IT TOOK just ten years for the digital CD to bury the analogue vinyl LP. Even
at its higher price, the digital disc has swept the board, thanks to its
glorious convenience and sound quality. A similar revolution is about to hit
television broadcasting. It promises hundreds of channels, sparkling picture
quality and a host of interactive services. But whether digital TV will succeed
in storming the analogue barricades is far from certain.
Broadcasters have still to convince the public that the benefits of digital
TV will be worth the inevitable extra cost. Their success will depend greatly on
whether electronics companies can drive down the cost of watching digital
pictures. And this, in turn, relies on whether the correct technical decisions
are made today: wrong choices could saddle viewers with expensive equipment that
would be outdated within a few years.
The US, Europe and Japan are all developing digital systems, but a curious
set of political coincidences has left one country at the forefront of the
digital TV revolution: Britain (see 鈥淪low motion鈥�, next page). The nation鈥檚
broadcasters and viewers have much to gain if the move to digital succeeds, but
everything to lose if things go wrong.
Advertisement
The push towards digital TV in Britain is coming not from a single direction,
but from three. First, the satellite broadcaster BSkyB, controlled by Rupert
Murdoch, has leased 14 transmitters on the new series of Astra satellites, due
for launch in August 1997. Shortly afterwards, Murdoch wants to begin
broadcasting several hundred digital TV channels. One of the obstacles facing
this venture is how to create a large audience for those digital channels
quickly. To watch these programmes, viewers will need new set-top decoders. They
will also have to point their dishes towards a different satellite, and install
new electronics for the dishes.
This problem for BSkyB is a commercial opportunity for the second driving
force behind digital TV, Britain鈥檚 cable companies. They could simply pick up
digital programmes, convert a selection to analogue form, and pump them to
subscribers鈥� homes. This way, the viewers would need no new equipment to watch
the programmes. Eventually, the cable companies could start giving their
subscribers digital set-top boxes and distribute the channels in a digital
form.
And finally there is the BBC. In May, straight after the government awarded
the corporation a new 10-year charter, its director-general John Birt announced
his 鈥渄igital dividend鈥濃�攁 plan that includes launching digital TV
broadcasting in late 1997. He predicts that the next ten years will see the BBC
sending digital programmes into people鈥檚 homes by whatever means is
appropriate鈥攕atellite, terrestrial transmitters, cable or telephone wires.
鈥淓ither we make the digital transition or we just let our beards grow and do
nothing. That鈥檚 not a sustainable option,鈥� says Birt.
If more channels are the goal, the analogue option is not sustainable because
it wastes so much of the broadcast spectrum. Each analogue station needs around
8 megahertz of frequency space. And transmitters using the same frequency must
be separated by several hundred kilometres, because even weak interference
creates herringbone patterns on the screen. So the frequencies allocated to
analogue terrestrial broadcasting support only a few TV channels in each
country. And while a satellite can deliver more channels, its transmitters often
cover several countries, so each country receives only a handful of channels in
its native tongue.
Switching to digital multiplies the number of TV channels that can be
broadcast in the same frequency space by a factor of about 10. This improvement
is due partly to new data compression methods, which mean that fewer bits are
needed to make up a picture, and partly to new transmission techniques that
allow more bits per second to reach the viewer鈥檚 aerial. How many digital
channels will actually be on offer will depend on the system that is chosen and
the wishes of the operator. The bitstream could be used to carry a few
programmes鈥擧ollywood blockbusters, for example鈥攚ith high-quality
pictures or a lot of different programmes鈥攏ews or sports programmes,
say鈥攚ith the same picture quality as today, or worse.
But if picture quality is not assured, other features are. Digital TV
promises to create a variety of interactive services. Viewers will send digital
commands back to the broadcaster by telephone line or cable link, which will
enable them to take part in competitions, vote in debates, gamble on games or
watch a centrally stored video at convenient times. Astra says that within two
years its new satellites will be able to respond to signals beamed up by viewers
from their home dishes.
Britain鈥檚 position at the forefront of the digital TV revolution owes much to
the swift pace of development in Europe. The impetus for this came, in turn,
through lessons learnt from the failure of an analogue TV project. In the 1980s,
manufacturers and broadcasters came together in a pan-European research project,
called Eureka 95, to develop high-definition TV technology. HD-MAC, as it was
called, was a hybrid system in which the pictures transmitted in analogue form
were accompanied by digital 鈥渉elper鈥� signals. Specially designed TV sets used
the helper signals to create wider, clearer images than we see today. MAC was
heavily patented to protect the European electronics industry from Far Eastern
competition. The European Commission tried to force European companies to use
MAC, but they rebelled. Electronics companies such as Philips of the
Netherlands, Thomson of France and Nokia of Finland had hedged their bets by
secretly developing digital TV systems in case MAC failed. In the end, HD-MAC
went nowhere.
In September 1993, hard on the heels of HD-MAC鈥檚 demise, Europe鈥檚
manufacturers and broadcasters set up the Digital Video Broadcasting group.
Their objective was to set standards for a system driven by commercial need, not
political whimsy. The group agrees standards for digital broadcasting, which it
passes to ETSI, the European Telecommunications Standards Institute, for
rubber-stamping. With remarkable speed, the DVB group hammered out three
standards: for cable, satellite and terrestrial broadcasting. All are based on
MPEG-2, an international digital compression standard. The prime differences
between the three DVB standards are in the way the compressed signal is packaged
for transmission.
The satellite and cable systems build on techniques used for sending
electronic mail over ordinary telephone lines. The latest modems send digital
bits by switching the amplitude of an analogue wave between several different
levels, each of which represents a 鈥渙ne鈥� or a 鈥渮ero鈥�. In addition, they
duplicate the analogue wave and transmit it out of phase with the original. And
this wave also carries digital data. The result is an extremely rapid stream of
digital pulses. But this system works only if the waves can travel freely,
without interference or distortion. This is fine for TV cables, which act like
electronic pipes, shielding the signals from distortion, and satellite signals,
which arrive at a dish aerial direct from space in a narrow unobstructed
beam.
Terrestrial TV signals, however, are far more prone to interference. On their
journey to the viewer鈥檚 aerial, they reflect off buildings and hills, so several
versions of the same signal arrive at the receiver with varying delays. On an
analogue TV set, this generates 鈥渟now鈥� and ghost images. On a digital system the
echoes so confuse the decoder that it stops working altogether.
The solution is to use a system called coded orthogonal frequency division
multiplex (COFDM), which was developed for digital radio. Instead of
transmitting a single, wide radio channel, the broadcaster puts out several
thousand narrow channels, packed tightly together in the electromagnetic
spectrum like the teeth of a comb. The digital code is split into a similar
number of streams, and each stream is sent on one channel. In this way, each
channel carries fewer bits every second, so there are relatively long gaps
between bits. Any unwanted reflections arrive in these gaps and the receiver
ignores them.
This elegant solution has, however, left Britain with a dilemma. Last year,
the BBC reckoned that in order to make digital terrestrial signals robust enough
to give good pictures in cities and valleys, it would need 8000 channels. But
electronics companies said that 2000 channels would be enough. They argued that
the chips needed for 鈥�8k鈥� reception would be as complex as a Pentium processors,
would cost too much, and would not be ready until 2000. In the light of this,
NTL, the company that runs the network of transmitters which serves Britain鈥檚
independent broadcasters, declared that 8k transmission was 鈥渄oomed to
蹿补颈濒耻谤别鈥�.
Gambling on quality
The DVB group has now decided to leave the choice of system up to individual
countries. Some, including Spain, may hold off the launch of digital services
until 8k chips become cost-effective. But to avoid delays to digital TV, Britain
has opted for a 2k system. This decision is a gamble, because 2k receivers will
never be able to pick up 8k transmissions. So once the BBC starts broadcasting
with a 2k system, the die will be cast. If the system is not up to the job and
8k technology is needed, all receivers will become useless. So being the
pathfinder could backfire on Britain.
When the hype and euphoria is stripped away, viewers will only benefit if
they are willing to buy the equipment needed to receive and decode the new
signals. Without consumer cooperation, the whole house of cards comes tumbling
down. One of the big selling points of digital TV has been that it will allow
viewers to enjoy movies in wide-screen format鈥攚ith a 16:9 aspect ratio
rather than the 4:3 of standard screens鈥攁nd this also raises problems.
Anyone wanting to pick up digital transmissions will need either a new TV set
or a set-top box that converts digital code into analogue signals, which will
then be fed to a conventional set. But viewers watching conventional sets will
not see wide-screen pictures in all their glory, so broadcasters are having to
decide how to transmit the digital programmes for viewing both on wide-screen
sets and conventional sets via a set-top box. Either 16:9 viewers will lose the
tops of actors鈥� heads, or 4:3 viewers will see 鈥渓etterbox鈥� pictures with wide
black bars at the top and bottom of the screen.
The BBC has now settled on a compromise that looks likely to become an
international standard. The target is to transmit all signals in 14:9 aspect
ratio. With this system, viewers with ordinary 4:3 sets will lose something from
the sides of their pictures, and see narrow black borders at the top and bottom.
Wide-screen sets will expand the signal and then trim the top and bottom of the
image to fit the 16:9 screen.
But new sets or set-top boxes are not all the public must buy. While existing
terrestrial TV aerials will, in most cases, pick up digital transmissions,
satellite viewers will need either a new dish or modifications to their existing
dish. Yet with the launch of digital TV barely more than a year away, BSkyB is
still selling dishes that will not be able to receive digital signals without
costly modification. And the company has not yet announced how it intends to
introduce digital receivers into viewers鈥� homes.
The DVB Group鈥檚 target pricing is 拢485 for a set-top decoder. This is
far too high for a mass market now educated to expect a complete satellite
system for under 拢100. In the past, Murdoch has subsidised analogue
satellite dishes, virtually giving them away in return for viewers鈥� long-term
commitment to pay for his subscription channels. If BSkyB wants to prime the
digital pump, it will surely have to work this way again.
At a recent conference in London, representatives from BSkyB and Pace,
Britain鈥檚 leading supplier of satellite equipment, gave clear clues to the
likely launch scenario for the digital revolution. Jim Beveridge of Pace and
Robin Crossley, one of BSkyB鈥檚 technical advisers, said that because of the high
price of even 2k COFDM circuitry, making a satellite set-top box that can also
pick up terrestrial channels will cost an extra 拢50. By comparison, making
a terrestrial receiver capable of also picking up satellite signals will add
only 拢5 to the cost. So it makes no sense, they said, to sell a set-top
box for terrestrial reception that cannot receive satellite channels too.
A dual-purpose box could also have other benefits, says Beveridge. The
encryption process can be made to work with a combination of national codes
beamed down from satellites and local updates sent via terrestrial transmitters.
Pace is planning a set-top box to receive both satellite and terrestrial
channels, which Beveridge estimates will cost about 拢300 to make by early
1998. That translates to a price in the shops of 拢500 to 拢600,
unless somebody subsidises production. The price, he says, should fall to
between 拢150 and 拢200 by the year 2000. The BBC has now spoken out
in support of the dual-purpose box. In mid-June Birt called on the government to
compel manufacturers to sell only equipment that will receive both terrestrial
and satellite broadcasts.
Britain鈥檚 terrestrial broadcasters reach 90 per cent of the population
through 50 main analogue transmitters. The rest are served by another 1000 relay
transmitters. The microwave links that pass TV signals around this network will
not have the capacity to cope with the multitude of promised digital channels,
which would mean that some remote areas will miss out on the digital revolution.
But this needn鈥檛 be so, Crossley argues. The BBC could rely on satellite
transmissions to relay its digital signals, either to its smaller transmitters
or direct to viewers鈥� homes.
All this leads logically to a launch scenario which has all viewers using the
same, dual-purpose box鈥攕ome for terrestrial reception and others, in more
remote areas, pulling in the same programmes direct from satellite. The same
scenario also leads to the hitherto unthinkable, an alliance between Murdoch at
BSkyB and Birt at the BBC. Without such an alliance, the digital dividend may
never be paid.
* * *
Slow motion
FOR some viewers in the US, digital TV has already arrived. Two years ago,
the aerospace giant Hughes launched DirecTV, but so far it reaches only 1.6
million viewers. The US government is now close to setting a standard for
digital terrestrial TV, but commercialisation has been bogged down because the
national plan was based on high-definition TV. It turns out that few viewers
want high-definition pictures, and even fewer are willing to pay for the costly
new sets needed to do them justice. The government鈥檚 new standard will allow a
flexible system for transmitting conventional-quality TV in digital form.
Japan is also on the slow track to digital broadcasting. The country was a
pioneer of an analogue HDTV system called Hi Vision, which the government
adopted in November 1991. It dares not repay viewers who have supported Hi
Vision by making their receivers obsolete.
On mainland Europe, there is a mess of alliances and competition. Players
include Canal+ in France, Rupert Murdoch鈥檚 BSkyB and German media giants
Bertlesmann and Kirch. The situation varies by country, and changes almost
daily.