ÐÓ°ÉÔ­´´

Television viewers throughout Europe will finally get their chance later
this year to judge high-definition television for themselves. Thomson of
France and Philips of the Netherlands, two of the main proponents of HDTV
in Europe, are planning to broadcast the Olympic Games in Barcelona to more
than 600 sites, ranging from shopping malls to leisure parks. Most previous
HDTV screenings in Europe have been to the electronics trade. The change
in strategy is designed to whet our appetite for the launch of a public
HDTV service in 1995, and to counter any criticism of the European Commission’s
plan to spend a large cache of European taxpayers’ money to safeguard the
interests of the European broadcasting industry and, in particular, the
industry’s investment in HDTV.

HDTV lets viewers sit closer to a wider screen, and so become more involved
in the pictures. This makes TV more like the cinema experience, where the
screen is so large and so wide that the audience is seeing some of the action
with peripheral vision. The 625-line TV in Europe displays 50 images a second,
each made up from 312.5 lines, vertically staggered so that they interlace
to create the illusion of 25 frames a second, each with 625 lines. Anyone
who sits close to a conventional TV will be acutely aware of the coarse
scanning raster of 625 horizontal lines (or 525 lines at 60 images a second
in Japan and the US). With the number of scanning lines doubled to 1250
(or 1125 in Japan) and the screen expanded sideways, HDTV lets the viewer
sit close enough to be seeing some of the picture with peripheral vision
without being disturbed by the line raster.

European manufacturers have been collaborating on the development of
HDTV for five years, with Philips and Thomson thought to be investing at
least a billion dollars each. Their aim has been to upgrade Europe’s MAC
(multiplexed analogue components) satellite transmission system to an HDTV
service, known as HD-MAC, and persuade the European Commission to make this
the broadcasting standard for Europe. As an intermediate step, the industry
has developed a medium-definition widescreen service, known as D2-MAC. With
MAC and its variants as standard in Europe, European manufacturers, which
hold key patents to the system, would be able to control the imports of
transmission and receiving equipment. The trouble is that millions of Europeans
already own receivers tuned to the PAL (phase alternation line) transmission
system, which cannot receive MAC signals. To get MAC off the ground, the
Commission has plans to spend more than £600 million subsidising
the simultaneous transmission of programmes in both MAC and PAL formats.
Though an expensive compromise, it may be the only way of steering TV viewers
in Europe from PAL to MAC and from there to HD-MAC.

Meanwhile in the US, the Federal Communications Commission looks set
to protect the interests of American manufacturers by standardising another
type of HDTV system. This system, which is scheduled to come into service
in 1995, will be incompatible with Europe’s and with that already operating
in Japan.

But there remains one great flaw in the grand plan to sell HDTV as a
domestic product. The technology of display screens lags far behind that
of electronic transmission. Manufacturers cannot produce TV screens cheap
enough for the home that do justice to the high-quality signal of an HDTV
broadcast, and there is still no sign of a breakthrough. Genuine HDTV clarity
is seen only on the professional monitor screens used by broadcasters in
studios or for public demonstrations.

So backward is screen technology that it is highly unlikely that the
TV sets that go on sale in 1995 will display high-definition pictures to
the standard of professional sets. In Japan, where research on HDTV began
20 years ago and sets went on sale last October for around $30 000 apiece,
the screens display pictures that fail miserably to do full justice to the
broadcast signals of the Hi-Vision system, the Japanese HDTV service. Even
medium-definition widescreen sets, which are now on sale around the world,
fail to do justice to the signals they receive.

Another drawback is the size of the equipment. This consists either
of a large box housing the receiver and screen, or a display screen on the
wall plus a projector that is fed TV signals from a receiver. Manufacturers
admit that people may find such bulky hardware unattractive in their living
rooms.

Nearly all widescreen TVs being demonstrated or in the shops rely on
the cathode-ray tube. They may use one large CRT built into a cabinet so
that it faces the viewer. Alternatively, they may use three small CRTs,
one each for the red, blue and green components of an image. A lens focuses
the three outputs into a full-colour picture, which is then projected onto
a reflective screen or through a translucent one. Two projection systems,
one sitting on top of the other, can be used to increase the brightness
of the picture on a large screen, but such a combination is far too expensive
for the average consumer.

Though the CRT remains by far the most cost-effective way of displaying
a large, clear, moving colour image, it is beginning to show its age. The
technology is almost 100 years old, and although it has been refined almost
to perfection it seems cumbersome by the standards of modern electronics.
To prevent implosion under atmospheric pressure, the evacuated glass tube,
which looks like a bottle with a long neck and a flattened bottom, is thick
and heavy and has a steel band around its rim. The inside of the bottom
of the tube, which will become the screen, is coated with a mosaic of red,
green and blue phosphor dots. Three electron beams, produced in the neck
of the tube and modulated by the broadcast signal, cause the phosphors to
emit light. Each beam is channelled onto the dots of one colour by perforations
in a thin sheet of metal, or ‘shadow mask’, which lies over the screen.
For accurate colour, the pattern of dots must match the pattern of holes
in the mask. This is achieved by using the mask as a template for positioning
the dots.

The mask is made by a photoetching technique. It is coated with a light-sensitive
material, known as a photoresist, exposed to ultraviolet light through a
master template, and then bathed in hydrochloric acid that eats away at
the metal to form the pattern of tiny holes. Ultraviolet light is then shone
through the finished mask onto the inside of the bottom, or face plate,
of the tube where it exposes another layer of photoresist deposited on the
glass. More etching leaves a mosaic of holes through which phosphor slurry
penetrates to the glass surface. The process is carried out three times,
once for each colour. Throughout this process, the mask and face plate travel
together as a matching pair. A tube plant sounds more like a zoo than an
electronics factory, with noises like elephants trumpeting as robots lift
and spin the hot, wet and heavy glass components.

BIG AND HEAVY

In a working CRT, magnetic coils focus the electron beams into fine
spots that match the size of the mask’s perforations, which guide them to
the phosphor dots on the screen. Each beam sweeps the screen in a horizontal
raster, and careful adjustment of the coils is essential to prevent it becoming
defocused at the edges of the screen. Such adjustments become even more
critical as the width of the screen, and thus the sweep of the electron
beam, increases. Widescreen receivers have a width-to-depth ratio of 16:9,
compared with the relatively narrow 4:3 ratio of traditional receivers.

All this, and the extra glass needed for widescreen tubes, indicate
why the first batches of 16:9 CRTs are costing several times as much as
4:3 tubes for the same picture height. Clearly, too, the mechanical construction
of a CRT puts a practical limit on its size. Above a diagonal measurement
of about 100 centimetres, the tube, and any TV set containing it, is just
too heavy for domestic sale and use. As a result, the standard size of a
widescreen TV, measured diagonally across the face of the screen, is 90
centimetres (36 inches), although manufacturers are also selling 80-centimetre
sets.

The dilemma for tube designers is balancing picture resolution against
picture brightness. The size of the phosphor dots and shadow mask holes
limits the detail that the screen can resolve, just as the coarseness of
the dots making up a newspaper photograph determines the detail a reader
can see. The smaller the pitch, the higher the resolution. But reducing
the pitch of the phosphor dots and the size of the holes in the shadow mask
cuts significantly the amount of energy that can be focused on each spot.
This reduces the light that the spots produce, so the picture appears dimmer.
This is why demonstrations of fine-pitch HDTV tubes, which are also too
expensive for the average consumer, are given in dark rooms.

When juggling compromises, designers must first consider the theoretical
and practical resolutions available for the transmitted signal. Vertical
resolution is defined by the number of horizontal scanning lines making
up the picture; for the HD-MAC system, Europe’s proposed HDTV service, there
are 1152 active lines. The rest of the 1250 lines define the black borders
at the top and bottom of the picture. Horizontal resolution is defined by
the number of pixels, or picture elements, along each line. For HD-MAC,
the theoretical maximum is 1920 pixels per line but the average consumer
screen can display no more than about 700 pixels per line. Even cameras
and recorders manage to pick up only 960 pixels per line, though this limit
is being extended all the time.

To resolve the detail of an HDTV signal, the tube must have a phosphor
pitch of 0.4 millimetres or less. HDTV tubes for professional users achieve
this pitch. But these tubes have been handmade and the pictures they display
are far too dim to satisfy domestic viewers, who are accustomed to watching
in a bright living room. Conventional 4:3 tubes have a phosphor pitch of
between 0.7 and 0.8 millimetres.

The medium-definition sets now on sale in Europe use 16:9 tubes with
a phosphor pitch of around 0.7 millimetres. So do all the widescreen sets
on sale in Japan, whether they offer high or medium definition. The Philips
widescreen TV displays 100 images a second, each made up from 312.5 lines,
vertically staggered so that they interlace to create the illusion of 50
frames a second, each with 625 lines. The company’s next move will be to
sell screens that display 100 images a second, each made up from the full
625 lines. The Ferguson Space System, which is produced by Thomson, does
not alter the frequency at which images are displayed, but repeats each
line to create the illusion of 25 frames a second, each with 1250 lines.
Nokia of Finland, another collaborator in Europe’s development of HDTV,
argues that this artificial doubling of line structure, from 625 to 1250
lines, degrades the quality of the picture unless expensive extra circuitry
is installed in the set. Instead, the company sells widescreen TVs based
on technology similar to that offered by Philips.

Though none of these widescreen TV sets display images of HDTV quality,
they do provide big, bright pictures at a price (around £3000) that
some consumers can afford. Handmade professional tubes with a phosphor pitch
of 0.4 millimetres cost ten times as much as tubes with a 0.7-millimetre
pitch; this is on top of the premium on the price of a 4:3 tube that is
charged for 16:9 widescreen tubes of the same height.

In Japan, Toshiba says it plans to reduce the phosphor pitch of the
widescreen tubes it makes to 0.6 millimetres. But experiments by Thomson
at its factory in Italy indicate that a reduction from 0.7 to 0.6 millimetres
to be counterproductive. The slight increase in resolution was far outweighed
by the reduction in brightness of the picture. Max Artigalas, head of the
company’s HDTV development team, says: ‘The target for tube designers is
to double the intensity of the electron beam. This would give a 50 per cent
increase in brightness while giving a 50 per cent improvement in resolution
from the use of a tighter phosphor pitch.’ But this approach brings its
own problems. Doubling the intensity of the beam shortens tube life and
also makes the metal shadow mask so hot that it may change shape, and so
distort the colours of the picture.

Such drawbacks with CRTs have led the makers of TV sets to consider
alternative screens. Liquid crystal display panels are the favourite. Sharp
in Japan expects to have invested $700 million in research and development
on LCDs in the three years to the end of 1992, while Sanyo will have got
through $560 million and Panasonic $350 million. Toshiba has a joint venture
with IBM in Japan called Display Technology Incorporated. DTI began producing
colour LCDs in May last year. Philips is looking for a partner to exploit
the research done at its LCD pilot production line in Eindhoven, and has
been talking to Thomson.

HEAVY ON THE POCKET

The first pocket TV sets, which were launched in Japan in the early
1980s, had small monochrome LCD screens. They were a commerical failure.
So was the next generation, with tiny colour screens. Japanese manufacturers
are now waiting to see how the current crop sells. They have provided better
backlighting and increased the size of the screen from 7.5 to 10 centimetres,
both of which have helped improve the quality of the pictures. But at £550
apiece, the sets cost as much as a traditional TV. Sharp makes a 22-centimetre
LCD screen that sells for around £2000, even without a TV tuner to
feed it signals. These larger screens are being used for specialist applications,
such as on bulkhead walls in long-haul aircraft, to provide in-flight entertainment
for seats that have no clear view of the main cabin screen.

An LCD panel is a glass sandwich. The liquid crystal, a clear fluid,
is trapped between two glass plates in a gap about 5 micrometres thick,
one-tenth the width of a human hair. Bonded to the surface of each plate
is a fine grid of metal electrodes, so thin that they are transparent. A
network of very thin metal tapes connects the electrodes to microchips round
the edge of the panel. The microchips convert the TV signal into pulses
of electricity, which are transmitted through the tapes to the electrodes.
Depending on the voltage applied to the electrodes on each side of them,
the liquid crystal molecules in the vicinity realign themselves so that
they either block light or transmit it. In this way, each electrode represents
a cell of light or darkness. Thin strips of red, green and blue filters
overlaying the glass add colour to produce an image from a mosaic of tiny
cells of light.

To connect or ‘address’ each individual cell with its own pair of wires
would require an impractically large number of individual connections. The
latest LCD panels use a network of transistors, one at the corner of each
electrode, which switch the voltages off and on. Amorphous silicon is used
to bond the transistors directly to the glass plates. The system encompasses
what is known as thin-film transistor technology, because the matrix of
transistor material is deposited as a thin film on the glass.

Making thin-film transistor LCDs involves depositing up to ten different
layers of conducting and insulating material on the glass plates, one after
the other, like making a giant microchip. Unlike microchip manufacturing,
however, one tiny blemish can ruin the entire device. ‘If one speck of dust
falls on the silicon wafer carrying a hundred chips, then it just kills
one chip out of a hundred,’ says Thierry Robin, managing director of Ferguson
UK and formerly with Thomson’s LCD division. ‘But if you increase the size
of the device, so that it takes up the whole wafer, then one speck of dust
kills the device. It only has to spoil one pixel and the whole screen is
²õ±è´Ç¾±±ô±ð»å.’

As with the phoshor dots on the screen of a CRT, the smaller the individual
liquid crystal cells, the more there can be, and the greater the resolution
of the picture. Although a small screen can give the illusion of clear pictures
from a relatively small number of cells, a larger screen needs more cells.
This, too, increases the risk of failure and reduces production yield. For
existing 4:3 TV systems, a 10-centimetre LCD panel will look acceptable
if it has 100 000 pixels, each a triad of a red, a green and a blue cell.
But a larger panel must have at least 300 000 and preferably 400 000 pixels
to give clear pictures. For HDTV, the number of pixels needed to do the
signal justice rises to at least 1.5 million, and no LCD panel has yet been
designed to provide this density of detail.

Out of all this gloom, there is one ray of hope. Designers are starting
to use small LCDs in video projectors. Light is pumped through the LCD,
as it is through the film in a 35-millimetre slide projector. Ideally three
LCD ‘slides’ are used, each around 2.5 centimetres wide, one filtered red,
one green and one blue. The three coloured images are aligned on screen
to form a full-colour image. The technology’s pacesetters, Japan’s electronics
industry, see this approach as more likely to be viable for widescreen
TVs than large LCDs, until at least the beginning of the next century.