With existing LCD technology, it is difficult to make flawless large screens. Colour screens are made up of thousands of liquid crystals which either block light or allow it to pass through, depending on whether they have a voltage across them. Each crystal has a white light behind it and a red, green or blue filter in front. And the voltage across each crystal is governed by an electronic switch.

Unfortunately, a small proportion of these switches develop faults during manufacture and show up as bright dots on the screen. A single fault is enough for a screen to be rejected. With large displays the chance of finding a fault is so great that they cease to be cost-effective.

The new TV, tentatively called Plasmatron, overcomes this problem by dispensing with conventional switches. Instead, each horizontal row of crystals is backed by a long plasma discharge channel filled with low-pressure gas. When a high voltage passes between two electrodes running along the channel, it triggers the formation of a cloud of electrons.

The plasma channel acts as a common electrode for all the liquid crystals in that row. By applying a positive voltage to a second electrode, each crystal can be switched “on” – that is, light will be allowed to pass through.

An image is created by switching on each channel in turn. As each one is switched on, positive charge is fed simultaneously to whichever crystals in that row need to be turned on.

Each plasma channel corresponds to one horizontal scanning line of a conventional TV. If a channel fails, then a whole scanning line will be faulty. But with only one channel for each row, the chance of producing a flawless large screen is much greater than with conventional LCD technology.

While the LCD panels are only 3.7 millimetres deep, the prototype TV sets are 20 centimetres deep. Sony expects to reduce this to 10 centimetres by the time the sets go on sale.

The technique is called Imaginography, and it applies the same logic that the human brain uses to analyse images. People normally see single images only in the plane of the object on which they are concentrating. Everything in front of or behind that plane is seen as a ghosted double image. For example, when looking at a distant object such as a picture on the wall, a finger held directly in the line of sight and at arm’s length appears as two fainter images.

Inventor Rahim Nader has defined the precise formula needed to work out how big and how far apart the double images should be to replicate this effect. So far, Nader has produced 3D photographs, videos, films and animated computer graphics.

For videos shot using Imaginography, two cameras are placed a small distance apart. If they are “looking at”, say, a bowl of fruit, the signals from the cameras are mixed so that the two images of the bowl are precisely superimposed in the final image. Objects behind or in front of the fruit bowl appear as double images, with overlaps showing. This setup – a “normal” central image surrounded by double images – fools the eye into seeing a 3D picture. The point of focus can be manipulated by moving the two images from side to side so that different objects appear as the central single image.

The two video signals could simply be synchronised and mixed for viewing on a normal TV set except for one hitch – overlaid colours interfere with each other. For example, if blue and green light are mixed on the TV screen they make yellow instead of the turquoise that people instinctively expect to see.

The best way of preventing the problem is called “space averaging” – combining the two images in a fine chessboard pattern using a computer. Because the signals do not overlap, there will be no colour interference. Nader says that space averaging also improves the look of pictures compared with simple overprinting (See Diagram).