The first 3DTV World Forum in London confirmed that while some broadcasters are boldly forging ahead, encouraged by recent cinema successes and the ambitions of display manufacturers, there is still much to learn about the illusion of depth presented by stereoscopic imaging. As yet there is little evidence of consumer demand for 3DTV in the home, which will only be created by the availability of programming and suitable displays on which to view it.
As informitv has previously reported, informa estimates that in 2010 there will only be around 100,000 homes that will actually be able to watch 3D broadcasts. Nevertheless, the major manufacturers are planning to sell millions of 3D televisions over the next year, not least to maintain a premium price point for their displays.
Sky is confident that 3D is the next revolution in television after HD. Arguing that innovation drives the entertainment industry, Sky compares the move from 2D to 3D with the transition from analogue to digital, black and white to colour, or silent movies to talkies.
Sky is already showing 3D sport in pubs and clubs and plans to launch a channel for home subscribers in the autumn, featuring over six hours a week of sports, film and entertainment programming.
Brian Lenz, their director of product design and development, believes that success will be judged by repeating the take-up profiles of high-definition, personal video recorders and the transition to digital. Without forecasting any numbers, he observed “You don’t win these races in the first six months.”
Danielle Nagler, Head of HD and now 3D at the BBC, is more cautious about the potential for public service broadcasting, with much still to be done to roll out high-definition. She suggested that the BBC is keen to get it right, rather than being first. The BBC is planning to simulcast its main channel in high-definition later this year but as yet has no plans for regular 3D transmissions.
That is not to say there is no interest in 3D. The current movie StreetDance 3D is backed by BBC Films. There are various other trials in train and there are hopes to capture some of the London 2012 Olympics in 3D for transmission in cinemas and on public screens, as well as for the archive.
NHK Media Technology, the commercial arm of the public service broadcaster in Japan, has developed extensive experience in stereoscopic production since 1990, including covering the opening and closing ceremonies at the last six Olympic games. There are no immediate plans for 3D broadcasting.
Meanwhile, the World Cup football in South Africa will be covered extensively in 3D. Sony is using the event to showcase its technology. Viewers in the United Kingdom will have to watch it in selected cinemas, as the BBC and ITV share the rights and so far seem unwilling to work with Sky to allow 3D coverage.
There is still much discussion over signal formats and standards. It seems evident that side-by-side presentation of two images within a single frame has pragmatic benefits for distribution using existing infrastructure. It seems equally evident that this will ultimately be replaced by a full-frame solution, however that may be coded.
In Korea they are envisaging a system that is backward compatible with existing digital terrestrial transmissions, using an MPEG-2 stream for the left eye augmented by an MPEG-4 stream for the right eye — an approach that appears fraught with potential problems.
The movie industry appears to have it all figured out with a 3D specification for Blu-ray discs and the connection from the player to the display is covered by version 1.4 of the HDMI specification.
It is all the more surprising that broadcasters have yet to agree a way to transmit the signalling of 3D formats. It seems that vertically integrated pay-television platforms will initially use some form of private signalling. The various broadcasting bodies have only just started looking at the problem.
The consumer electronics industry is keen to promote 3D to drive sales of new televisions, worth $100 billion a year, led by Korean manufacturers like Samsung, which has the largest share of sales with 23% of the market.
Samsung showed its latest models, which are capable of accepting stereoscopic signals in various formats, but also synthesising a sense of depth from conventional channels. Based on similar techniques as those used to create higher frame rate displays, the results of 2D-to3D conversion were surprisingly effective and seem likely to improve as processing develops but will remain more limited than material that is specially shot stereoscopically.
The problem that remains is the dark glasses. The latest active shutter designs are lightweight but require recharging. Passive polarised glasses are cheaper and it seems likely that they will be available in fashionable designs and with prescription lenses. Nevertheless, peering through special spectacles for a special event may be one thing, but it seems unlikely that viewers will feel comfortable wearing them for general viewing.
3D is likely to be driven by occasional special events, major movie releases and video games. As far as general television viewing goes, the consumer electronics companies need to take account of the typical viewing environment. The television screen rarely has the undivided attention of the viewer, who may well be simultaneously engaged in other activities.
Dr Neil Dodgson, an expert in 3D at the University of Cambridge in England gave a balanced account of the history and prospects for stereoscopy, ignoring the issues of technical standards and covering the bigger picture.
Many directors are still tempted to use 3D as a gimmick but like all new innovations in moving pictures, from sound, to colour, to widescreen, it may eventually become a standard part of the storytelling toolset.
It is generally understood that some people cannot actually perceive stereopsis correctly. Figures quoted range wildly from 2% to 20%, but there is apparently little scientific research to back this up. There has also been almost no work on the effects on young children, or the short and long-term effects of viewing stereoscopic images.
Dr Sean McCarthy of Motorola, chair of the 3DTV working group of the MPEGIF, has a background in vision science and bioengineering. He urged delegates to go back to their respective organisations and governments and call for further research.
The ultimate hope is for autostereoscopic displays that do not require glasses. Such displays exist but the viewers need to remain in particular viewing zones to see the effect correctly. Again this presents problems in the home environment. For this reason such displays are more likely to be seen in handheld applications, where the viewing position is more restricted and easily adjusted.
For engineers, stereoscopic vision seems like another technical problem to solve in order to reach ever greater fidelity of reproduction.
However, the perception of depth is only partly dependent upon binocular vision. Stereoscopy can produce a very strong illusion of depth that when carefully controlled can be very convincing. Otherwise the conflicting cues presented can be confusing or even uncomfortable for the viewer.
Furthermore, the stereoscopic representation is no more real than any other picture. It may provide an impression of depth but it cannot be confused with reality. It remains an intriguing phenomenon that a very thin display can produce images that appear to extend in front of and behind the screen plane, but only a fool would mistake the figures represented for real people.
Finally, stereoscopic display is not really 3D at all. The view does not change as you move with respect to the image and it is not possible to look around objects. Needless to say, technologists are working on that, but it could be some time off before we see it in the living room.
Dr William Cooper of informitv was one of the session chairs for the first 3DTV World Forum in London.