Dr. Dobb's Journal November 2003
For years, the Federal Communications Commission (FCC) has been spurring the U.S. economy toward digital television. In fact, the FCC has set an aggressive goal of completing this digital television transition by 2006 so that the Commission can free up the analog spectrum for other uses.
In 2000, the 2006 target appeared out of reach as the HDTV rollout was mired in consumer apathy and suffered from a dearth of content. Given the sluggish rollout, HDTV was relegated to wealthy early adopters who enjoyed purchasing advanced technology and didn't mind the paucity of HD channels. Although the number of viewers was minuscule, broadcasters such as CBS continued to pour millions of dollars into HDTV equipment so they would be positioned for the eventual digital switchover.
However, due to plummeting television prices and the increasing availability of content, HDTV has transitioned from the exclusive domain of the elite to widespread availability and growing market awareness. In addition, in most major U.S. cities, viewers have access not only to local Over-the-Air (OTA) HD signals, but also HD content from national broadcasters such as ESPN, HBO, and Showtime.
While HDTV was morphing into a consumer platform, there was a simultaneous explosion of broadband cable and DSL connections in U.S. homes. During the earliest phases of the HDTV rollout, the majority of Internet connections were 56 kbps. Consequently, file sharing of pirated HDTV streams recorded at 19.2 mbps was impractical since it could take a few months to download a single feature-length movie. Furthermore, since only early adopters could afford HDTV displays and set-top boxes, the potential impact of such piracy was infinitesimal.
By contrast, today, there are millions of people with both HDTV and broadband Internet connections, and the entertainment industry is petrified that if you combine high-speed Internet and HD streams with easy access to advanced file-sharing programs and web sites, you have the building blocks for mass piracy. While it's true that it still takes days to download a three-hour HD movie at today's broadband transfer speeds, this roadblock will vanish as connection speeds increase and compression technologies improve.
The rapid evolution of the Internet has spurred CBS and other broadcasters to issue an ultimatum to the government: Either protect their investments in HD technologies by mandating the use of a "broadcast flag"—a sequence of bits embedded in a TV program that signals that the program must be protected from unauthorized redistribution—or they'll cease HD broadcasts and potentially derail the digital-video transition. Since pirates will now be able to make high-definition digital copies of precious content, these companies believe they are in an extremely vulnerable position and only government intervention can rescue them from the relentless advances in technology.
Although the entertainment industry gives the impression that there is widespread agreement on the concept of a broadcast flag, consumer-rights organizations are equally passionate that it is a threat to individual freedoms. In reality, neither view is precisely correct. The broadcast flag by itself is benign. Rather, the proposed rules and regulations attached to it are what ignite controversy.
In 1997, the Advanced Television Systems Committee (ATSC) issued the "Program and System Information Protocol For Terrestrial Broadcast and Cable (Revision A)" or PSIP specification. This document (ATSC A/65-A) and associated amendments define the data structures that are the fundamental building blocks for the digital-video industry.
The essential feature of ATSC A/65-A is a data structure called a "descriptor" (see Table 1). Each descriptor has a tag (or identifier), a length, and a block of data (the exact meaning of this data block varies by descriptor tag). While the ATSC has defined certain core descriptors such as video, audio, and caption needed by current video streams, additional descriptors can be added to the specification as digital-video technology evolves.
The Redistribution Control Descriptor (RCD), found in Amendment A of the PSIP specification, was one of the first descriptors added by the ATSC (see Table 2). The ATSC says that when this descriptor is present, "technological control of consumer redistribution is signaled." In other words, the RCD lets content providers alert MPEG-2 transport stream demodulators that there are restrictions on the interfaces on which they are permitted to redistribute (or forward) this content. While this isn't an explicit form of copy protection, it is an implicit flavor because it can restrict the type of devices that are authorized to make copies.
If you look at that RCD data structure in Table 2, you'll see that it contains only one field—rc_information. At a high level, the RCD alerts demodulators that redistribution of the content has restrictions, while the rc_information field gives specific details on the meaning of these limitations. Alas, the ATSC has yet to provide details for this field and indicates that it may be addressed in a future revision of the specification.
Given that the ATSC hasn't completed its work with the rc_information field, what we're left with is the rather nebulous RCD and no guidelines on what to do when we encounter it. Although the ATSC probably didn't realize the implications of leaving the rc_information field undefined, the ramifications of this indecision were immense because, a few years later, the industry became obsessed with RCD and eventually nicknamed it the "broadcast flag."
Because the RCD definition is so vague, the Broadcast Protection Discussion Group (BPDG), a subgroup of the Copy Protection Technical Working Group (CPTWG), was tasked with researching the issue and publishing a draft that could be considered by Standards bodies and governmental agencies such as the FCC.
The BPDG final report was completed on June 3, 2002, and submitted to the FCC. The FCC, in turn, requested comments beginning on August 9, 2002. The most important aspects of this draft revolve around the protection status of digital outputs and the video quality permitted on both analog video connections and legacy computer digital interfaces.
The BPDG report requires that encryption be enabled on digital outputs if the stream contains the RDC. In "Digital, Analog, and High-Definition TV" (DDJ, November 2002), I pointed out that there are two copy-protection technologies for HD digital-video interfaces—High-bandwidth Digital Content Protection (HDCP) and Digital Transmission Copy Protection (DTCP). The BPDG document would require HDCP to be enforced on DVI connections and DTCP on 1394 interfaces (see Figure 1) when the RCD is present.
While newer HDTVs have DVI/HDCP connections, the majority of first-generation DVI HDTVs and virtually all DVI computer monitors don't support HDCP and, as a consequence, they pose a potential security risk. The BPDG draft recommends that the video image produced by set-top boxes be constrained (or reduced in resolution from High Definition to Standard Definition) by the MPEG-2 demodulator before it is transmitted over a nonHDCP DVI connection to the display (see Figure 2).
Again, in "Digital, Analog, and High-Definition TV," I mentioned the DTCP descriptor has a specific feature that enforces constrained images on analog connections when the content creator requests it (see Table 3). In contrast, there is no such field in the RCD. As a result, the BPDG final report explicitly states that analog outputs should not be constrained simply because the RCD is present in the stream. If the content creator is concerned about piracy via analog connections, they can't rely on the RCD to protect them.
A third issue that the BPDG report addresses is legacy computer DVI connections. Although it was a source of much debate within the group, the final recommendation was that computers could transmit high-definition images over nonHDCP DVI interfaces without requiring the video to be constrained. Conversely, consumer electronic devices such as set-top boxes are required to constrain video if the RCD is present but the display is not HDCP compliant.
Because the BPDG report represented a broad coalition of companies ranging from movie studios to consumer electronics and computer manufacturers, it produced a consensus report. And like most consentaneous recommendations, there were aspects of the report that were controversial and significant disagreements remained. For example, the companies represented by the Motion Picture Association of America (MPAA) were strenuously opposed to computers being able to transmit unconstrained HD video over unprotected DVI connections. Consequently, when the FCC requested comments on the BPDG report, the MPAA submitted a draft proposal to the FCC that eliminated this capability.
Besides the MPAA, thousands of other interested parties have flooded the FCC with comments on the BPDG proposal. Large companies such as CBS/Viacom submitted comments indicating that they will cease HD broadcasts in the 2003-04 season unless the FCC issues a positive ruling in favor of the broadcast flag.
On the other hand, consumer-advocacy groups such as the Electronic Frontier Foundation (EFF) and the Computer & Communications Industry Association (CCIA) have filed comments expressing their concerns about the implications of the broadcast flag. For example, the EFF believes that if adopted, the BPDG draft will stifle open-source, digital-video development projects. To explain, in today's environment, anyone can download the MPEG-2 specifications from the ATSC web site and create an innovative digital-video product. However, if the broadcast flag proposal is adopted, open-source developers would have to license proprietary security protocols such as DTCP or HDCP from external parties. Since open-source projects typically can't afford such expenditures (or the legal risks associated with licensing proprietary security protocols), the EFF believes this proposal will have a chilling effect on digital-video development.
The CCIA's comments are even more biting. It feels that a few key players controlled the decision-making process of the BPDG and that critical consumer-advocacy groups had no meaningful input on the final proposal. Furthermore, they believe that private industry should have been given the opportunity to resolve this issue before the FCC got involved.
All comments regarding the broadcast flag were originally due to the FCC by December 2002. However, due to the immense interest and volume of comments, the FCC extended the deadline to February 2003. In the end, the FCC received over 6000 comments on this incendiary topic.
This may be one of the most difficult issues the FCC has ever tackled. It must try to balance the entertainment industry's legitimate concerns about rampant Internet piracy against the possibility that consumer rights will be trampled and open-source, digital-video development will be devastated. In the end, the CCIA may be correct—these issues are better solved by the market rather than by a government agency.
While the FCC's final decision is still unknown, what we do know is that the broadcast flag proposal under consideration at the FCC will not cause analog outputs to be constrained nor will it disable older HDTVs with analog component inputs. And for that, millions of consumers will be grateful.
DDJ