Background

The conventional audience for television broadcasters is one which wishes to watch programs. Most watching is plural in the sense that more than one person is watching the screen at any one time. Viewing is a leisure pursuit which seldom requires any physical or mental activity from the viewer other than to select the channel which might provide the best viewing. It is almost as if television were the opium of the masses (to mis-quote Karl Marx). It is the carrier of entertainment.

Of course there is a vast market for training and informative uses of television. Pedagogic applications embrace schooling class rooms of pupils through to providing individuals with a customised presentation. A more interactive multi-media form of presentation can be provided but it will need to consider both the nature of the audience it seeks to serve and of the technology through which it has economic access.

The television audience

The conventional view of the television audience is that it wishes to watch television. One of the challenges which has arisen in the development of broadcasting strategies in digital television broadcasting has been to arrive at an understanding of just what degree of interaction can be promoted and accepted by the majority of the audience. It is accepted that there will be a fraction of the audience which would interact - if only we could understand just what this interaction activity involves. Perhaps, in the past, audiences have done so through treating the television as a background noise and by talking over it. Nowadays they are likely to be playing or working with a PC either programming, working or perhaps engaging in Internet activity engaged in what has been called Infotainment. Such a person is not really a television viewer and no amount of advertising revenue or reduction of a licence fee will make him one.

Television based Internet interactivity trials have been conducted and it is reasonable to suppose that some useful data has been acquired from these trials. It would appear that none have continued into a maturity. Many observers have questioned the oversold "wisdom" that Internet will be the saviour of television.

There is an intrinsic difference between the use of the television for interactivity - particularly Internet oriented activity and broadcast television programming. For the viewer the use of interactivity oriented around the Internet is very much a single person experience with the display terminal. By contrast the television service is very much seen by many people at the same time. The differences exist also in the proximity of the viewer to the screen … the PC user is typically seated within 2 times the screen height whilst the television viewer is typically 7 times that height distant from the screen. Apart from the sense of involvement it is important to recognise that the plain visibility of the text on the screen is much affected by this.

But that’s to see interactivity with the television screen as an image of the PC oriented Internet service. There are other levels of interactivity. In the UK, Teletext Ltd makes a very attractive business out of the approximately 192kbit/s offered in the VBI lines of the ITV and Channel 4 television services. Apart from the very simple but effective television program guide the pages which viewers most commonly visit include the travel, weather and stock market pages. The principal revenue comes from the carriage of advertising.

From this we may learn that a consistent reliable presentation of information will be accessed by television viewers. We need also to recognise that humans are only sufficiently awake for a small amount of the week such that they can watch television or play with PCs. In the UK the available viewing time is of the order of 25 hours per week - it is around 42 hours in USA and 16 hours in Sweden. Further, it does not seem to be growing greatly. Indeed, in the UK conventional cinema attendance is increasing and this is despite the ready availability of multi-channel television and the increasing numbers of home PCs which are fitted with modems.

Some technical issues

There are key differences in the technical performance of the display devices which used in PC and television viewing. Many consumer television displays may do well to resolve 2MHz horizontally whilst the PC user’s screen is commonly able to resolve 20MHz. Add to this the effect of interlace and the effect of big screen flicker for the television viewer and it is clear that the displays are not likely to be equal. This greatly affects the amount of information which can be reliably placed on screen at any one time.

Presentation on a television screen has been the preserve of the broadcaster. Television screens display the same picture in the same way on all television screens. Partly this reflects the very much simpler technology present in television and, in turn this reflects an intrinsically conservative approach to change which viewers have towards the technology they buy. Television receivers are very much bought on the premise that they shall have a long and useful lifetime. The same anticipation is not so of a PC even though the baseline cost to the consumer is similar.

Most television displays do not even deliver the full picture to the viewer. This is because they are typically set up to "overscan" by some 10%. This can be contrasted with the PC screen where every addressable pixel will be visible.

By the same token the broadcaster and the viewer have come to expect that televisions will be uniform in their presentation and in the way in which the viewer can be expected to interact with it. A good many of the television viewing public have no desire to become technologically literate … it’s simply not where their life interest lies.

Developments in interactive presentation

There are two approaches which are commonly proposed … either the television approaches the PC in terms of its display precision and its attitude towards user interaction or the television medium develops a valid paradigm for its own audience to enjoy. This is not to deny any cross-over between the audience types but it strongly suggests that there is a specific technology requirement for the television viewer.

For this technology we are looking to fix the presentation irrespective of the software within the receiver and irrespective of the manufacturer of the receiver. The last thing the broadcaster wants is a host of desperate help-line calls in the late evening as a new application crashes receivers across a nation. The cost of call-out in the morning could be a serious drain on profitability. The broadcaster is looking to be able to provide controls to retain the viewer and this is important as the viewer will presumably need to go to sleep at some stage in the day.

For the television program broadcaster a favoured technology solution is based on MHEG. The reasons are well rehearsed and include the fact that it is an open international standard, that it handles the presentation of text and graphic objects in all receivers ion the same way, that it can handle television originated media streams precisely, that it has "access" to external peripherals such as keyboards, modems and memory and, finally, that when a complex procedural application must be delivered to the viewer use can be made of Java.

In fact the combination of a DVB compliant receiver incorporating MPEG2 decoding, the common interface as a means of delivering conditional access and further functionality and the use of MHEG as an API provides the basis of a receiver which uses entirely open standards and which has no requirement to embed proprietary technology. As a consequence such a receiver should find wide acceptance in an open market where the consumers confidence in the long term stability of the product will help to establish digital television.

Relating MHEG functionality to the WWW paradigm

The problem comes when budding service providers confuse the function of the television with that of the PC. Since the intrinsic differences between the broadcast television service and the PC provided Internet access should be now apparent, the issue which remains is what is it that needs to adapt in order that the information of one delivery medium can be handled using the presentational paradigm of a quite different one. Note that the provision of a return channel is not an issue.

It is this requirement which defines the problem adequately and which points the development direction for such a service for some time to come. As an indication of some of the drivers consider that for both practical reasons to do with plain visual acuity and the usual television screen viewing distances and for the reason that the broadcast regulator may validly insist that text should be readable that the smallest acceptable character size is such that a capital V occupies 16 horizontal lines. This will not allow the use of a text size such as this paper is being authored with (Arial at 10 point can provide around 100 characters across an SVGA screen) but limits the authoring to closer to 60 characters.

We should consider too the man machine interface with which television is most commonly blessed. The humble infra-red remote control need not be thought of as the underfunded last_thought_of tool which is packed with the television. Rather, it represents a challenge to the system designer to provide a powerful means of giving the viewer control over a wide range of functions. In this respect the details of the man machine interface are either admirably sparse or annoyingly inadequate. The tragedy would appear to be that there is no single universal standard. However it may be fair to suggest that the use of the computer keyboard is not the correct substitute and, indeed some of the more adventurous television man machine interfaces have started to show what could be done whist remaining true to the requirement for simplicity and for maintaining the relationship to the television rather than to a PC.

Much of this suggests that the conventional WWW HTML documentation needs to be mapped in some reliable manner into the television’s MHEG environment. The time honoured approach is that the broadcaster or service provider bears the cost of any complexity or any receiver legacy issue which arises as a consequence of developing technology. Some sources of information may be mapped relatively painlessly using automatic means whilst others will require re-authoring in order that the display of information is appropriate.

This rather suggests that whilst the full Internet compatible sites may remain accessible to those receivers and viewers who can handle the presentational demands suited to a PC it may be valuable to provide a parallel site which is accessible to television based display systems. These would include both conventional digital television receivers and dedicated so-called Internet STBs which are designed to be operated in conjunction with a television presentation.

Some concluding observations

A commercial free to air broadcaster depends on delivering an audience to an advertiser in order to assure an income. In an environment in which regulation is applied evenly across all broadcasters program quality should not suffer. Indeed it can be easily shown that even when only a part of the broadcast spectrum is placed under full regulatory control that high quality standards can be maintained.

For an audience which can only be awake for a fraction of each day it is important that the complex issue of delivering a relatively passive viewing audience is achieved whilst being able to provide a level of non-obtrusive interactivity for that part of the audience which desires it.

Since, from this viewpoint it does not seem sensible to divert the whole audience into a full confrontation with the Internet, it seems relevant to provide for that degree of interactivity which capitalises on the presentational quality and man machine interface which is achieved with digital television systems rather than to risk the viewer’s loyalty and faith by promoting, as a presentational platform, something which is forever changing and which, mostly as a consequence, is both technologically demanding to understand and unreliable.

For the television viewer the interactive delivery system needs something like MHEG if not exactly MHEG and it is therefore clear that when information from other sources is to be presented to the viewer it must be mapped carefully into this domain.

It seems that rather than adapt the television so that it is based on the same volatile software and hardware technology demands as the PC it may be sensible to determine a derivative markup language and related presentational and authoring tools which would result in the ability of a PC to display with the same intrinsic consistency and reliability the information which is made available to the television viewer. Naturally these would include the access to whatever return channel technology was present