they were, how long would it take them to recover their investment in HDTV equipment?

Current estimates for equipping a theater for high definition tape or disk display start at around $250,000 per screen. Estimates for satellite distribution make it even higher, about $300,000 per screen. But the average annual gross revenue per screen in the U.S. is about $220,000. If ten to fifteen percent of this is used to offset the cost of the video equipment, it could take more than a decade before the investment cost is recovered. Another way to look at this is to say the cost for equipping every screen in the U.S. for HDTV distribution would be around $6 billion.

That doesn't include a penny for operating and maintaining the equipment. Nor does it consider the cost--human as well as financial--of replacing projectionists with potentially higher paid video engineers. But satellite dishes and video equipment must be maintained by electronics technicians, while many of today's multiplex theaters have automated equipment that can be run by a single manager-operator.

There is also a "non-cost" issue surrounding satellite distribution that must be considered: it greatly limits the exhibitors' flexibility in scheduling and maintaining quality. For example, the start of a popular film is often delayed because of the time needed to seat large crowds. Will satellite transmission require a fixed start time, even though a percentage of the audience is not seated? And will inclement weather create signal losses that further decrease the quality of the presentation?

All these questions add up to one simple one: what is the advantage, in cost or quality, of turning a movie theater into a giant HDTV television set?

Nevertheless, there are some areas where HDTV theaters do make sense: those that take advantage of television's unique power of immediacy, its ability to make us experience live events while they're happening. Thus, presenting concerts, sports, live theatrical performances and other special events in special venue HDTV theaters does give audiences a quality advantage over traditional closed-circuit television presentations. Similarly, audiences for certain non-entertainment events, such as continuing education programs, medical conferences, business meetings, and so on, will also benefit from their presentation in an HDTV theater.

In other words, the true value of electronic theaters will not be as a replacement for current film theaters, but as an addition to them, giving people more reasons for going out to a theater. And that will be the key issue facing the theater business in the future.

As HDTV becomes popular in the home, it will educate audiences to expect even higher quality images in a theater. Certainly, they will not want to pay to go to a theater, if it offers the same image and sound quality they can get at home. So exhibitors will have to continue offering audiences an experience that is distinctly better. When that experience involves an entertainment motion picture, it will be produced--and displayed--on film.

John W. Johnston has been with Eastman Kodak's Motion Picture and Television Products (MP&TV) Division for 17 years in various sales and marketing positions. He began his career with Kodak as a sales and engineering representative in California and Texas. In Hollywood, Johnston had account responsibility for several major motion picture studios. More recently, he served as marketing programs director for MP&TV's worldwide business planning group. Currently he is district sales manager in Atlanta. Johnston holds a BS from California State University, and is a member of SMPTE and AICP.

Field Emission Displays For HDTV
By Chris Curtin

Introduction

The field of electronic displays emerged about 60 years ago when the CRT was commercialized as a practical device for television images and for oscilloscope and radar presentations. After World War II, with the widespread introduction of entertainment television, first in monochrome and then in color, followed by the "information explosion" caused by the introduction of digital computers in the 1960s, electronic displays (CRTs) grew into a substantial industry.

From the beginning, CRTs had unpopular attributes, such as excessive bulk and weight, and the need for implosion and X-ray protection. More recently the analog nature of the CRT is increasingly a nuisance in the digital electronics world. However, its outstanding advantages, including ever-increasing resolution in full color; the efficiency of its cathodoluminescent screen; its simplicity, durability and low cost continue to resist effective substitution.

From its inception as the display workhorse, the CRT has undergone continuous attempts to reduce its bulk. Some progress was made by increasing deflection angles up to 100 degrees or more, and from the earliest days, true sandwich constructions also were attempted. These activities continue energetically today. Flat CRT designs generally fall into two classes: those incorporating one or more electron beams, which are deflected or guided towards the phosphor screen by means of electrostatic field-forming structures; and those embodying linear or area-type cathodes, the emission from which reaches the screen through grid structures addressed in an X-Y manner.

The most successful flat CRT is the vacuum fluorescent display (VFD). This is a modest device, incorporating discrete symbology structures, and is used as an indicator in the appliance and automotive industries.

All attempts to develop a design capable of matching traditional CRT performance have failed to date, due to structural complexity and/or cathode limitations. Nevertheless, the goal of developing a flat display technology which can inherit the pedigree of the CRT, especially the resource of cathodoluminescence, has become even more desirable today, in view of the poorly competitive alternatives to the CRT which the display community has developed so far.

Field Emission Displays (FEDs)

The evolution of high-density matrix-addressed field-emitter arrays has been the result of work by Capp Spindt and associates at SRI International. In contrast to conventional field emitters which require an ultra-high vacuum or heating to obtain stable emission, the micro-miniature Spindt cathodes can be comfortably operated at pressures typical of commercial CRTs. This key factor, supported by life tests, plus a series of improvements in the fabrication processes for microelectronics, has allowed the transition of the technology from the research laboratory into a full scale engineering development program directed at commercial applications.

SRI International has demonstrated cathode current densities of 1000 amps per square cm; and a very recent announcement from Philips described a device at 8000 amps per square cm. These results are significant advances in the state of the art, far beyond the requirement for entertainment displays. Using efficient cathodoluminescent phosphors, 100 footlambert FEDs can be built with 100 microamp per square cm cathode arrays, which are six orders of magnitude below these recent developments.

An FED consists of two pieces of glass, separated by small (invisible) spacers, and containing a high vacuum (10-6 torr). The faceplate carries the phosphor screen (as in the traditional CRT) and the baseplate supports the cold-cathode field emitters as well as the row and column driver chips. Bonding the driver chips to the glass substrate will greatly reduce the number of connections to the interface board, thereby improving the cost and reliability. A typical assembly is shown in Figure 1.

In Figure 2, a cross-section view of an FED, the phosphor dots are aligned opposite the intersections of the row and column electrodes. Since the emitter holes are only one micron in diameter, ten to 100 tips are deposited at each intersection to provide redundancy for emission uniformity, manufacturing yield, and life. The XY matrix is addressed via low voltage waveforms applied to row electrodes (upon which the emitter tips have been deposited) and the column electrodes which constitute the accelerating anode or gate. The non-linear characteristic of the resultant diodes eliminates the "half-select" problems seen in some flat display technologies.

The insulating pillars are incorporated to support the atmospheric pressure and to allow the display area to grow without increasing the thickness of the faceplate or the baseplate. The Figure 2 cross section is not to scale, but typical dimensions might be: phosphor dot and cathode 0.1 x 0.1 mm; spacer height of 0.1 mm; faceplate and baseplate thickness of 1.1 mm.

The field emission display combines the best features of both the traditional CRT and the more recent flat panel alternatives:

)

In projecting the applicability of new display technologies to HDTV applications, a primary consideration is power consumption, since the display area will be quite large and today's capability of household circuits is fairly modest. Figure 3 shows the power consumption of an HDTV display having the same incremental resolution as today's 19-inch NTSC receiver:

The important element in this analysis is the net efficiency of the final product, not the inherent efficiency of the light-generating mechanism. For example, a traditional CRT uses a green phosphor with an efficiency of 60 lumens/watt, but wastes 60 to 70 percent of the beam energy in the shadow mask. A further reduction occurs due to the transmission of the black matrix (~60 percent transmission), and the use of less-efficient red and blue phosphors. The end result is a net efficiency of less than two lumens/watt for a white picture! Since the FED does not utilize a shadow mask, it is anticipated that the net efficiency will be three lumens/watt, even at the lower operating voltage.

It is projected that alternative display technologies will have much lower efficiencies than the FED, and will have marginal performance in this situation. For example, liquid crystal projection systems are striving for one lumen/watt and small backlit LCDs hope to achieve 1.5 lumens/watt. Color plasma displays use efficiency phosphors, but require an inefficient discharge to generate ultraviolet light to excite the phosphor. This combination seems to limit the expectation of color plasma displays to one lumen/watt. Finally, electroluminescent (EL) displays will have a net efficiency around 0.5 lumens/watt due to the low efficiency of the thin film blue phosphors and the large capacitive power load of this "solid state" technology.

HDTV Display

An HDTV display fabricated with field emission cathodes could have the following set of performance characteristics:

Display area 3 x 5.3 feet (6.1 foot diagonal)
Resolution 1024 x 1800 white pixels
Brightness 100 FL (white)
Viewing angle 160 degrees peak
Power consumption 530 watts peak; 106 watts average
Thickness 1 inch
Weight 50 pounds

Cost

In addition to the critical question of power and efficiency (brightness), the manufacturing cost is critical to the sucess of an HDTV consumer display. The FED lends itself to high volume manufacturing techniques by virtue of its requirement for standard thin-film processing steps which are being automated by the semiconductor industry. A careful analysis of the FED compared to an active matrix LCD (today's leading flat color technology) indicates that the LCD will be 30 percent less than the LCD due to fewer manufacturing steps and higher yields.

While the feasibility of such a large display must be proven, some of the elements are already in place. Vacuum systems to deposit thin films on windows for offices and homes are common and are used in production environments. The lithography equipment to pattern the emitter holes will need to be developed, as will the glass handling jigs and fixtures. The drive circuits for an HDTV display will prove to be a formidable challenge, and some proprietary techniques that promise to be cost effective are being explored. Another alternative, which is available to any technology, is to build a mosaic display from smaller tiles. However, it is not one that the author would prefer.

Schedule

The field emission display will initially be developed for PC applications where light weight, low power, full-color displays are needed. The next set of products will be aimed at workstations, first at 1024 x 1280 resolution, and next at 1280 x 1600 where the costs should be less than a comparably-sized CRT. Finally, the delivery of 40-inch, 60-inch, and 80-inch diagonal HDTV sets will take place in the late 1990s, soon after the beginning of commercial broadcasting. The price of these products will be comparable to the high-end (35-inch) NTSC sets currently available, easily affordable by the consumer who demands the best in picture quality.

Summary

In summary, the field emission display promises to improve upon the performance of the traditional CRT, while overcoming its disadvantages in size and weight which limit the display area below the optimum for HDTV.

Chris Curtin is a founder and vice president of product engineering at Coloray Display Corporation. Prior to starting Coloray, Curtin held a variety of engineering and management positions at Tektronix, including general manager, Imaging Displays Business Unit; general manager, Display Devices Division; CRT Manufacturing manager; and CRT Engineering manager. He holds two U.S. patents, has published numerous papers on CRT and display technology, and is an active member of the Society for Information Display. Curtin holds a BS in physics from Lewis and Clark College in Portland, Oregon, and has participated in the Tektronix Executive Development Program.