NIST GCR 03-859
Economic Impact of the Advanced Technology Program's HDTV Joint Venture

2. Overview of the HDTV Joint Venture Project and Technology Outcomes

The high-definition television (HDTV) joint venture (JV) project led to new technological innovations that reduced the cost of the conversion to digital television (DTV) broadcasting for most television stations and quickened the introduction of new digital studio technologies. Sarnoff Corporation, a research and development (R&D) firm with expertise in video and broadcasting equipment, assembled a unique team in the form of a JV to investigate new approaches to creating and operating digital studios. The National Institute of Standards and Technology’s (NIST’s) Advanced Technology Program (ATP) awarded the JV additional funding support, thereby making the research effort a joint ATP and JV project. This section provides an overview of the project, its goals, and its technology outcomes. It also identifies the underlying drivers leading to the JV’s formation and charts and evaluates the JV’s pathway to success.

It is important to note that the scope of DTV broadcasting has broadened beyond HDTV to include standard-definition television (SDTV), DTV whose quality exceeds that of today’s analog broadcast television but does not have the same high resolution of HDTV. Throughout this report, DTV includes both HDTV and SDTV. However, this report maintains the “HDTV project” title given to the project by ATP and the JV members.

2.1 THE HDTV JOINT VENTURE PROJECT

The national conversion to DTV broadcasting required U.S. television stations to initiate digital transmission operations no later than mid-2003. According to the Federal Communications Commission (FCC), “the digital transition will increase efficient use of the spectrum, expand consumer choice for video programming, and increase the amount of spectrum available for public safety and other wireless devices” (FCC, 2003b). Analog television requires a much wider spectrum allocation because empty channels must be left between co-located broadcast signals to prevent signals from interfering with one another. Digital broadcasting will allow the FCC to permit adjacent channel broadcasting, thereby freeing up significant portions of the spectrum for other uses. The ATP award that enabled the HDTV JV preceded the adoption of the FCC standard by approximately one year.

The conversion to DTV broadcasting has profound implications for the cost and organization of studio operations.⁽¹⁾ Under FCC rules, broadcasters are not required to install new digital studio equipment; they can meet FCC digital broadcasting requirements by converting their analog signals to digital. However, many broadcasters are investing in digital technology for their broadcast studio as part of their DTV broadcasting strategies. Under the analog broadcasting paradigm, broadcasters replaced or upgraded their existing studio equipment on an as-needed basis. But the digital conversion planned by many stations entails a complete overhaul of studio equipment and operations as the technologies used to manipulate and pass through digital signals are different from those required for analog.

Digital studio conversion is costly, not only in terms of equipment costs, but also in terms of installation, operations, and maintenance. Commercial stations faced a May 1, 2002, deadline, and public television (PTV) stations, a May 1, 2003, deadline to commence transmission operations in digital. Both deadlines passed with many stations not being on air because of construction delays and funding issues, among other types of setbacks. Regardless of individual station adherence to the deadlines, while commencing DTV transmission, broadcasters maintained their analog signal on air because the overwhelming majority of television audiences still use analog technology to view programming. Maintaining both digital and analog transmission operations places financial pressures on all stations’ operating and facilities budgets.

The situation was further complicated by the fact that as the DTV broadcasting conversion commenced, digital equipment had yet to achieve satisfactory performance specifications. In the early 1990s, efforts were concentrated on creating digital transmission standards for the delivery of a digital broadcast signal to the home and less so on the innovation that would be required to create effective technologies for managing the digital studio that would create and transmit that signal (Dickson, 1995).

These technology and business drivers converged to form a case for the development of a coordinated strategy for new technologies for DTV broadcasting. The technology driver for more efficient studio technology and the business driver for cost-effective solutions led Sarnoff to create a JV to investigate new technologies for managing and accomplishing digital studio operations.

Integral to the proposed JV was the assembly of a cross-functional group of digital studio equipment suppliers that would leverage synergies from coordinated R&D. In the words of Frank Marlowe of Sarnoff, “The aim of [the JV] is to drastically reduce the cost of … equipment so that the [broadcasting] industry can move forward as rapidly as possible…. Until the technology is developed and made cost effective, there will not be an HDTV broadcasting business” (Dickson, 1995).

2.1.1 Rationale for ATP Involvement

DTV broadcasting deadlines catalyzed R&D efforts to make the conversion both efficient and more economically viable for stakeholders. NIST’s ATP was uniquely positioned to organize and support these R&D efforts. ATP competitively selects high-risk, high-reward ventures that may yield significant social and economic benefits. ATP has funded several projects through its focused program “Digital Video in Information Networks” that funds projects to develop technological innovations aimed at increasing the national digital video technological infrastructure (ATP, 2003). When the JV was envisioned, it was seen as an opportunity to energize national DTV conversion and enhance U.S. competitiveness in the broadcast-technology sector (Yoshida, 1998).

ATP’s commitment to the development of new DV technologies provided an avenue for Sarnoff and its JV team to receive third party support for its vision of a new, more economical model for managing digital broadcasting operations. Furthermore, ATP provided support to a research venture that would otherwise not have been funded.

The project’s goal was to develop and commercialize a comprehensive suite of technologies to enable efficient DTV and HDTV studio operations. At the time the project was conceived, DTV stakeholders were concentrating on adopting a DTV standard for terrestrial broadcasting based on Moving Picture Experts Group, Version 2 (MPEG-2) compression. Essentially, stakeholders focused their efforts on the efficient delivery of HDTV to the home. However, Sarnoff believed that “considerable effort would also be needed to achieve a cost-effective flexible HDTV studio and that such a studio should be based on compression technology to help manage the very high data rates required for HDTV” (NIST, 2001).

Studio technologies available in the mid-1990s lacked the capability to efficiently and economically manage HDTV’s enormous data streams—1.5 gigabytes per second when decompressed. Video feed is either locally originated at the studio or distributed in compressed form by networks via satellite to the studio. Conventional technologies require stations to decompress satellite feed to perform studio manipulations, including video edits and insertions. In addition, if the feed is not to be aired immediately, it must be stored locally. According to Sarnoff, bandwidth, or data-throughput capacity, was very expensive in the early 1990s and therefore was a significant roadblock to DTV conversion. At that time, technologies did not exist that could transfer or store uncompressed data streams efficiently.

Sarnoff recognized that substantial technological innovations would be required to provide digital studios with equipment that could perform studio operations without the necessity of decompressing the content distributed by television networks.

The JV envisioned a new breed of broadcasting studio and established technical objectives to produce innovative technologies that would permit processing, distribution, control, data management, and broadcast of compressed video streams. The individual studio components that comprise the JV’s technical objectives are summarized in Table 2-1.

2.1.3 Organization and Mechanics of the HDTV JV

Sarnoff assembled a cross-functional team of broadcasting and information technology industry leaders to develop the requisite technologies. Participation from such a broad array of firms was deemed essential, as potential individual technology outcomes did not, according to participants, “make sense” on their own. For any resulting technology to be truly viable, it needed to coordinate well and be interoperable with other studio technologies. The JV leveraged the individual competencies of each member, such as broadcast equipment R&D, networking, signal processing, and advanced computing, to create a whole greater than the sum of its parts. Each member was able to contribute its domain expertise at periodic meetings and through reviews of each member’s research progress. The original JV group comprised nine firms (see Table 2-2), though Sarnoff invited additional firms to fill gaps in technical competency as nonmembers through subcontracts or informal relationships.

Source: NIST, 2001.

The JV project ran from late 1995 through 2000. Although the project was originally scheduled to run 3 years, it was extended an additional 2 years to adapt to changing technological and market conditions and to accommodate operational trials of technology outcomes. Though JV members and ATP concurred that the project changes were necessary, the project remained consistent with its original technical plan and total project budget (NIST, 2001). It became apparent to JV members that a test bed for R&D outcomes would be highly beneficial, and the New Jersey Network (NJN), a public television (PTV) licensee, was recruited by Sarnoff to join the team. The addition of 2 years allowed the JV to refine its technologies by leveraging technical developments occurring outside of the JV and in reaction to emerging market conditions. The JV recognized the opportunity to collapse many of the individual devices into a single high-performance computer.

Sarnoff formed AgileVision, a second JV co-owned by Mercury Computer Systems, a manufacturer of ultra high performance parallel computer systems. AgileVision did not join the JV officially, yet became the commercialization outlet for much of the JV technology.

ATP provides funding on a cost share basis; JV members must provide more than 50 percent of total JV funding. In this instance, JV members provided 52 percent of the JV’s total funding, or $30.1 million, and ATP provided the remaining funds, $28.4 million. Total R&D expenditures were therefore $58.5 million.

The JV set for itself seven technical goals, all of which were met successfully, though only five of the accomplishments have been incorporated into commercial products. The project’s R&D outcomes are summarized in Table 2-3. Those that have not found full-scale commercial outlets could yet have an impact. For example, IBM and Sun both indicated that their R&D efforts (goals six and seven in Table 2-3) are molding approaches to new product development efforts in other areas. This section provides a general picture of individual JV member efforts, the degree of success, and the current state of commercialization for those efforts. It analyzes member comments and identifies social and economic benefits.

Sarnoff led the overall system design and system integration for the project. Additionally, Sarnoff co-developed several technologies with other JV members. These efforts were most concentrated on the development of two technologies: compressed bit stream switching for splicing, edits, cuts, and spatial effects and a server technology to provide multiple compressed HDTV streams with demanding quality-of-service constraints. (The other technologies in which Sarnoff was involved are discussed under the headings of their teaming partners.)

The compressed bit stream switching technology allows end users to pass compressed digital signals through their digital studios while inserting local content and logos and performing basic manipulations. The technology removes the need to decompress and later recompress signals to perform basic tasks and, therefore, mitigates equipment, bandwidth, and storage pressures in the studio. The server technology permitted compressed domain multicasting: real-time processing of transport feed of four SDTV channels or one HDTV channel.

The compressed bit stream switching and server technologies were commercialized in 1999 through a spin-off entity named AgileVision. AgileVision was operated jointly by Sarnoff and Mercury Computer Systems, a manufacturer of advanced parallel computing systems. It became clear that without explicit arrangements some JV technologies would not make their way to the marketplace. Thus, AgileVision was created as a commercializing entity and worked closely with Sarnoff during the last 2 years of the venture. Mercury Computer Systems was involved in AgileVision because the AgileVision system operated on Mercury’s advanced high-performance parallel computer. AgileVision’s system, the AGV-1000, has garnered several emerging technology awards since its release, including those from Broadcast Engineering and the National Association of Broadcasters (NAB). AgileVision was purchased by Leitch Corporation on February 8, 2002. AgileVision is no longer a stand-alone entity but has become one of Leitch’s product lines.

The AGV-1000, billed as “DTV-in-a-box,” has been well received by the Public Broadcasting Service (PBS) community. AgileVision is an economical choice for many PBS stations. The AgileVision unit allows them to buy a basic system for delivering a digital signal to customers for a few hundred thousand dollars rather than approximately $2 million. Most PBS stations apply for grants from the National Telecommunications and Information Administration (NTIA) for funding of such capital purchases. Thus, cost considerations factor prominently in their equipment investment decisions.

AgileVision is an integrated unit that runs with a high-performance computer. The PBS content distribution model calls for most PBS stations to pass through compressed streams of either four SDTV channels or one HDTV channel, depending on program schedules. The AgileVision system is well suited for this purpose, because it allows the stations to multicast signals while inserting select local content and logos. In this report, “AgileVision” is synonymous with the AGV-1000 and the JV technologies it embodies.

IBM was involved in two areas of JV research: file transfer architecture and video browsing and query. Though neither technology was commercialized through product offerings, IBM transferred knowledge through various consulting agreements with clients. In addition, IBM and, as will be discussed later in this section, Sun Microsystems indicated that the published results of uncommercialized research provided the DTV equipment industry as well as end users with constructs for approaching DTV conversion and digital studio organization.

A functional digital studio needs an advanced network for delivery of digital audio, video, and data. IBM investigated prospective applications of Asynchronous Transfer Mode (ATM) in the studio environment. ATM is a robust technology widely used by the telecommunications industry to deliver large amounts of data quickly. As a group, the JV became aware that working with compressed bit streams made control of these streams challenging. ATM offered the JV the opportunity to route large amounts of data quickly and efficiently through the studio. IBM was ultimately responsible for these routing and networking technologies, and Sun Microsystems was responsible for the command-and-control of the information as it was distributed throughout the studio.

After the initial three project years, the potential of collapsing all the ATM networking functions into a single computer became evident. With the addition of AgileVision, the need for the ATM networking within the studio declined. While ATM networking still holds potential application in large studios and distributed operations, the market pull for this aspect of the project declined significantly.

No technologies were ultimately directly commercialized. However, some technology transfer probably occurred through consulting agreements between IBM, an electronics manufacturer, and a major cable news network to develop an archival system.

The principal impact of IBM’s effort in this area was on the future of networking and control design for digital studios. According to IBM representatives, although there was “very little explicit commercialization that emerged from the venture, from the point of view of influencing the community [the project] had a significant impact.” Knowledge developed by the ATP project influenced other manufacturers, and when reviewing these manufacturers’ products, according to IBM, kernels of JV research are apparent in their product designs, particularly in the realm of control. Manufacturers appearing to have been influenced by papers and conference proceedings include Leitch, Harris, and Miranda. In addition, several of the most salient points that appeared in an article describing a technical roadmap for how broadcasters would convert to digital broadcasting in the SMPTE Journal (SMPTE, 1998), the journal of the Society of Motion Picture and Television Engineers (SMPTE), were developed and investigated during the ATP project.

Sun commercialized the technology in its StorEdge Media Central product line, which was available for 2 to 3 years. The product achieved only limited success, partially because defender systems were entrenched and partially because Sun could not provide the resources to fully develop the infrastructure needed to support the product line in an era of economic uncertainty. The product was withdrawn from the market. A derivative product, Sun Media Appliance Platform, based on Java technology, was released in September 2002.

To permit adjacent signal broadcasting, the FCC decreed extremely tight spectral shoulders for each new channel. Only with a very “clean” signal could broadcasters reliably use adjacent channels. Unfortunately, the power amplifiers used in high-power broadcast transmitters used non-linear amplification (for efficiency), which produces lots of unwanted energy in adjacent bands. The traditional approach for dealing with these unwanted signal components was to apply filters to the transmitter output. These filters, however, are very large and expensive, and significantly degrade the desired signal. With the exception of Thales, all major broadcast transmitter manufacturers petitioned the FCC to relax the specification of spectral purity.

Thales’s concept was to apply predistortion, sometimes referred to as precorrection, to prevent digital signals from bleeding over into adjacent bands, a technique that has been used successfully in analog broadcasting. Simply put, the technology anticipates outof- band spectral products and injects signals equal in amplitude but 180 degrees out-of-phase into the broadcast stream before it reaches the amplifier. The technology automatically cancels out the out-of-phase signals, eliminating the distortion that would otherwise appear in adjacent channels. According to Thales, “Once the operator pushes a button, the problem takes care of itself.” The technology is known as digital adaptive predistortion (DAP).

Thales believes DAP impacted the cost of the digital conversion for its customers, particularly during the 2 years in which its competitors had no equivalent products. DAP reduced the cost borne by broadcasters in meeting their digital transmission mandates. The first economic benefit was the price of the transmitter: Thales purchasers were able to buy 25 kilowatt (kW) transmitters for the same price it would have cost them to buy 20 kW transmitters from competitors (once filters had been installed). They obtained 20 percent more power for the original cost. The new technology reduced set-up time for transmitter installation by an engineer to 1 hour from 1.5 days. Finally, DAP allowed transmitters to more easily meet performance requirements while reducing the number of operations and maintenance hours. These benefits led to Thales receiving an Emmy Award for Technical Excellence from the Academy of Motion Picture Arts and Sciences in 2003.

2.2.5 MCI

MCI developed a digital interface technology that permitted the delivery of “theater-quality digital video over standard digital telecommunications facilities” (NIST, 2000). Individuals knowledgeable of MCI’s JV-related research are no longer with the company; thus, it was not possible to interview the firm for this analysis. The company reported in a 2000 NIST project background piece that their digital transport service is “being used by a major cable TV programmer at 40 percent savings (saving tens of thousands of dollars monthly)” (NIST, 2000).

In addition to the organizations above, several other companies participated in the JV in some fashion, but their activity is not explored in depth in this analysis. For example, Philips Laboratories withdrew from the JV near the end of Year 3 when Philips made the business decision to move its research division from the United States to Europe. In all instances, the role each parting firm played was reallocated to another JV member or subcontractor.

JV members unanimously agree that the project was a productive and valuable experience. However, as is often the case with complex JVs, the HDTV project was at times affected by firms that left as their business strategies or circumstances changed. Understanding that each company’s competency was required to make the project successful, new members and subcontractors were invited to join to compensate for competency losses. Still, the interpersonal and cross-collaborative relationships developed during the project were the key to its success, though the comings and goings of JV members caused some disruption.

Members indicated that the technologies they developed during the course of the JV would not have been developed in the absence of the JV and ATP funding, or if so, not as quickly. Without NIST ATP funding, Sarnoff doubts that JV members would have expended the resources to develop these technologies, let alone make the effort to perform the research jointly. The venture was able to generate significant synergy through facilitation of communication and knowledge exchange. Firms gathered quarterly at each other’s facilities for meetings and to demonstrate technologies and share insights. This coordination was “invaluable.”

The “DTV-in-a-box” technology developed by Sarnoff, and later commercialized by AgileVision, would not have been developed in the absence of the JV’s project. In turn, the transition to DTV broadcasting would have been significantly more costly for PTV stations that adopted it, as will be discussed in Sections 3 and 4. Thales’s introduction of a new transmitter embodying the DAP technology would not have occurred as quickly.

The JV most likely also improved DAP’s quality and lowered its costs while accelerating its introduction. As another example, IBM might have been involved in this type of research, but the JV brought the firm into it sooner. The collaborative JV also developed IBM’s knowledge much further than it would have been had it chosen not to participate or had the JV never existed. Similarly, Sun Microsystems stated that it collaborated with firms and technologies outside of its traditional business model, exposing it to new venues for its own R&D and potential market segments. Sun had been interested in exploring media asset control and management. When the JV presented itself as an opportunity, Sun welcomed it. The ATP funding multiplied Sun’s resources and allowed the company to conduct research earlier than it would have in the JV’s absence.

Interviews with JV members suggest that the venture yielded positive technology outcomes in general, although each member individually met with varying levels of success. Sarnoff and its spin-off, AgileVision, as well as Thales each indicated that they were able to develop and successfully commercialize projects undertaken as part of the JV. The technologies embodied in their product offerings are a beneficial alternative to current and defender technologies. IBM and Sun Microsystems achieved technical success, yet market constraints usurped any potential viability for prospective products. Nonetheless, both firms reaped intellectual property benefits that may inform future technology investigations. The quantitative analysis will focus on the AgileVision system and Thales’s development of DAP, while the benefits accruing to the efforts of other firms will be discussed among future directions in Section 5.

The first area of potential benefit to be analyzed is AgileVision’s impact on the PBS member station market. Both Sarnoff and AgileVision indicated that AgileVision’s AGV-1000 has both onetime and ongoing benefits for this market segment. The hypothesized onetime benefits consist of reduced equipment costs and faster installation turnaround. Ongoing benefits are hypothesized to stem from less labor required to operate the AgileVision unit compared to an alternative equipment installation.

Benefits associated with DAP are also analyzed. The interview with Thales indicated that the company’s JV technology allowed them to introduce a digital transmitter that had cost, installation, and operations benefits over competitor products for two years. The transmitter more easily met FCC spectral requirements than did competitor products, allowing end users to reap routine maintenance benefits and avoid costly filtering. End users also avoided the need to purchase more powerful transmitters to compensate for losses associated with installing more powerful filtering technologies. Furthermore, other transmitter manufacturers invested in R&D to compete with Thales’s technological innovation and introduced products to their customers sooner than would have occurred without the ATP JV project.

Among the suite of JV technologies developed, these two technologies have quantifiable economic benefits. Although qualitative benefits are associated with other JV members’ technologies, there is no indication that those technologies are accruing any quantifiable benefits at this time or will in the near future.

____________________
1. Not all broadcasters chose to install new digital studio equipment; many broadcasters are converting their analog signals to digital using conversion equipment to meet FCC requirements.

2. Comark Communications was the original name of this JV member. During the course of the JV, the entity’s name changed from Comark Communications to Thomcast following a merger. The entity is now known as Thales Broadcast & Media.

Return to Table of Contents or go to next section.

Date created: July 12, 2004
Last updated: July 13, 2004