First-Hand:Digital Television: The Digital Terrestrial Television Broadcasting (DTTB) Standard
Contributed by Stanley Baron, IEEE Life Fellow
The International Telecommunications Union (ITU), a special agency of the United Nations, is assigned the responsibility of developing international agreements on wired and wireless communications. The ITU considers the changing state of global communication networks and services and develops recommendations for the promotion and harmonization of developments in the field as they apply to the international community.
Responsibility for broadcast services was originally pursued within the ITU's International Radio Consultative Committee (CCIR). The CCIR was created in 1927 for the purpose of serving the needs of the broadcast community and was incorporated into the ITU in 1947. The ITU underwent reorganization in 1993, and the CCIR was reborn in March of that year as the ITU's Radiocommunication Sector (ITU-R).
In 1991, member nations petitioned the ITU to investigate requirements for enhanced, advanced television broadcasting services. The ITU-R formed a special Task Group (Task Group 11/3) in January 1992 to develop Recommendations (the ITU refers to its standards as "Recommendations") for advanced television broadcasting services.
At its final meeting held in Sydney, Australia in November 1996, Task Group 11/3 finalized international agreements defining a complete digital terrestrial television broadcasting (DTTB) system that could be applied to broadcasting in both 50Hz and 60Hz environments, regardless of whether channels of 6 , 7 or 8 MHz bandwidth were employed.
Narrowing the focus
In 1991, the worldwide broadcast community was divided as to which technologies might best serve their needs. In 1987, the United States Federal Communications Commission (FCC) had called for proposals for an advanced television system. Twenty-six different systems were offered for consideration. Most of these systems were based on analog technology, but four of the proposals were based on digital technology. After study and testing the different proposals, the FCC concluded that the future needs of the United States would be best met using digital technology. Further, the FCC concluded that each of the four proposed digital systems offered some advantages and disadvantages, and that the best approach would be to pick and choose the best technologies used in each of the four digital systems. The system developed based on this combination of technologies was termed the Grand Alliance System.
One of the four digital systems was developed by the Advanced Television Research Consortium (ATRC), a team consisting of NBC, Philips, Sarnoff Laboratories, and Thomson-RCA. The ATRC system was based on the ISO/IEC MPEG-2 digital compression system and data service multiplex and transport. It used a multiple carrier modulation technique for transmission. The other three systems employed various proprietary technologies. The ATRC provided the first over-the-air digital HDTV simulcast in the Americas (and possibly in the world) on 30 September 1992, proving that the MPEG system could support the required services. HDTV programming was broadcast over a UHF channel from the WRC-TV antenna in Washington, D.C. and received 68 miles away, while an NTSC version of the program was simultaneously broadcast on WRC-TV, NBC channel 4 in Washington, D.C. The MPEG-2 digital compression system and data service multiplex and transport used in the ATRC system was adopted for use in the Grand Alliance system. I participated as a member of the ATRC team that developed the system and also participated in the system tests during September 1992.
The FCC designated the AC-3 audio coding system proposed by Dolby and the 8-VSB single carrier modulation system offered by Zenith as part of the Grand Alliance System. At the request of the FCC, the Grand Alliance System was documented by the Advanced Television Systems Committee (ATSC). I chaired the ATSC technical committee charged with documenting the Grand Alliance System. The ATSC began with the Grand Alliance System and added features and capabilities that would enhance the ability of broadcasters to meet the viewing audience’s expectations and to facilitate interfacing with cable and satellite systems.
Europe was looking at a broad range of compression methods and the European community was actively supporting a system that relied heavily on advanced analog technology termed MAC (multiplexed analog components). However, the Nordic countries (Sweden, Norway, Denmark and Finland) were working on a system that also employed the MPEG-2 compression system and data service multiplex. The Nordic HD DEVINE system employed a COFDM multi-carrier modulation scheme.
The Japanese proposed a proprietary MAC system developed in Japan that was designed primarily for satellite broadcasting.
Organizing the effort
In response to the petition to investigate a system for enhanced, advanced television broadcasting services, the ITU issued an urgent question concerning digital terrestrial television broadcasting. A meeting was held in Geneva during the first week of November 1991 to develop agreement on the terms of reference for a Task Group to respond to the urgent question and to identify a chairman and one or more vice chairmen to lead the Task Group. At the meeting, agreement was reached on the terms of reference for the Task Group and naming me, Stanley Baron (USA), as chairman, and Terry Long (United Kingdom) and Osamu Yamada (Japan) as the two vice chairmen.
The Australian delegation volunteered to author a ballot to the member nations. The ballot was approved by a sufficient number of administrations to launch the Task Group, designated Task Group 11/3, in January 1992.
The assignment involved: (1) defining a fully digital system for enhanced broadcasting services that would include a higher resolution picture, surround sound, and employ digital broadcasting technology that used spectrum more efficiently than that employed by conventional television services; (2) defining the technologies that were to be employed, and (3) providing agreement on a set of system parameter values. This effort to develop a consensus for enhanced broadcasting services would deal with a set of technical problems of great complexity, since several competing technologies were being offered for consideration.
ITU procedures dictate a minimum period between the approval of an ITU committee and its first meeting to allow the member nations to prepare documents and appoint delegations. The rules require that decisions on how the Task Group should organize its work be made by a vote of the delegates at the first meeting. The rules also provide for election of ‘Special Rapporteurs’ who might serve as the chairmen of any subgroups established by the Task Group to be made by a vote of the delegates at the first meeting.
In January 1992, after consulting the ITU rules, I spoke with Dr. Giuliano Rossi, the ITU senior counsel assigned to the Task Group to establish a date for the first meeting. The first available date was in December 1992. I explained to Dr. Rossi that I did not want to waste the six to twelve month period between approval of the Task Group and its first scheduled meeting. Rossi counseled that ITU rules would allow me to start work and to appoint individuals as 'temporary' Special Rapporteurs. These individuals would continue in their roles if approved at the first meeting of the Group. In the interim period, work could begin.
Outline of Work
Prior to the first meeting, I generated a series of documents for distribution to the member-nation administrations. The documents were designed to provide guidance in meeting the Task Group's assignment to draft Recommendations leading to a single worldwide digital television standard. The guidance included an Outline of Work. The outline of work provided a list of issues to be considered by the Task Group in preparing its Recommendations. The document also presented a set of draft Recommendations for consideration, an outline of a report (or reports) to be generated either as annexes to Recommendations proposed by the Task Group, as Reports to carry the work of the Task Group forward, or in the form of a possible tutorial Report. The list of issues was based on a system model. The system model divided the DTTB system into four areas the Task Group should consider exploring: (1) source coding and compression, (2) the service multiplex and transport, (3) the physical layer and modulation system, and (4) planning factors and implementation.
I submitted the draft Outline of Work to five individuals who I hoped would serve as Special Rapporteurs [the two vice chairmen, Mr. Long and Dr. Yamada, and Thomas Ryden (Sweden), Richard Barton (Australia), and David Wood (EBU)] for comment and suggestions for improvement. In accordance with my discussion with Dr. Rossi, I had appointed Barton, Ryden, and Wood to serve as temporary Special Rapporteurs.
The Outline of Work was intended to be used as a guide for Administrations and other interested parties preparing documents for consideration by the Task Group. After incorporating the suggestions made by the five Special Rapporteurs, I forwarded the document to the ITU Secretariat for distribution to interested parties. The text of the Outline of Work asked all administrations to review the document and forward any suggestions for improvement to the ITU with a copy directly me by 31 August 1992.
I submitted an initial work plan to the ITU early in 1992 calling for presentation of a set of Recommendations and Reports by December 1996. The task with the longest planning cycle was the selection of the modulation scheme as the science in that area was not fully proven.
ITU rules provided that the two vice chairmen had the right to chair a subgroup that dealt with a technical area in which they had an interest. Yamada chose to chair the effort on the service multiplex and transport (TG11-3/2). Long chose to chair the subcommittee responsible for the effort on planning factors and implementation (subgroup TG11-3/4). I then assigned source coding and compression to Ryden (TG11-3/1) with Ryden focusing on audio source coding and Brian Roberts (New Zealand) leading the effort on video source coding. The physical layer and modulation system documentation (TG11-3/3) was assigned to Barton. Wood was assigned responsibility for reviewing the Outline of Work and developing guidance for a minimum standards set for a DTTB system.
The Task Group met for the first time at ITU headquarters in Geneva, Switzerland, 14-18 December 1992. The meeting was attended by more than 115 delegates representing 43 national administrations and international organizations. At the first meeting of the Task Group, the individuals I had selected as temporary Special Rapporteurs and their assignments were approved by the delegates, and the effort was launched.
Prior to and during the first meeting, I suggested that the focus of the Task Group should be to document a set of Recommendations that bring the greatest benefit to the consumer, thereby, maximizing the probability of success of adoption by the world community. I understood that various national administrations and regional groups would be pursuing goals that they saw as providing the most benefit to the interests they represented. My goal was to find a set of solutions that provided the most benefits to the consumer. The cost to the consumer of access to media is dependent on the volume of manufacturing; the greater the volume, the less the cost. Defining a system that described a limited set of tools that could be used worldwide would best meet this goal.
The world community had developed a standard, MPEG-2, for the basic coding and multiplexing of video, audio, and data signals for use in digital media. The MPEG-2 standard was developed for television applications in which channel bandwidth or recording media capacity is limited and the requirement for an efficient transport mechanism is paramount. The MPEG-2 standard provides a set of tools that may be used to describe a system. Task Group 11/3 based its work in many ways on both the work and the philosophy behind the MPEG-2 standard. Employing a similar approach, the set of Recommendations defined by Task Group 11/3 established a set of tools that could be used to provide a DTTB service applicable to broadcasting in both 50Hz and 60Hz environments, regardless of whether channels of 6, 7, or 8 MHz bandwidth were employed. Another goal was to construct a system that allowed international exchange of program and data services across media boundaries. To accomplish this goal, it is necessary to agree upon a set of basic elements for DTTB. In October 1994, I worked with and David Wood to develop a Recommendation that provided the basis for agreement on a minimum set of basic elements defining a DTTB system. The agreement provided for the following:
- A base system capable of conveying a single HDTV service or a number of conventional quality services that includes:
- A common, multiple level compression syntax for both video and audio;
- Coding video sources in conformance with the MPEG-2 standard at the Main Profile, Main Level (MP/ML) or higher;
- Coding audio sources in conformance either with the MPEG-2 standard Level II or the AC-3 standards, utilizing a single set of integrated circuits (ICs) capable of decoding both MPEG-2, level II and AC-3 audio coding.
- A common program multiplex standard where the service multiplex and transport conformed to the MPEG-2 standard and used a common Service information and Header descriptor system that included a common standard for identification;
- A common recording standard for program interchange for each of the levels of compression used;
- A common electrical and mechanical interface standard at the data stream level;
- Channel coding and modulation schemes using either the 8-VSB system where single carrier systems are appropriate or COFDM technology where multiple carrier systems are appropriate.
As noted, at the time of the Task Group's first meeting in December 1992, most European administrations and Japan were interested in pursuing MAC analog technology. Fortunately, by the second set of meetings in the spring of 1993, the European community had concluded that a digital solution afforded important improvements in use of the radio frequency spectrum, and they began to focus on the application of digital technology. Much of the early work on discrete-cosine transforms that were the basis of the MPEG system compression scheme originated in Europe, and that background and the early success of the Nordic countries in developing their digital system helped move the European community and the Task Group to a consensus.
The Digital Terrestrial Television Broadcasting (DTTB) Model
The Outline of Work described a model of a DTTB. The model was divided into four areas of interest with subgroups assigned to develop the required Recommendations and Reports. The Task Group used the model as the basis of its investigations.
The four subsystems of the system model are as follows (reference Figure 1): (1) Source coding and compression, (2) Service multiplex and transport, (3) The physical layer including the channel coding parameters and the modulation scheme, and (4) Planning factors (which includes consideration of both the transmission and receiver environments) and implementation strategies.
Source coding and compression refers to coding methods designed to reduce the large data stream created when images are represented by a sequence of individual digitized picture elements (pels or pixels) or when sound is represented by digital audio samples. Source coding may also include error protection techniques that are appropriate for application on the video, audio, and ancillary digital data streams. The term "ancillary data" included system control data, conditional access control data, or data associated with the program audio and video services such as closed captioning. Ancillary data can also refer to independent program services. Compression involves processes that reduce the bit stream containing the image, sound or ancillary data information in such a way as to minimize the number of bits needed to represent the information and to be able to recreate a representation of the original source at the receiving point without noticeable or unacceptable degradation.
The service multiplex and transport module divides the digital data stream into packets of information, uniquely identifying each packet or packet type, and provides appropriate methods of multiplexing the video, audio, and ancillary data stream packets into a single program data stream. Multiplexing also provides the capability of combining different program data streams into a single broadcast channel for simultaneous delivery. In developing an appropriate transport mechanism, interoperability between digital media such as terrestrial broadcasting, cable distribution, satellite distribution, recording media, and computer interfaces must be a prime consideration.
The physical layer includes the channel coding and modulation scheme. The channel coder takes the compressed data bit stream and adds additional information that can be used by the receiver to recognize and reconstruct the images, sound, and ancillary data from the transmitted signal. This includes any additional data added to the multiplexed data stream to provide protection against partial loss of the signal. The characteristics of the channel coder are selected to support the modulation scheme adopted for the system and the medium through which the data must be transported. Modulation is a mechanism whereby the protected data stream is imposed on one or more carrier signals for transmission. These transmission systems are referred to as single-carrier and multiple-carrier schemes, respectively.
Planning factors and implementation strategies include consideration of the characteristics of the transmission media and receivers and discussions of strategies appropriate for the introduction and implementation of a DTTB service, taking into account other existing broadcasting services.
Summary of Task Group Recommendations
Task Group 11/3 completed its work in November 1996 and produced a set of Recommendations and Reports that defined a unique DTTB system. Two subsets of the standard's set of tools were described in detail: System A (ATSC) for use in 60Hz, 6MHz channel environments and System B (DVB) for use in 50Hz, 7 and 8 MHz environments. The differences between the two subsets were minimized and harmonized with respect to the video and audio coding and transport levels so that there are no conflicts and single, "plug and play" decoders are possible.
A list of the international Recommendations and Reports produced by the Task Group accompanied by a list of the Recommendations and Reports described in the Outline of Work for comparison purposes is shown in Table I.
The Recommendations and Reports produced by the Task Group defined a complete DTTB system and provided for the following:
Video coding: During the first year of the Task Group's effort, consensus was reached to adopt a modified version of the MPEG-2 standard as a mechanism for video source coding for broadcasting applications. MPEG video encoding profiles and levels for use in the DTTB standard were reduced from more than 20, first to four, and finally, to two. These two profile/levels are the Main Profile at Main Level (MP@ML) which defines the consumer appliance for standard digital television services only, and the Main Profile at High Level (MP@HL) which provides for a consumer appliance capable of decoding both HDTV and standard digital television services.
Agreement was achieved on the dropping of two profile/levels: the Main Profile at 1440 (MP@14) and the Spatially Scalable Profile at 1440 Level (SC@14). The MP@14 profile/level had been considered for use in services that were not quite HDTV (EDTV). It has been shown that the MP@HL profile/level described in the DTTB standard can accommodate those services efficiently. The Spatially Scalable (SP@14) profile had been considered for use for accommodating spatially and temporally scaled services. After a thorough investigation by the EBU, consideration of this profile was dropped as spatially and temporally scaled services were shown to be extremely "bit hungry" and inefficient for use for terrestrial broadcasting.
The number of profiles and levels included has an impact on the cost of the consumer appliance (receiver) and the cost of providing programming content for the international market. The set of DTTB video subsystem tools are defined in ITU-R Rec.BT.1208 and allow content producers to provide programming in conventional, wide screen, and HDTV formats. The existence of single decoders capable of decoding the entire set of tools defined in ITU-R Rec. BT.1208 fully met the request of the World Broadcasting Union for unique global broadcasting systems leading to single universal consumer appliances.
Audio coding: Selection of an audio source coding system was more controversial. There were two primary candidates, the European supported MUSICAM system and the United States supported Dolby AC-3 system. MUSICAM was part of the MPEG-2 specification and backwards compatible with prior systems. AC-3 was found to be a more efficient system that had already been adopted for use in motion pictures and on DVD. (In surround sound tests conducted by the EBU to evaluate the performance of the two systems, the AC-3 system operating at 600 kbps, were found to provide equivalent performance to a MUSICAM system operating at 900 kbps.) Both systems had already been deployed in other media and had found wide acceptance.
After investigating how the MUSICAM and AC-3 systems worked, I suggested to the Task Group 11/3 delegates that decoders for these two systems could be viewed as a series of resources (arithmetic units and memory) under the control of an instruction set. Since each system could provide three levels of audio performance (monaural, stereo, or surround sound), each decoder could be viewed as a series of resources with three resident instruction sets. The appropriate instruction set was selected for the desired level of service. Since the resources required to implement the two systems were very similar, a dual decoder would not require the building of two separate decoders, but could be implemented employing a single set of resources with six resident instruction sets, three for MUSICAM and three for AC-3. After discussing this architecture with manufacturers of audio equipment, it was concluded that the cost to the consumer purchasing a digital television receiver with a dual decoder would be less than a 0.25% increase over the cost of a receiver with a single system decoder capability. After further discussion, the Task Force accepted this proposal, and the ITU Recommendation provided for both MUSICAM and AC-3. The set of DTTB audio subsystem tools are defined in ITU-R Rec. BS.1196 and allow content producers to choose between both MPEG1 backward compatible MUSICAM and the AC-3 compression and coding tools. The existence of single decoders capable of decoding the entire set of tools defined in ITU-R Rec. BS.1196 fully met the request of the World Broadcasting Union for unique global broadcasting systems leading to single universal consumer appliances.
Agreement was reached that any new television service should provide for a range of multi-channel services from a single channel (monaural) to 5.1 channels for each service. One or more services, each comprised of between 1 and 5.1 channels, can be accommodated depending upon the data capacity of the multiplexed bit stream. A typical speaker placement is shown in Figure 2.
Digital television systems provide compression mechanisms that produce bit streams as low as 32 kbits/s for voice/dialogue services to 384 kbits/s for 5.1 channel sound reproduction. The coding provides for transmission of one (monaural) or two independent (stereo) sound channels, as well as matrixed services which may include L,C,R (Left, Center, Right) and surround sound. The surround sound options include a single surround channel (S) or a surround sound pair Surround Left (SL) and Surround Right (SR). In addition, a low frequency enhancement (LFE) channel capability may be added to any of the matrixed services. The LFE channel is defined as having a limited frequency range (20 Hz to 120 Hz) and allows the listener to extend the low frequency content of the sound format in terms of both frequency and level. It essentially duplicates the sub woofer channel used in digital film sound formats. Since the LFE channel is coded at a lower bit rate, it constitutes the '.1' in the '5.1' notation.
Transport level: The service multiplex and transport provides the foundation for the DTTB system. It is a constrained subset of the MPEG-2 standard tool set defined in Rec.BT.1299. The assignment of packet identification as described for System A and System B was harmonized to avoid the possibility of decoder errors. Systems that conform to the subset of the MPEG-2 transport defined in the Recommendation including the use of the Descriptor Tags and Table ID assignments, allows the development of single devices capable of decoding the entire set of tools defined. The existence of single decoders capable of decoding the entire set of tools defined in Rec. BT.1299 met the request of the World Broadcasting Union for unique global broadcasting systems leading to single universal "plug and play" appliance for use by consumers without the need to consider the specific subset used. Since MPEG-2 had been originally developed for video storage applications such as DVD, it required some minor additions to allow for use in the broadcast environment. In order to facilitate the acceptance of these changes by the MPEG committee, I invited the chairman of the MPEG committee, Dr. Leonardo Chiarglione to participate in the Task Force meetings. Therefore, Chiarglione was familiar with the concerns of the Task Group, understood that the changes were necessary, and arranged to have the MPEG committee incorporate the necessary changes.
Physical layer: The major obstacle to agreement on a standard for the modulation system was the lack of uniformity in the use of the broadcast spectrum throughout the world. Countries that had adopted the NTSC system developed spectrum plans employing 6 MHz channel bandwidth. Countries that had adopted the PAL and SECAM systems developed spectrum plans with channel bandwidth that ranged from 6 MHz to 8 MHz. Therefore, the available bit capacity of the systems would vary geographically.
Some nations, such as Australia, Canada, and the United States, developed national broadcasting systems that focused on local broadcasting. In these environments, one or more national broadcasters produce and distribute programs nationally to local service providers, but allow for local insertion of commercials and locally generated programming such as news throughout the broadcast day.
Other nations developed national broadcasting systems that are truly national in their content. In these environments, one or more national broadcasters produce and distribute programs that are broadcast nationally without local modification. The ‘local’ broadcaster is simply a re-transmission tower. These latter national broadcasters would find the most efficient use of the spectrum coming from a system that allowed for the use throughout the nation of a single frequency network (sfn) that was impervious to reflections and deployed receivers that responded to the strongest sfn signal in the local environment. In dealing with the variety of broadcasting environments, the Task Force could not agree on a single approach to the transmission standard. Two different approaches were documented termed ‘COFDM’ and ‘8 VSB’. COFDM was less susceptible to interference and would accommodate national single frequency networks.
8 VSB was slightly more bit efficient (carried more bits per MHz) and was designed for use in 6 MHz environments where the insertion of local content dominated. 8 VSB (see Figure 3) was adopted for use in System A (ATSC).
COFDM (see Figure 4) was adopted for use in System B (DVB).
Although a set of COFDM parameters was proposed for use in System A by a group of broadcasters including ABC and NBC in the United States and the CBC in Canada, the FCC Advisory Committee deemed the proposal as coming too late in the testing schedule, and the system was never accepted for testing.
Recommendations concerning the DTTB physical layer (channel coding and modulation scheme) are defined in Rec. BT.1305 and take into consideration the existing 6, 7, and 8 MHz allocation of channel assignments and the need to accommodate differing environments and planning factors. The set of Recommendations and Reports can be viewed as providing a single, compatible system solution for DTTB within the practical physical limitations of the current worldwide channel assignment environment.
Multi-program capability and interoperability with other media: The application of digital signal compression technology to coding television signals combined with packet identification accommodates multi-program transmission in the existing channels. Compressed digital television systems offer the prospect of considerable improvement in service quality while appreciably improving spectrum utilization as compared with analog transmission methods. This capability can also be exploited to deliver multiple digitally compressed television programs instead of a single conventional, enhanced, or high definition program. These digitally compressed television signals can be accompanied by digital high quality sound, coded conditional access information and ancillary data channels.
Furthermore, the same approach could be implemented in the transmission of multi-program signals or stereoscopic television services over existing digital satellite or terrestrial links or cable television networks. Task Group 11/3 paid particular attention to constructing a digital architecture that could accommodate both high definition television (HDTV) and conventional television services in the terrestrial broadcasting environment and was interoperable with cable delivery, satellite broadcasting, and recording media. The approach taken provides harmonization between services by using a unified, common method of video and audio source coding and a unified, common service multiplex and transport. As noted above, two different subsets of this unified set are defined; System A (ATSC) and System B (DVB). The two subsets are compatible and single decoders can be provided that can extract either subset from the data stream. This unified transport data stream is then provided with a framing structure, error protection mechanism, and modulation scheme appropriate to the distribution media. The common transport is seen as a "container" and facilitates the interoperability of the signal through different delivery media. This results in a common data stream after demodulation in the receiver which simplifies the complexity of the consumer receiver appliance.
HDTV: During the final meeting of the Task Group, several nations indicated an interest in the 1080 line system as the basis of international agreement on an HDTV standard for program interchange. Such an agreement would be beneficial to content producers who have global market concerns. The then existing 1152 line/50 Hz (Europe) and 1035 line/60 Hz (Japan) HDTV standards was seen as specific to certain nations or regions. The 720 line (USA) standard was seen by some administrations as too close in performance to the existing progressive scan versions of the 625 line conventional broadcasting standards for consideration. Only the 1080 line standard existing in 50 Hz, 60 Hz, and 24 frame per second film compatible versions was seen as providing sufficient improvement over all existing, conventional formats to form the basis for international agreement.
Picture Aspect Ratio: The 4:3 picture aspect ratio (the ratio of picture width to height, expressed as H:V) of conventional television systems was adopted from the aspect ratio of motion picture films adopted by the film industry in the 1930's. The early Japanese proposal for a wide screen HDTV service suggested a picture aspect ratio of 5:3 to replace the conventional television standard picture aspect ratio of 4:3.
The development of the16:9 wide screen picture aspect ratio adopted by SMPTE and the ITU began in 1984, in response to requests by the Hollywood film community. The Hollywood community had petitioned SMPTE to begin work on a standard for electronic production for film. The cost of access to a technology of interest is dependent on volume, and it is accepted that volume is driven by standards. The film community was beginning to rely more on electronic generation for special effects and animation, and a standard was deemed important. Over the decades, the film industry had become accustomed to using many picture aspect ratios. Electronic production would most likely be restricted to one aspect ratio. The problem to be resolved was what single aspect ratio would best serve the needs of the creative artists.
Dr. Kerns Powers of Sarnoff Laboratories studied the issue. He drew rectangles of all of the popular film formats and overlaid them one upon the other. He discovered that all formats would fall within a rectangle with a 1.77:1 aspect ratio (reference Figure 5). Further, he discovered that all of the rectangles fell outside an inner rectangle, which also had a 1.77:1 shape. This meant that a 1.77:1 picture aspect ratio provided the ideal 'shoot an protect' format that would allow motion pictures produced electronically to be released in any aspect ratio. Kerns Powers provided his results to SMPTE, and the international community adopted the 16:9 aspect ratio (1.78:1) for electronic film production as it offered the most favorable compromise for a single electronic standard.
In viewing Kerns Power's work on electronic film production, the community working on DTTB realized that the 16:9 aspect ratio would also provide the most efficient format for this new service, and the Japanese 5:3 ratio was replaced by 16:9 (16:9 = 5.333:3). The 16:9 aspect ratio had been incorporated in SMPTE and ITU standards for HDTV services.
Completing the assignment
Since a Task Group is a temporary committee with a defined lifetime and organized for a specific purpose, one of my statements of guidance also provided for continuation of the work by standing ITU committees with suggestions for transfer of responsibilities to specific permanent ITU committees. Since a technical standard is never "finished", the work should be viewed as an evolutionary process. The NTSC standard, for example, was established in 1940, modified to include color in the early 1950s, modified later for stereo sound, and continuously modified as technology allowed for system improvements. The list of suggested transfers of assignments was adopted at the Task Group's final meeting.
The Task Group completed its work on schedule in November 1996, producing a set of Recommendations that went beyond providing for enhanced television services. It offered broadcasters the ability to construct a digital highway into each home that allowed for a range of digital services. In the case of North America with the narrowest bandwidth channel of 6 MHz, the digital highway would support a data rate of 19.3 Megabits per second. This could be used to transmit an HDTV service with surround sound and supporting data, or multiple standard television services including educational services, or provide a data path to download text, still pictures, and other information and services.
Looking to the Future
One of my first actions after I was elected chair of the ITU Task Group on digital television was to prepare opening remarks for the Task Group’s first meeting in December 1992. Many of the delegations were seeking to incorporate technology supported by their national interests into the work of the committee. In many instances, this technology did not necessarily provide benefits for the consumer (the viewing public), nor did it necessarily provide for future expansion of capability. I was hoping that the technology adopted by the Task Group, and therefore, by the ITU and its parent organization, the United Nations, would allow for digital communications services supporting the consumer’s future need to access the entire spectrum of digital information that certainly would become available over the next quarter-century, and to facilitate doing so via wireless communications.
I believed that within the next quarter-century, the marketplace would introduce and support inexpensive personal communication appliances (PCAs) combining the functions of telephones and computers. I further believed it was important for the technology adopted by the Task Group to permit broadcasters’ content to be accessed by these devices. I, therefore, argued for adoption of technology providing a set of tools for broadcasting television which would be compatible with and support the consumers’ need to access the world’s fund of knowledge.
My goal of developing a set of tools that allowed universal access to knowledge was one of the reasons that I argued for adoption of the compression technology and transport standards tools provided by the ISO-IEC MPEG committee. However, I recognized that the MPEG tool set had been originally developed primarily for use in recorded media. Wireless media is more significantly impacted by interference (noise) concerns. I was looking to the Task Group to build on the MPEG tool set to provide for identifying and enabling future services and to identify error detection and correction schemes to support reliable communications in the more difficult wireless environment.
I incorporated the text of my opening remarks in the book I coauthored with Professor Krivocheev, titled: “Digital Image and Audio Communications: Toward a Global Information Infrastructure,” published in 1995. A portion of the text of my remarks appears on pages 263 and 264 as the first three paragraphs of the section titled "The Information Age". The text found in the book is as follows:
“In the 21st century, no nation or region of the world will be able to afford to allow communication services provided to its citizens to be second-class as compared to services provided in the remainder of the world. In the coming age of information, advantages will accrue to ‘information-haves’ societies at the detriment of the ‘information-have-not’ societies.
“The vision of this new age sees the citizen of the 21st century communicating with the world using an inexpensive personal communication appliance (PCA). The PCA will support individual mobility by allowing this citizen of the world community to be contacted regardless of location. The same ‘personal contact number’ will be able to be ‘dialed’ anywhere in the world and contact the intended person independent of whether the person is at home, at work, or in transit. The PCA could allow the citizen of the world to retrieve data from the world’s libraries, images from the world’s museums, and sounds from the world’s environment. … The PCA could also allow citizens of the world to act as sources, sharing their experiences and dreams with others.
“The human character is tremendously diverse, having developed a broad spectrum of cultural traits and experiences. A world community that is able to communicate with and better understand its diverse constituencies is a world that is best equipped to both protect and maintain the individual cultural heritages. A world in which individuals can share their own cultures with others and develop an understanding of those other cultures is a world in which individuals can better develop respect for the worth of their own and other cultures. Armed with modern tools of communication, the world community will be better equipped to solve the global problems of the 21st Century.”
When the book was written in 1995, the Task Group had not completed its work, but most of the standards documents describing the digital system had been either approved or were in the final stages of editing. Therefore, I included the following paragraph on page 264 of the book:
“The ITU-R has met its commitment to provide the tools necessary to implement a global information infrastructure necessary to support the information age. The important results achieved by the ITU confirm its vitality as an international forum for communications activities. The ITU continues to play a dominant role in the new era of digital communications that will characterize the 21st Century. The technologists have provided a set of tools that allow the world community to establish an inexpensive, integrated, ubiquitous, and extensible communications system. The question now is what we will choose to create with those tools.”
A quarter-century later, when I see individuals using their smart-phones or smart-pads as they engage in their daily activities, or when I see individuals using those devices to capture still pictures or video and send it to their friends, family members, business associates, or share their experiences with the world community, I am satisfied that the work of the ITU Task Group met the challenge of this vision.
As a First-Hand History, this article was written in the first-person. However, it should be obvious that the accomplishments of Task Group 11/3 reflected work done in hundreds of laboratories and several consortia around the world and made available to the Task Group members in the published work of the European Digital Video Broadcasting (DVB) project, contributions to the United States Advanced Television Service program, projects organized by the Japanese Broadcasting Technology Association (BTA), the International Organization for Standardization (ISO) and International Electrotechnical Commission (IEC) Moving Pictures Experts Group (MPEG), the committees of the Advanced Television Systems Committee (ATSC), European Broadcasting Union (EBU), the International Telecommunications Union (ITU), the Society of Motion Picture and Television Engineers (SMPTE), and the many national delegations that participated in the work of the Task Group, among others.
The Task Group’s success can be directly attributed to the continuing guidance provided by the leadership and staff of the International Telecommunications Union (ITU), including Mr. Pekka Tarjanne, Secretary General; Dr. Richard Kirby, Director ITU-R; Prof. Mark I. Krivocheev, Chair of ITU-R Study Group 11 (Television); the members of the staff of the ITU-R in Geneva, Switzerland, and in particular Senior Counselors, Mr. Richard Nickelson and Dr. Giuliano Rossi; and Secretariat, Mrs. Renata Zecha.
As indicated in the text of this First-Hand History, the bulk of the effort was carried by the Special Rapporteurs: Richard Barton (Australia), Terry Long (UK), Keith Malcolm (Australia), Brian Roberts (New Zealand), Thomas Ryden (Sweden), David Wood (EBU), and Osamu Yamada (Japan).
The world community owes a debt of thanks to all of the individuals and organizations who were involved in, and contributed to, the Task Group’s successful completion of its assignment.
- ITU-R, Question 121/11, "Digital Terrestrial Television Broadcasting."
- CCIR Document 11 3/2, "Outline of Work for Task Group 11/3, Digital Terrestrial Television Broadcasting," 30 June 1992.
- Four of these five individuals served as Special Rapporteurs for the life of Task Group 11/3. In 1993, Ryden became Chair of TG10/3 and was replaced as Special Rapporteur for Source Coding by Roberts. Starting in 1993, Barton shared responsibilities for developing the Physical Layer with Keith Malcolm (Australia).
- IEC/ISO 13818, "Generic Coding of Moving Pictures and Associated Audio Information," (MPEG-2).
- Chairman, ITU R Task Group 11/3, "Report of the Second Meeting of ITU-R Task Group 11/3, Geneva, 13-19 October 1993," 5 January 1994, p.40.
- ITU-R Document 11-3/TEMP/1, "Chairman's Opening Remarks," October 1994.
- ITU-R Document 11-3/19, "Recommendation for Main Elements of a Common Digital Terrestrial Television ITU-R Standard," 14 September 1994.
- ITU-R Document 11-37, Draft new Recommendation, "The Basic Elements of a World Wide Family of Systems for Digital Terrestrial Television Broadcasting," 2 November 1994.
- Chairman, ITU-R Task Group 11/3, "Report of the Second Meeting of ITU-R Task Group 11/3, Geneva, 13-19 October 1993, 5 January 1994, p.40.
- S. Baron and K. Powers, "Common Image Format for International Television Program Exchange," 131st SMPTE Technical Conference, Los Angeles, CA, October 1989.