Digital Audio Radio: Two perspectives

The Dimming Prospects for DAR in the United States

by Ralph Justus, Director of Engineering
Consumer Electronics Manufacturers Association

103rd Audio Engineering Society Convention
The Broadcast Forum on Digital Audio Broadcasting
September 26, 1997, reprinted by permission.

Consumers want their music digitally. Our research and sales statistics confirm this. Digital alternatives to radio are already the preferred method of listening to music at home.

This was recently re-affirmed by a study CEMA conducted earlier this year where we surveyed Direct Satellite System owners. The results establish a significant. new listening trend. We surveyed 1,000 DSS owners, 87% of whom had DSS for less than 12 months. The DSS audio programming package contains 30+ channels of CD-quality audio, commercial-free with a wide range of formats.

We found that 69% of DSS owners were listening to their digital music channels regularly -- -an average of 10.5 hours per week. Sound quality is the most important reason (52%) they-prefer digital audio radio, ranking ahead of "no commercials" (31%). When DSS audio channels are introduced into the home, it is likely they completely supplant over-the-air pleasure listening. Previous CEMA research indicated consumers. on average, listened to 11.8 hours of radio per week at home. Unless total radio listening drastically increased after DSS, digital audio pulls away approximately 3 billion over-the-air listening hours per year. And DSS unit sales will increase 12% in 1997. When asked whether they would be interested in receiving CD quality sound from their over-the-air radios, some 58% of DSS owners were interested. The importance of sound quality was ranked by age showing 80% of those under 26 years old believed it to be important.

Those are the listeners of the future.

The trend is clear. Consumers with access to higher quality audio overwhelmingly prefer it. and more importantly, devote their listening time to it.

In 1991, we saw over a dozen technologies proposed to broadcast digital audio. They promised CD-quality sound. immunity to interference, and ancillary data capacity. Consumer electronics manufacturers saw real potential for a new class of audio products with wide-spread appeal to listeners -- DAR. This would also be the opportunity for radio broadcasting to participate in the digital revolution.

CEMA formed its DAR Subcommittee to centralize system evaluations through an industry-led process, fair and open. We proceeded to:

• open the door to formal proposals,
• test and evaluate them,
• select one and
• recommend it to the FCC to approve

However, by the time we completed the system tests and evaluations, events confounded the ability to complete the process.

Let's now look at a few highlights of those findings:
This shows the systems tested.

• the Eureka 147/DAB system in both 224 and 192 kbps modes
• the VOA/JPL satellite system S-band
• the AT&T/Lucent IBAC system
• the AT&T/Lucent/Amati IBOC system, with both dsb and ssb modes
• the USA Digital Radio FM-1, FM-2 and AM IBOC systems

None of the IBOC systems were field tested.

Multipath performance is key to DAR's success. Urban mobile reception is the harshest broadcast environment. Most radio listeners are here. And the car is where we find the strongest sales growth of CD players -- In-dash CD players had a 30% unit sales increase just this past July. In short, DAR will live or die depending on how it performs with mobile reception.

We defined multipath scenarios representing typical reception environments using both Rayleigh and Doppler fading. Other tests examined airplane flutter and overall delayspread/Doppler performance. The good news is that wide-band OFDM was an excellent choice by Eureka-147. It handles frequency-selective fading and flat fading with an elegant, intricate multi-carrier design of coding, interleaving, symbol period and guard interval. This makes it resistant to most all types of multipath - better than the other systems tested. The bad news is that it’s wide-band and requires new spectrum.

Receiver characteristics limit IBOC use. Here you see IBOC digital energy inserted in the first adjacent FM channel. Any wideband FM receiver sees this energy as increased noise. In the lab tests, we observed a signal-to-noise degradation of up to 24 dB on some receivers. We estimate that IBOC is not compatible with perhaps more than 90% of in-use receivers.

Spectrum congestion further limits both IBOC and IBAC implementation. Here we see 47 received signals measured throughout the FM band at a typical location in New Jersey.

This drastically limits IBOC digital coverage. Using the lab test results on 1st adjacent digital-to-digital susceptibility, only 14 of these 47 stations would be listenable if IBOC were implemented. IBAC is also constrained by current spectrum use.

Field tests in San Francisco showed general agreement with the lab tests, but with one very startling result: S-band frequencies fade early and often. That alone is not so startling, but the implications are. Let's view a tape and you will see what I mean.

TAPE INTRO

• numerous S-band outages due to foliage, signs, etc. on the coast route
• chart showing outages on South route ("clear" reception)

Audio quality subjective tests compared each system to a CD reference. These rigorous tests were conducted at the Communications Research Centre. Some systems performed very well. Others, such as the AM IBOC system at a low bit-rate, were judged in the "annoying" range under clear channel conditions. This was not very promising.

The more serious IBOC incompatibilities and performance handicaps overshadow their inherent audio quality capabilities. We could spend more time debating the relative merits of MUSICAM and PAC and their reduced bit-rate performance. And I know this is near and dear to the hearts of many in AES. But in the IBOC context, it's much like having a fashion review of the necktie worn by a 1600-pound gorilla. It just doesn't matter. Until IBOC can overcome its multipath, coverage and interference challenges, any audio quality advances will never reach listeners.

We conclude that:

• IBOC does not produce CD-quality audio compatibly with existing stations, and still provide robust digital coverage.

• IBAC has similar problems. but more importantly, needs spectrum already used by FM stations.

• S-band satellite can't succeed without a host of terrestrial gap-fillers in every city. Without them, program interruption is repeated and continuous.

• Eureka-147 works -- and works well. Unfortunately, it also requires new spectrum.

What we do with such wisdom is unclear.

Eureka-147 was unquestionably the superior performing system. in most all categories of tests. It provides CD-quality sound, unimpaired and seamlessly. And, not to be overlooked. has a host of other attractive features shown here. The test results demonstrate the technical superiority of a new-band/wide-band DAR implementation. The Eureka-147 designers deserve credit for their achievement. (I'm amazed to this day that it was designed by a committee.) Unfortunately it is all dressed up with nowhere to go -- here in America. It needs new spectrum to exist. But there currently is no such allocation initiative in the U.S.

IBOC does not require new spectrum. It could be implemented quickly on the existing broadcast infrastructure. USADR and NAB list many IBOC advantages (including economic? regulatory and enhancement rationales) -- for broadcasters. Unfortunately, IBOC does not work for listeners. Some say it cannot work. They say that the theoretical data capacity of a narrow RF channel fundamentally limits its performance. The test results so far seem to confirm this view.

Rather, I should say IBOC is "functionally challenged." But, let's be specific: It has reduced data capacity due to its narrow RF channel occupancy. This forces quality trade-offs in both audio and transmission coding - which limits the robustness of coverage and mobile reception. Injecting digital energy in the 1st adjacent channel, degrades reception of both the host station and interferes with other stations. In its current incarnation, it does not meet the "CD-quality" threshold - not enough to become widely accepted with listeners.

Consequently, as you will hear, the latest IBOC R&D efforts lower their reference from a "CD-qualify" standard. to something more generic like "digital quality," "FM-quality," or even, "better than AM." I doubt these marketing slogans will attract new listeners in droves -- not in today's digital climate. IBAC is also narrowband, limiting its multipath performance under flat-fading conditions. It also needs "vacant" unused FM spectrum. But there's none to be had -- not in the major markets where implementation is critical.

For these reasons, receiver manufacturers conclude that a successful DAR service in the United States requires a new spectrum allocation -- for a wide-band implementation. The original 50 MHz S-band allocation was cut in half by Congress and auctioned to the "Wireless Communications Service." leaving the surviving satellite DARS applicants with a 12.5 MHz-wide channel each. But it's at the wrong frequency for a compelling broadcast service. The outages from blockage are too severe and repeated - and would require scores of terrestrial repeaters in every city to restore service. Even then, the Doppler spread at S-band may limit reception to speeds under about 40 mph. Apparently, the urban mobile market does not enter into satellite DAR business plans.

The shame of it is that S-band could have been designed to work. Diligent designs of symbol duration, guard interval and carrier spacing might work, but at the expense of requiring greater overall signal bandwidth. However. when Congress cut in half the DAR S-band allocation, it eliminated any possibility for this approach to succeed.

My outlook for what will occur?

IBOC advocates may continue to look at their designs, perhaps even improving on past efforts. But they have a long way to go. Perhaps a major new technology breakthrough may be found. This would be welcomed news and clearly needed for IBOC to succeed.

But all we have heard so far simply adjusts the current designs - trading off quality, compatibility and coverage - to force-fit the current "IBOC-only" policy directive of broadcasters. I look forward to hearing something new later in this session [or this year? - or next year?] - if it's truly new. But it may already be too little, too late.

"CD-quality" means more than just audio fidelity and dynamic range. It also means interference-free. And that is a very high hurdle to clear for mobile reception. Today, IBOC performance pales in comparison to the Eureka-147 system. The IBOC proponents and broadcasters admit this. Broadcasters continue embracing IBOC (warts and all). But. the public couldn't care less where the sound comes from, as long as they like it. And what they like is "CD-Quality." Not, "Better than AM." Not even, "Near-CD Quality."

"Interference-free" also means a wide-band system is needed to handle flat fading multipath. Studies show that 2 MHz-wide channels are ideal. Eureka-147 already compromised that somewhat by going to 1.5 MHz channels. But the current IBAC and IBOC configurations are only 20-25% of that -- not enough to successfully counter flat-fading multipath.

The Eureka-147 system is the best technology developed. Those arguing against it do so NOT because of its capabilities, BUT RATHER to see whether something less capable might suffice. I personally do not believe that a "good enough" approach will succeed in the long term. Digital technologies are revolutionizing both consumer and professional electronics NOT simply because they are digital. Rather. because digital is better. And not just marginally better, but approaching the best that we can attain.

I believe that is where we should set our sights.

Perhaps after another year or two, others will recognize the value of the new-band/wideband approach. Then we will be where Europe was more than a decade ago.

New spectrum is required for a successful DAR service. Where will it come from, and who will seek it?" remain unanswered questions.

So far. manufacturers hesitate to begin an initiative for DAR spectrum. Right now it would have low probability of success. The opposition is the US government (including the White House, Congress, the Defense Department, the FCC, and NTIA), the aeronautical industry. and the broadcast industry. Formidable foes, especially when combined. Manufacturers will more likely put their attention on other, more established and profitable markets.

Without spectrum, the prospects of having a successful DAR service in the U.S. are very dim. Meanwhile, radio's market share continues to shrink. Digital media listening and sales continue to grow.

Finally, all this delay in DAR may have served a purpose. Soon, the public will have multi-channel audio pervading their homes and cars with DTV and DVD: CD-quality (or better) with discrete 5.1 channel surround-sound. Multi-channel has not yet entered the DAR discussions. Soon, "CD-quality" may not be enough for listeners. But "digital multichannel" might spur the market to excitement.

It's at least something to think about for the future.

I look forward to your comments and questions.

Ralph Justus joined the Consumer Electronics Manufacturers Association, a sector of Electronic Industries Association, as Director of Engineering in February, 1991, where he works on the issues of electromagnetic compatibility, radio and television interference, high definition television, digital radio systems, non-ionizing radiation, and consumer electronics /cable television compatibility.

Previously, he was Director of Engineering, Regulatory and International Affairs for the National Association of Broadcasters. There he worked on issues involving broadcast radio and television technologies, auxiliary and satellite systems, spectrum management, proceedings at federal agencies including the Federal Communications Commission, Environmental Protection Agency, Federal Aviation Administration, the Department of State, and participation in the International Telecommunication Union.

You can reach him at ralphj@eia.org

Local digital radio gets closer to reality

by John Merli, special correspondent
Broadcasting and Cable Magazine, reprinted by permission

Prototype system begins field tests in spring; if successful, stations could be on air in three years.

National satellite radio moved a step closer to existence (at least on paper) this month when the FCC issued licenses to CD Radio and AMRC. For thousands of current terrestrial AM and FM broadcasters who could join the digital era via digital audio broadcasting (DAB), it's been an anxious and frustrating waiting came. But the wait may be coming to an end.

Late next spring, a new DAB prototype being developed by USA Digital Radio (USADR) and Lucent Technologies should be ready for extensive field tests. USADR. whose partners include

Westinghouse, CBS and Gannett, says that if all goes according to plan—and there are several ifs still to consider— DAB could begin emerging within three years. Radio broadcasters have heard all this before, but this project team already has worked out many of the bugs that plagued earlier prototypes.

Back in 1994. following a series of technical mishaps (mainly interference and multipath problems) during field tests of a different in-band/on-channel (IBOC) scheme in San Francisco, broadcasters. engineers and their respective companies decided the only solution was to start over. They went for a "total redesign'' of a workable IBOC system, according to NAB's John Marino, vice president for science and technology. NAB had strongly supported the original IBOC effort and backs the current endeavor (although not financially), while repeatedly characterizing proposed subscription-based national satellite radio services as a threat to radio's traditional localism.

Today, in a Baltimore laboratory and elsewhere. USADR and Lucent Technologies are attempting to create a new type of digital transmission standard for the radio industry. Once again, it will be IBOC. which will allow the use of current frequencies now held by most broadcasters. The much-publicized Eureka system being implemented in Europe (and probably Canada) via L-band frequencies is technically possible in the U.S., but not available; L-band frequencies already are being used by the U.S. military.

The USADR/Lucent collaboration which began a few months ago, partly stems from Westinghouse’s new interest in digital radio following its acquisition of CBS. CBS Radio Group has long been active in digital development Lucent. barely more than a year old, brings the resources of the old Bell Labs

Lucent devised PAC (perceptual audio coding), which has been incorporated into the current system. Suren Pai, Lucent director of business development, says PAC is an audio compression technology that eliminates audio frequencies that the human ear cannot readily perceive. It then compresses the essential audio that remains. PAC was not part of the first IBOC efforts a few years ago.

Jeff Jury, vice president at Westinghouse Wireless Solutions, says the ongoing research in Baltimore (as well as in Cincinnati and New Jersey) is rapidly leading engineers toward the final—and crucial—stage of real-world field testing: "We were faced with correcting the 'first adjacent-channel' interference problem and the multipath distortion [from earlier IBOC schemes] caused by tunnels and trees and other objects. By the second quarter [of 1998], we should be out in the field [and testing] this new system."

Jury says the IBOC team in Baltimore already has worked through a "theoretical design" of a newly revised DAB solution and is into the "detailed design," which eventually will include lab simulations and other tests. Preliminary talks with the FCC could begin in a few months. He predicts that within three years or less "you'll begin to see a real aggressive push" toward bringing DAB to the industry and consumers.

Lucent's Pai says this latest IBOC scheme will cause only a minimum amount of disruption to the basic structure of the radio industry: "Frequencies will stay the same. although we don't know how much power will be required yet. Basically you need less digital power for the same signal strength as analog. Most transmission sites can remain the same, too." Westinghouse's Jury adds that local broadcasters still will need to purchase new digital exciters ("probably under $20,000") and likely will have to retune their transmitters. Other start-up costs are expected. but nowhere near the digital costs on the television side.

In the current IBOC project, Pai says, "for the first time we have a world-class team that understands what [local broadcasters need], and we’re using a much more practical and realistic approach" to convergence. The sole radio broadcaster directly involved in the effort is the CBS Radio Group. Its director of engineering, Glynn Walden, is a seven-year veteran of past IBOC wars and the only broadcast engineer involved in the project.

"We continue to make little breakthroughs every week," Walden says. "Some of the discoveries were originally methods and approaches we had devised for other situations. That's one of the big advantages of having a track record on IBOC research." Another advantage, he believes, is having a team that is specifically attuned to conducting product-development projects.

Incorporated into the current IBOC system will be the potential for LCD-type displays on new receivers for a variety of possible data—weather forecasts, sports scores. product information (ideally popping up on the LCD as a specific advertising spot is being aired)—and other enhancements for the consumer. Off-air "data broadcasting" to niche customers also remains a future possibility and another potential source of station revenue.

Any new digital standard will have to be teamed with the current analog standard as part of a still-undetermined transition period. This is a complicated problem to solve," says NAB's Marino, "and a lot of compromise is involved to keep the current analog system viable while adding digital and creating a receiver that will allow both." Lucent's Pai also maintains that getting both analog and digital on new receivers is "a business issue. It's the chicken-or-egg thing. Manufacturers have to know that consumers can actually pick up digital signals before they can proceed to build and sell the new receivers."

Once a new IBOC system goes forward into the marketplace, the auto industry may need four to five years to introduce DAB into new car models. Pai points out. however, that nonvehicular receivers—as well as aftermarket car radios—could be on store shelves within a couple of years after final approval.

And when DAB eventually reaches the American consumer, what will IBOC digital really sound like? Westinghouse's Jury says the FM signal will take on "near-CD quality", while digital AM will sound "closer" to today's analog FM.

Pai stresses that although not all terrestrial radio stations will be able to convert to digital at the same time, eventual conversion will be inevitable—"digital itself is inevitable. You see that in every aspect of life, and radio is no exception." A transition period. no doubt, will span several years, perhaps similar to the conversion timetable mandated for digital television.

"The world is going digital," Pai says. "There is no going back."

John Merli spent 14 years in public radio -- as reporter and news director at Minnesota Public Radio for nine years, and as a daily magazine host and producer at WAMU 88.5 in Washington, DC. He was a press secretary in the U.S.Senate for two years, and from 1986-97 served as a public affairs officer at the National Association of Broadcasters. Shortly after writing this article, he was named Associate Director of Strategic Planning for Digital Television --which he admits is the longest title he's every had. He can be reached at jmerli@pbs.org.

Reproduced with permission from BROADCASTING & CABLE, copyright 1997 Cahners Publishing Company, a division of Reed Elsevier Inc.
BROADCASTING & CABLE is a trademark of Cahners Publishing Company, a division of Reed Elsevier Inc. All rights reserved.