The end of analog television is at hand. Pundits have predicted the death of analog before, but such forecasts were couched in caveats. Now governments are setting firm dates and planning for life after analog, when vast amounts of bandwidth will become available for new uses and the broadcast TV scene will change.

Around the world, governments have begun the analog shutdown, and it will accelerate rapidly during the next five years. In Germany, Berlin killed off analog in 2003, Munich did it this year [see photo, "Getting Ready"], and the rest of the nation is scheduled to follow suit by 2010. In the United States, Congress likely will legislate January 2009 as the shutoff date. The end-of-analog date in France is 2010. In Japan, it's 2011. The United Kingdom, which turned off analog broadcasts in one Welsh community this year as an experiment, is slated to phase out analog completely by the end of 2012.

After analog television is phased out, digital over-the-air transmission will be the only game in town for those receiving free TV signals through antennas.

If television comes to you by cable or satellite, you won't notice a thing. Satellite television is already digital, and so is much of cable. But eventually you will reap diverse rewards that you might not even connect to changes in TV broadcasting: better cellphone reception, opportunities to download video to your cellphone [see illustration, "Playing Soon"], and mobile broadband Internet. And, in the United States, you might see a modest dip in the federal budget deficit when the government sells off 108 megahertz of the old analog broadcast spectrum for as much as US $50 billion, by some estimates.

If you do rely on broadcast television, you'll notice the changes even sooner. The first one might be a little painful: you'll need a new TV set or, at minimum, a new tuner costing at least $100 [see sidebar, "Countdown to the End"]. With a new high-definition set, you'll see a big improvement in the TV picture. Most digital programming is broadcast in HD, which brings the crisp, detailed images so prized by sports fans (who are determined never to lose sight of the ball or puck) and feared by news anchors (who know that viewers can see every bit of makeup they plaster on). Along with those sharp pictures comes digital surround sound—if you add the speakers.

In some countries, mainly in Europe, broadcasters have no plans for terrestrial broadcast of high-definition television. Nevertheless, digital broadcasting should bring other potential benefits. Some broadcasters may send out multiple standard-definition channels, perhaps "narrowcasting" shows to niche audiences or providing supplementary material, such as an interactive experience, with regular shows.

In any massive technology change, particularly one with so much money at stake, there are winners and losers. I'll get to those. But first, to understand why this enormously valuable part of the spectrum will soon be up for grabs in an unprecedented high-tech rush, we have to go back to the late 1990s.

The United States was the first country to broadcast digital TV, in 1998, and its mechanism was basically followed by other countries in their own systems. So the U.S. experience is illustrative.

In the late 1990s, the Federal Communications Commission (FCC) loaned each TV broadcaster a second channel in the existing broadcast bands, 54 through 806 MHz. Interspersed among the broadcast channels are some spectrum gaps that minimize interference between them. To further minimize interference, the FCC skipped certain channels in a geographic region; for example, if channel 4 is assigned in one metropolitan area, the nearest channel 3 broadcaster is in a different metropolitan area. The skipped channels are known as taboo channels.

Each channel occupies 6 MHz, and that hasn't changed. Rather, because digital transmission is less interfering and also less subject to interference, and because digital channels operate at lower power levels than their analog counterparts, the FCC assigned second channels into analog taboo channels. The FCC deemed the modest increase in the overall level of interference acceptable during the transition.

At the time of the bandwidth loan, Congress set year-end 2006 as the date when analog service would officially cease and the extra channels would be "returned." At that point, the digital channels, with their low interference characteristics, could be repacked into less bandwidth—a swath between 54 and 698 MHz. The move would free 108 MHz of spectrum—the upper end of the UHF band, or TV channels 52 to 69—for other uses. To put the potential value of that 108 MHz in perspective, note that the entire AM radio spectrum is less than 1.2 MHz. All local area networks using IEEE 802.11b and 802.11g, the most common forms of Wi-Fi, occupy just 83.5 MHz. Congress looked forward to a lucrative spectrum auction to help balance the federal budget.

The 2006 date, however, came with a caveat: on a market-by-market basis, at least 85 percent of households would have to own at least one television that could receive digital signals.

It has been clear for months that the 85 percent criterion will not be met next year, so the U.S. plan will be delayed [see sidebar, "Countdown to the End"]. But for how long? Now, many of the affected players—consumer electronics and computer manufacturers, along with communications and other companies interested in using the recaptured spectrum—do not want a "soft date." Instead, they have been agitating for a hard one, with no further chance of delay.

Although Congress has yet to pass legislation to set such a date, both the House of Representatives and the Senate seemed in late summer 2005 to be converging on 1 January 2009.

Shortly before any hard date, the band rush will begin. Congress, eager for the money, is pushing the FCC to start the auctions as soon as possible. The Congressional Budget Office is advising that the auctions be delayed until after other, unrelated spectrum auctions are completed. Spreading them out will prevent a sudden glut of bandwidth, thus optimizing returns. Auction winners would require a year or two to gather the money they'd need to invest in developing their newly acquired spectrum segments. So for them, if bandwidth is to become available at the end of 2008, auctions in late 2006 or early 2007 would be ideal.