Techno-Exegesis

by Joseph Battaglia  (sephail@2600.com)

The past century has seen many changes in the way radio content is delivered.

Outside of amateur radio, the dits and dahs of Morse code no longer fill the airwaves, FM broadcasting listenership far outweighs the number of those still tuning into the AM (MW) bands, satellite radio is becoming a standard feature in cars, and "podcasting" seems to be the new buzzword amongst the youth.  Most of the changes have been positive, expanding the medium and improving its overall quality, while others threaten the very nature of radio itself.

Shortwave broadcasts have always been an excellent source of international news and perspective, while AM/FM broadcast bands keep us up to date with local events.  Anyone can take a receiver, be it made in the last month or left over from the days of vacuum tubes, and tune into any number of local or international stations packed with news, entertainment, and of course, propaganda.  You choose what you want to hear, and it's all available for free.

But the days of the average person listening to international shortwave broadcasts are quickly passing, causing stations such as the BBC World Service to cease their broadcasts to North America, yet millions are willing to pay for a subscription to satellite radio.

Frequencies now broadcasting analog television signals will become silent in just a few years, and in their place will be private content, owned and controlled by the highest bidder.  New proprietary digital modulation schemes on our broadcast bands threaten to quickly antiquate billions of radios, as well as our freedom to choose what we listen to.

Licensing on new modulation schemes prevents hobbyists from writing their own software to demodulate signals that were previously completely open.  Where is all of this leading?  Into the hands of private enterprise, it seems.  While corporations have always had some control over the content on our airwaves, we now seem to be much more willing to give up the medium than ever before.

In 2001 and 2002, the radio industry saw two new players: XM and Sirius.

These companies, after paying nearly $80 million each to the FCC for frequency allocation in the 2.3 GHz band, became the first two commercial satellite radio providers in the United States.

In just a few years, both companies saw exponential growth, with millions of new customers subscribing to their service in later years.  It's not cheap, either.

The current $12.95 a month plus setup fees is a far cry from the free local broadcast radio we're all used to tuning into on our way home from work.  But where else can you turn to get high quality commercial-free content that follows you around on those cross-country trips?

Well, at least one of those claims is true.

Let's first take a look at some of the technology behind satellite radio.  Both providers live in the microwave S-band: Sirius from 2.320 GHz to 2.3325 GHz and XM from 2.3325 GHz to 2.3450 GHz, with 12.5 MHz of bandwidth each.

According to Chriss Scherer's article "The Final Countdown for Satellite Radio" in Radio World, the total data throughput is 3.28 Mbps.  Analysis with the Shannon-Hartley theorem shows that this is a fairly conservative data rate, allowing for reception of signals that are weaker than the background noise level (as is common with spread spectrum modulation schemes).

This allows for decent reception in noisy or weak signal areas, but is also a very crippling bandwidth limitation for the providers. 3.28 Mbps isn't much, and with modern encoding algorithms requiring at least 64 kbps/channel (or slightly less for talk) to reproduce acceptable sounding music, they're quite limited in the number of channels that they can offer - unless they decrease the bandwidth used by each and, along with it, audio quality.

Sirius promises over 125 channels while XM promises 160.  But how?

At 64 kbps, they should only be able to fit 50 or so channels not counting any additional overhead.  So they obviously decrease the bandwidth consumption even more to cram in that many channels, resulting in audio quality that can no longer compare to what's offered by local FM broadcast stations.  And people are paying for it!

Well, that's fine.  They're not interfering with the conventional broadcast bands, people seem to like it, and it's up to the consumer to subscribe anyway.  So where's the harm?  My concern stems from their success.  We no longer seem to care about the fidelity of what we listen to and while many would claim that this is unfounded and that satellite radio "sounds just fine," consider, for a moment, cellular phones.

Little more than a decade ago, nearly all cellular telephones used the AMPS protocol, which was little more than some digital signaling on top of a purely analog voice channel.  We're talking about real narrow bandwidth analog FM here - high quality stuff.  The voice quality was more limited by the telephone network's codecs than by the wireless modulation scheme and the calls sounded great.  As more and more people began using the cellular networks, more efficient use of the spectrum was necessary to keep up with the call volume.

As the years progressed, new digital standards (TDMA, CDMA, GSM, etc.) were introduced, giving providers a way to limit the bandwidth used by each channel.  More time went by and providers began doing everything they possibly could to increase the capacity of their cell sites, limiting bandwidth as much as possible and leaving us with what we have today: little more than barely intelligible shitty sounding audio.  And you can't argue with that!

While satellite radio is really its own beast, new digital modulation methods are being tested on our conventional broadcast bands as well.  A good example of this is Digital Radio Mondiale (DRM), which is an open standard for broadcasting data in low-bandwidth conditions using In-Band On-Channel (IBOC) technology.

Developed for cheap and easy implementation, DRM can be utilized with preexisting transmitters and receivers, requiring only minor modification.  Although it can be used on any of the AM bands, it is now most commonly found in the shortwave bands.

DRM promises to increase the audio quality of these low-bandwidth AM broadcasts, although a DRM capable receiver (or a modified conventional receiver with software decoding) is required.  It allows for a choice of three MPEG-4 audio codecs, depending on content type: HE-AAC for higher-quality audio and CELP or HVXC for low bitrate voice-only audio.

DRM can operate within the standard frequency allocations (i.e., the 10 kHz channels which are already assigned) in either a hybrid mode (AM+DRM) or DRM-only mode, and allows for multiple digital channels to be present.  It can even be used with a bandwidth of 20 kHz for higher quality audio or channel multiplexing but requires two adjacent channels to be allocated to the station, something many broadcasters do not have available.  Bitrates for single channel (10 kHz) operation range from 8 kbps to 20 kbps, and up to 72 kbps if more bandwidth is used.

DRM can be considered a step forward for the shortwave listening community, which is often plagued with fading as well as man-made and atmospheric noise.  Good DRM decoders can often overcome these issues, resulting in clear, static-free audio.  The meta-data included in DRM broadcasts can also help identify the station and content, a feature that's extremely useful when tuning around the enormous realm of the shortwave bands.  There have already been many radio hobbyists who have posted instructions for modifying popular shortwave receivers for use with software decoders (both open-source and commercial) which utilize a PC and a sound card, allowing for extremely low-cost DRM reception.  As more and more stations begin implementing DRM, my hope is that it will breathe more life into the overall interest in these fascinating bands.

Not all of the new digital methods, however, have these benefits.

Many commercial stations in the FM broadcast band are now touting the phrase "...now broadcasting in high-definition HD Radio!"  Despite being completely inaccurate, not much detail about the technology is being presented by the stations, leaving customers puzzled about what it all actually means.

In fact, HD Radio (HDR) actually stands for Hybrid Digital Radio, another IBOC digital encoding method developed by iBiquity and approved by the FCC for use in 2002.

But unlike DRM, HD Radio is proprietary, thus third-parties wishing to integrate the technology into a receiver must pay licensing fees to the company.

Although it seems that most stations are still in a "testing" phase, hidden dangers exist if the standard catches on.  For now, HD Radio operates on the sub-carriers of FM stations - that is, beyond the bandwidth required for the L+R (monaural) baseband signal (0 - 15 kHz), usually just above the L-R (stereo) signal (23 - 53 kHz) and Radio Data System (RDS) - 57 kHz) sub-carrier.

That's already a lot being jammed into the 200 kHz bandwidth allocation and, according to Carson's Bandwidth Rule, all that "stuff" with a 75 kHz deviation is already exceeding the limit.

HD Radio promises to extend the used bandwidth to almost 400 kHz and can end up causing some serious interference problems, even though most areas have stations spaced at least two channels (400 kHz) apart.  Receiving those distant FM stations stuck in between the locals will quickly become a thing of the past.

All of this, again, does not help audio quality.  FM broadcasting in itself is an extremely high quality means of transmitting audio, the passband being from 50 Hz to 15 kHz, an enormous chunk of the audible frequency range.

In fact, many people cannot even hear much past 15 kHz, let alone below 50 Hz.  High-quality receivers can reproduce extremely good audio in strong signal areas without the need for any type of digital modulation.  As we've seen with other forms of digital modulation, stations wishing to add more "channels" to their broadcasts will decrease the bit rate available to each, leaving us with more crappy audio.

iBiquity's hold on their proprietary technology is also a huge danger to us, the hobbyists.  We can't easily investigate the quality of their encoding or implement our own method of decoding without legal ramifications.  While we have free range to tinker with open standards such as DRM, our hands are tied when it comes to HD Radio.

Worse, if the standard sticks, stations will begin using more and more bandwidth for the digital modulation until the entire broadcast is in proprietary HD Radio format by first removing the stereo separation data, then the entire analog signal, leaving no fallback and billions of antiquated radios.

Clearly, this is the wrong path for us.

The importance of open standards is rarely ever understood in the corporate community, yet hobbyists, those who develop much of the technology in use by the corporate world, have always seen the need for them.

Historically, demand for competition has sorted this out, but in an age when monopolies seem to be sprouting up in all sorts of niche markets, I'm afraid of what might possibly happen if it doesn't.

I've covered only a few of the new concerns in radio, but there's so much more out there: the threat of "rights management" on top of digital radio, Broadband over Power Lines (BPL) interference to our shortwave bands, the sale of portions of our broadcast spectrum to private enterprise, and more.

We've fought similar battles before.  This is yet another that needs our attention.