Digital Multicasting Rollout Begins

The consortium of major broadcasters pushing digital radio are wasting no time deploying what they believe to be its “killer app,” multicasting – the ability to split a single radio channel into multiple program streams. Earlier this month they announced the rollout of digital multicast signals in several dozen markets. The broadcasts introduce industry-coordinated secondary program channels featuring formats like “Classical Alternative,” “Coffee House,” “Female Talk,” and “Extreme Hip-Hop,” and some miscellaneous strangeness. For now, these channels will be offered without commercials.
My ongoing research into digital radio is dredging up lots of interesting information, much of which has a direct impact on the viability of multicasting.
1. The maximum bitrate of an HD Radio broadcast signal is 96 kilobits per second. This bitrate is achievable only on FM channels, and multicast streams must fit within this restriction (digital AM signals max out at 48 kbps and presently have no multicast capability). The way stations are deploying multicasting involves dividing a digital radio signal into one 64 kbps stream and one 32 kbps stream, or two 48 kbps streams.
The most elegant encoding algorithm in the world cannot produce “near-CD quality” sound with such limited bandwidth (considering uncompressed CD audio has a bitrate in excess of 1,400 kbps, HD Radio signals convey less than 10% of the original program audio data).
2. The HD Radio broadcast framework, called In-Band On-Channel (IBOC) digital audio broadcasting, was not even designed to accommodate multicasting until National Public Radio undertook crash development of the capability in 2002. Interestingly, up until that point NPR had been a somewhat reluctant participant in the HD Radio development process, having expressed concerns throughout the 1990s that the IBOC framework was too cumbersome, prone to interference, and bereft of the bandwidth necessary to provide robust digital signals.
NPR was an early backer of the digital radio broadcast framework adopted by most of the rest of the world. Had it actually put any actual effort behind support of this alternative, might it have found traction with the FCC? We’ll never know: the FCC signaled early on that it would champion no standard, preferring instead to let broadcasters find their own. Commercial broadcasters then cooked up the scheme to plunk digital signals onto bandwidth divided up by analog rules, leading to the magic of today.
fmibochybrid3. IBOC digital radio signals do not fit neatly into spectrum partitioned for analog use. They take advantage of the FCC’s “spectral emission mask” rules to occupy up to twice as much spectrum as analog radio signals do. However, as the signal widens, FCC rules require that its relative power drop off by orders of magnitude.
An HD Radio signal is essentially a hybrid analog/digital signal. Its configuration keeps the analog portion of the signal nearest the center frequency and puts the digital data on “sidebands,” which are broadcast at 1/100th the power of the analog signal. Thus listeners should expect the coverage area of multicast channels (especially new secondary channels) to be somewhat limited. (See figure at right; click for a larger version with explanation.)
4. There stands a good chance that digital radio broadcasts will be locked down with anti-copying technology. The National Association of Broadcasters and Recording Industry Association of America are corresponding over the idea, which would effectively end the long-standing ability to tape radio broadcasts for personal, noncommercial, fair-use purposes.
At a recent Senate Commerce Committee hearing on such “broadcast flag” copy-protection, which is already in the works for digital television, RIAA chief don Mitch Bainwol effectively admitted this under pointed questioning (:58, 902K) by Senator John Sununu (R-NH). Sununu uses XM in his example but the context of the larger discussion encompassed both terrestrial and satellite broadcasting.
While the NAB thinks the RIAA’s worries about people cherry-picking digital copies of songs from the air is overblown (64 kbps copies? no thanks), it is amenable to doing something. That something won’t be good.
A recent review of HD Radio in USA Today is mixed: the author tested a $500 HD clock radio (!) and found it “challenging” to receive digital signals at his home in Northern New Jersey. And don’t forget that these extra channels will also cost broadcasters extra money, thanks to the proprietary nature of the IBOC framework itself.