Bleak forecast
HDTV’s future remains mostly cloudy, with only patchy clearing.
By Maury Wright, Executive Editor

As HDTV approaches a critical phase in its rollout, a gaze into the future reveals a spotty mix of clear pictures ahead, with reception problems dominating the near-term forecast.

But why start off with the bad news? Let's first consider what's going right in the DTV world. A number of consumer-electronics vendors, including Zenith, Panasonic, Sony, and others, now sell HDTV products. Thomson Consumer Electronics (maker of RCA and Proscan) is leading the way in affordable products. For $650, you can buy RCA's DTC100 set-top receiver, which can tune in both terrestrial and satellite HDTV signals. RCA has bundled the DTC100 with a 36-in direct-view monitor for $2500, although the monitor doesn't offer the full glory of HDTV's maximum 16:9 aspect ratio.

Missing pictures

So based on consumer products, you might ascertain that HDTV is ready to move beyond a micro-niche to a much larger group of early adopters.

You'd be wrong. Those early adopters will have very little to watch on their new gear. Not much HDTV. Not even much regular-resolution SDTV (if you need a primer on DTV terminology, see the sidebar, "Decoding DTV").

NBC broadcasts The Tonight Show in HDTV, so if your local NBC affiliate pumps out an HDTV signal, you may be in luck (though the value of watching Jay Leno's monologue in high-def is questionable at best). A look at a recent week's schedule revealed that CBS televised six prime-time HDTV programs and Fox aired two. DirecTV offers the most HDTV content, delivering HBO and pay-per-view movies 24 hours a day.

So what does an early adopter see on a DTV channel when the broadcaster isn't showing original digital programming? It depends on the station. Some broadcast NTSC pictures converted into DTV, but the quality is little if any better than an analog picture. Others run prerecorded HDTV sequences, which are compelling to look at briefly, but won't sit well with a consumer who just dropped anywhere from $2000 to $10,000 on new equipment.

ROAMING SET: Nokia’s Mediavision mobile appliance offers Internet access via a cellular modem and DTV reception based on the European DVB-T standard.

The networks have been waiting for others to foot the bill for delivering HDTV programs. So far, the consumer-electronics companies—the same ones who have invested heavily to develop HDTV products—have bankrolled a lot of the HDTV programming. Thomson paid for CBS to deliver HDTV coverage of the NCAA basketball finals, and Sony did the same a week later when CBS aired the Masters golf tournament in HDTV. Panasonic sponsored ABC's high-definition broadcasts of Monday Night Football last season. Mitsubishi and Samsung underwrite primetime HDTV programming on CBS.

But as Dave Arland, director of government and public relations at Thomson points out, the subsidies are about to end. Arland says Thomson has invested more than $50 million in HDTV product development. "It's time for the broadcasters to stop whining about the cost of HDTV and get on with it," he says.

The public interest

The networks, however, seem anything but sure about where to head next. They're concerned about how many consumers HDTV will be able to reach, how they can make money with HDTV, and whether they should emphasize HDTV or SDTV. In fact, people often misunderstand the FCC's rules about what the networks and their affiliates must broadcast digitally. The FCC imposed a staged timetable, which mandated the start of DTV broadcasts last year for stations in the largest markets, and by May 1, 2003, for all commercial stations. But all this really means is that the stations must digitally transmit NTSC-quality video.

There's substantial disagreement in the industry as to whether the networks are truly serving consumers if they only deliver SDTV pictures via DTV signals. After all, the airwaves are a public resource. Via the FCC, consumers gave the DTV spectrum to the broadcast industry in return for a promised move to DTV and a promise to relinquish the NTSC spectrum after the complete transition to DTV (scheduled for 2006). Certainly, most people thought the barter included HDTV broadcasting.

Dave Arland

Some prognosticators believe that SDTV televisions will prove more popular than HDTV models in the short term. Bob Stokes, director of operations for DTV at Motorola, expects that sub-$1000 SDTV sets with 480-line resolution will appear by year's end in 25- to 27-inch sizes. The sets will decode HDTV signals and downconvert them for display on the SDTV screen, and will also include NTSC tuners. The sets would attract consumers by offering better picture quality than regular NTSC sets.

The networks and TV stations, however, still have one overriding concern: the consumers' ability to receive DTV. As the Inside the Digital Den feature in this issue suggests, DTV reception is far from plug and play today. Some broadcasters, led chiefly by Sinclair, assert that the ATSC's chosen modulation scheme, called 8-VSB (see sidebar, "Decoding DTV"), is fundamentally broken. The FCC, the ATSC, and the Consumer Electronics Association respond that 8-VSB is fine, it's just the implementations that are immature.

For certain, DTV presents complex challenges. You have separate audio and video data that must be decoded and then synchronized. The ATSC standard defines many data channels with unique program IDs that different broadcasters might use in different ways. Throw in PSIP (program and system information protocol) data, which carries program listings and other information, and the conditions are ripe for failure.

In fact, horror stories about poor HDTV reception have abounded recently. In some cases, the problems have been limited to audio synchronization and other minor glitches that clearly can be fixed. Sometimes, however, reception difficulties result in a blank screen. Unlike your analog TV, which shows a bad picture when reception is poor, a DTV set yields either a perfect picture or no picture at all. The Sinclair-led contingent blames all reception problems on 8-VSB. In reality, something as simple as a problem with the PSIP data can lead to a blank screen. Such simple problems will be ironed out as DTV broadcasting systems and consumer products both mature.

Rescue effort?

Still, Sinclair is lobbying the industry to at least reconsider an alternative to 8-VSB. The so-called Grand Alliance of companies, which formed back in 1993 and developed the ATSC standards, considered two modulation schemes: COFDM and 8-VSB. The group chose 8-VSB because theoretically it offers a superior carrier-to-noise ratio. In plain terms, that means 8-VSB offers greater range at a given transmitter power. The group chose 8-VSB based on the belief that it would allow a broadcaster to reach, with a digital signal, any viewer that could receive the broadcaster's analog signal.

However, it appears that COFDM is superior to 8-VSB when it comes to the receiver's ability to accurately tune a signal in the face of multipath conditions. Multipath refers to a scenario where a receiver detects, over time, multiple instances of the same signal. One signal is the original, while the others are reflections from buildings or other obstacles. A regular TV exhibits ghosting when exposed to multipath. A digital TV can't disentangle the real signal from the echoes, so it shows no picture at all. Current 8-VSB receivers have significant problems in areas that are prone to multipath, such as urban canyons.

The million-dollar question: Can 8-VSB receiver designs solve the multipath problem? As of yet, no one has the final answer. COFDM is used in receivers developed for the European DVB-T system. These mature designs are widely deployed in places like the United Kingdom. Meanwhile, most 8-VSB receivers use first-generation demodulator ICs.

Improvements to 8-VSB demodulators will surely come, but will they solve the problem? Motorola's Stokes answers absolutely. "There's nothing wrong with the ATSC standard, but rather some early implementations that don't handle ghost cancellation," he says. The problem originates in the demodulator's equalizer section, which is designed to deal with the echoes. First-generation demodulators could only handle multipath if the original signal and all echoes were detected within a window as short as 12 microseconds. Stokes says Motorola's newest design can operate with a 40-microsecond window.

On the other hand, Mark Schubin, a consultant and leading independent authority on DTV, paints a picture of a more complex problem. He notes that especially with indoor reception, an echoed signal may be stronger than the original signal. Moreover, the echoed signal could actually arrive earlier than the original signal, if for instance the echo arrived through a window while the original traveled through a wall.

Schubin agrees that a longer equalizer window is important and concurs that demodulator vendors are delivering on that issue. But he points out three additional problems that must be resolved.

First, the equalizer window must be able to lock onto the strongest signal, even if it's an echo and regardless of whether it arrives before or after the original signal. Second, the demodulator must more quickly deal with echoes once they're detected.

Finally, the demodulator needs to provide a function called synchronous detection. Synchronous detection helps the receiver function when atmospheric or other conditions alter signal characteristics. Most current receivers will continuously try to lock onto the strongest signal, but the signal claiming the title of "strongest" may change from moment to moment. The signal processing involved in constantly locking on to a different signal can prevent the receiver from ever successfully demodulating the signal. Synchronous detection would allow the receiver to choose one signal in the equalizer window and stick with it, even if it's not always the strongest.

Schubin expects demodulator vendors to solve the first two problems in what he terms their fourth-generation chip sets. He anticipates such chips shipping in 2001, with consumer products following the next year. Schubin isn't convinced, however, that the demodulator vendors know how to implement synchronous detection, and no vendors are ready to describe just how they will address the problem.

Proponents of 8-VSB ask whether the industry should even consider a major change based on indoor reception problems. After all, indoor reception is a problem with NTSC signals as well. Many consumers already receive DTV via rooftop directional antennas, and that's the scenario under which the ATSC standard was conceived. Moreover, proponents point out that COFDM would require higher transmitter power and that COFDM products are far more susceptible to impulse noise from appliances such as hair dryers.

Petitioning

Most 8-VSB proponents—and the consumer electronics vendors are the staunchest of these given their investments—wish the COFDM controversy would just go away. But don't count on it. Sinclair floated a petition requesting that the FCC allow broadcasters, and therefore the market, to decide between the two technologies. Thus far, the FCC has turned down the request, winning cheers from the CEA and a number of manufacturers of both ICs and consumer products. But the FCC is asking for comments on the matter in preparation for its biannual DTV review meeting, scheduled for this month. Stay tuned to www.commvergemag.com for relevant news.

The 8-VSB folks claim a change is completely impractical given the huge amount of money broadcasters have spent on 8-VSB. Sinclair counters that a $7500 card could convert an 8-VSB transmitter into a COFDM transmitter. The 8-VSB proponents say a change would be unfair to consumers who have already bought DTV products. The CEA cites strong sales numbers to buttress this argument, but the COFDM side claims that the sales forecasts mix satellite and terrestrial numbers in order to inflate the size of the DTV customer base. Moreover, as Schubin points out, "Broadcasters using 8-VSB today can legally transmit signals that early 8-VSB products can't receive."

The lab tests that led to 8-VSB's selection assumed that the main obstacle to DTV reception would be noise. In real deployments, it appears the real obstacle is interference. Sinclair is on a mission to prove this point. At the recent NAB (National Association of Broadcasters) show, the company demonstrated COFDM reception in the Las Vegas Convention Center. The company had one of its local stations, 17 miles away, broadcast a signal at a power level of 200 watts. Sinclair tuned the signal in not only with an indoor antenna connected to a DVB-T receiver, but also on a Nokia Mediavision. This quintessential convergence product is a mobile Internet appliance and DTV receiver. It includes a color LCD, a DVB-T receiver, an integrated antenna, and a cellular modem.

The 8-VSB proponents immediately charged that the ATSC standard was designed for fixed-position reception and that the Mediavision demo was simply a publicity stunt. Sinclair and other COFDM proponents have long pointed out that the technology would be better than 8-VSB for mobile applications. And broadcasters do plan to use spare DTV spectrum to deliver data services to fixed and mobile customers. Certainly the Mediavision demo helped make the mobile point. But the demo also showed that even a simple antenna could receive the COFDM signal inside the convention center—a relatively hostile environment for TV reception. Moreover, 200 watts is far less transmitter power than many broadcasters use; some DTV stations broadcast at 1000 watts. On the other hand, flat Las Vegas probably offers the ideal terrestrial-DTV geography, despite the presence of the faux Eiffel Tower, the replica Great Pyramid, and other structures.

Whether the 8-VSB crowd likes it or not, the NAB demonstration may sway public opinion on the modulation scheme. At the very least it will cause more people to take a closer look at just what Sinclair is proposing.

Sinclair has always maintained that it isn't asking to replace 8-VSB; it just wants to give broadcasters a choice. Opponents say doing so will delay the adoption of DTV by alienating existing and potential customers, perhaps interfering with 8-VSB reception, and requiring consumer-electronics companies to support both standards. Off-the-record, 8-VSB proponents even claim that Sinclair is actually trying to delay the move to DTV, so that it won't have to spend money equipping its broadcast empire for DTV.

Try both

Schubin sees it differently, and believe me he is agnostic when it comes to this issue—his consultant relationships with companies on both sides mandate a neutral stance. Schubin finds no reason that the FCC shouldn't open the door to COFDM evaluation. He points out that the ATSC would have to develop allocation tables and specifications for any COFDM deployment, and that realistically it would be three years or more before broadcasters could turn on COFDM signals. "Within these years, if 8-VSB can be improved and the reception problems solved, then there would never be any COFDM deployment anyway," he says. If 8-VSB ultimately proves unsuitable, Schubin figures, then the Sinclair concept would move the industry immediately toward a COFDM backup plan.

That's what Schubin thinks the FCC should do. In reality, he believes no action will be taken on COFDM. In fact, he argues that DTV signal distribution will ultimately move to cable and satellite. Those technologies offer the only near-term ways to eliminate reception issues, he asserts.

"HDTV will kill the local affiliate and independent stations," Schubin continues. In the near term, he believes, HDTV will be marketed as a premium service, because the networks seem to have no other way to earn revenue beyond the ad revenue they already make on NTSC programs. And even though they got the DTV spectrum for free, the networks want to recover the investments they're making in HDTV. In the past, the networks have stood behind the local stations, even going so far as to prevent satellite customers from getting out-of-market broadcasts. However, for the right price, Schubin believes, the networks might start selling programming, such as HDTV sports, to cable or satellite operators.