Designs on the Games
Tuesday 7th October 2008
Savoy Place, London
Organised by PhoneAbility and the IET
Watching the Games on our mobiles
by Jim Slater
Please note: There is a more detailed technical paper available for this presentation.
I had always considered this the holy of holies prior to today. I guess I have given four or five talks here and I never thought the time would come when I would have to stand up here and say to Stephen, "I am an engineer, sorry about that!".
So forgive me. Let's see how we get on. Hardly anyone these days watches TV on mobile phones but that will in fact rapidly change in the coming years. I have this dream, I am sure it will be a reality, that when I visit the Games site in East London, my mobile phone will not only enable me to stay in touch with home and the office throughout the day; it will allow me to take photos of the events as they happen. But it will also enable me to watch colour television pictures. What of?
Well, the news, other events happening in realtime, and more importantly, events happening at the other venues whilst I am in the London stadium.
In between the athletics events, as I wait, I will be able to watch the sailing in Weymouth, beach volleyball in Horseguards Parade - that struck me as most unlikely! - football at Wembley and gymnastics in the Millennium dome.
But for this vision of mobile television to turn it reality, a great deal of further work needs to be done. It's only the remarkable track record of our existing commercially competing mobile telephone companies that can lead us to have any confidence at all that everything will be in place to enable a viable mobile television system in just four years' time.
Now, it's not that there's any shortage of commercial ideas or plans. Quite honestly there are more potential technical methods of providing mobile television than anyone could reasonably want. If technologies and standards are considered to be a good thing, well, we've certainly got plenty of them. But this very plenitude may mean things aren't as good for us as they might be.
Before we get involved in competing broadcast standards and their multiplicity of abbreviations, let's look at the basics of the various ways in which television pictures and sound can be carried to your mobile phone. Remember, we are looking at a device like this or perhaps slightly larger.
What does mobile TV mean? What it really means is providing video and audio content that's either streamed or broadcast direct to your mobile phone. It's important that we understand the key differences between the two terminologies and technologies, streaming and broadcasting.
Most of us will understand about broadcasting. It's a process that sends data from one source to many receivers. We are familiar with the concept of the huge transmitter masts there at Crystal Palace - a cartoon version of it, of course! - which radiates TV signals that can be received by millions of people in Greater London and all around.
But streaming over a mobile telephone service is something very different. It's one to one data transfer. Many of us are familiar with using our home computer to call up a video clip or on a remote server via the Internet.
The use of the widely known Internet protocol IP makes it simple and very straightforward in general for us to get a continuous stream of data. That stream of data represents the picture which is coming from the distant server on which it's stored. It's up to our computer then to re-assemble the bits to generate a picture and the sound that goes with it.
Now, if we have a high bandwidth Internet connection, the data flows smoothly, the pictures and sound are excellent. But if the bandwidth is restricted, either because we are using an old fashioned telephone line and modem system rather than a broadband connection, or because the Internet can't cope with the vast amounts of data required to transmit at times of peak usage, delays in receiving that data mean that the pictures break up and the sound stutters, completely ruining the experience.
One of the first lessons we are going to have to learn is that for any mobile TV service we are going to have to consider potential bandwidth limitations.
Now, early mobile phone services were designed to carry digitised audio telephone calls. In the most efficient way possible, therefore the systems needed only to carry modest quantities of data. But as the need for carrying other forms of information which began with e-mails, picture files and now video files, as that becomes apparent, it was realised that new mobile phone systems would be needed.
So enhancements to the current second generation mobile phone system allow data to be transferred at up to 384 kbits per second under favourable conditions.
But there is currently a move to go even further forward to the so-called third generation or 3G systems - a generic term covers a range of future wireless network technologies.
These will allow for much higher data rates, 2 mbits a second will be possible and typical on a regular basis. But there are also slightly longer term plans for high speed access systems able to deliver speeds up to 14.4 mbits.
Since the network of broadcast transmitters suitable for mobile TV covering the country doesn't exist, we've got 1152 broadcasting transmitters but not ordinary mobile TV signals, but most of the population is already well covered by second generation mobile phone services. There are masts everywhere you look.
3G is currently available in all the highly populated areas and along the main transport arteries. It is not surprising that apart from a few broadcast experiments, the only existing mobile TV services that there are and that you could choose are operated by mobile phone companies and they are streaming the data over their 3G networks.
Now, although data rates are limited, careful compression of the pictures by the operators means that acceptable pictures can be obtained on hand-held displays and the data rates are well within the capabilities of the current 3G systems.
Now, the main 3G operators offer mobile TV, or perhaps it's more accurate to say they offer video services using data streaming, and they do work. But there is one big snag.
The same data streams that are used to carry mobile phone calls are also those that are being used for Internet access and for TV pictures, but the TV pictures require vastly more data than the phone calls.
Now, since it's common for mobile phone companies to charge for the amount of data that's downloaded and since there is currently little competition in the field, watching streamed TV on your mobile at the moment can be horrendously expensive.
Cost will be an important factor in determining whether mobile phone television will be successful.
Again, cost isn't the only problem. With mobile TV being transmitted over data circuits, you have to note that the streamed forms of mobile TV are actually unicasting - they are not technically broadcasting - because what's happening is that each access to a programme requires a data link to be set up from the programme provider's server to the computer in your hand-held phone.
Every new viewer then requires the phone company to provide more server space and it uses more of the network bandwidth.
So in technical terms, we can say that the required network capacity is proportional to the number of users. The more users there are, the more network capacity you use, the more expensive things are going to be for the operator.
So it can be both technically and commercially difficult to provide high quality mobile phone pictures to large numbers of subscribers simultaneously over a 3G telephone network.
Using the service that we have now which is called UMTS, the universal mobile telecommunications system. All right, so just as you thought you were starting to understand the clear distinction between streaming and broadcasting, let me tell you that UMTS can in fact offer a broadcast mode MBMS.
I have deliberately taken out a lot of the technical data in the written version of this paper - it's on the Tiresias website - to fit in to the time space we have today. But I will try and give some of the essential detail.
We have seen that the required network capacity in UMTS is proportional to the number of users. Now, if you are MBMS, the broadcast service, you are effectively operating a one-way point-to-multipoint service, like a broadcaster, but there are still significant problems for the operator, since if he introduces MBMS to this service, that takes up a large amount of the network capacity. If he isn't careful, he won't have enough capacity left to carry the phone calls which is his main business where the money is coming in regularly.
The slide shows that UMTS operators were offered groups of paired frequencies, so uplinks and downlinks in various bands.
But in addition to that, the point of this bit of the talk, they were also given an unpaired frequency band from 2010 to 2025 MHz which is unpaired, no up or downlink there, just a transmission band. That allows the telephone operators to actually provide broadcast data.
One operator has shown that up to 14 TV data channels can be transmitted inexpensively in that 5 MHz of unpaired spectrum.
So although it's possible for mobile phone operators to sell services based on data streaming, there seems general acceptance that the real future of mobile TV lies with the broadcasting model rather than unicasting data one by one. So broadcasting means that effectively unlimited number of users can pick up the same signal and they can all watch digital content and they are not competing for limited bandwidth resources.
Now, various ways of broadcasting multimedia or mobile TV via broadcast networks exist, and they have been the subject of trials and rollout in lots of different parts of the world. They include a whole range of the abbreviations and acronyms, DVBH, ISDBT, media flow, and DXB to be available soon. Don't worry about it all!
Digital multimedia broadcasting, DMB, is based on the existing DAB, digital audio broadcasting, system. Many of you I am sure will have DAB radios.
This system is already in operation in Korea and Germany and China is using a variant too. In the UK, BT carried out trials using the DAB spectrum to distribute TV and radio channels to Virgin mobile phone subscribers who were able to watch the news, music video and radio channels.
The DAB system has been modified to DAB-IP, which can share the multiplex capacity, share the frequency bands if you like, used for the existing digital audio services to carry these extra transmissions.
So any operator can choose to use existing spare capacity on DAB networks to start offering mobile TV now, without waiting for new spectrum to be made available. We will see later that that really becomes important.
It's also important that about 90% of the UK population is covered now with DAB digital audio broadcast radio signals, so again, in theory, 90% could use this coverage to carry mobile TV as well.
Another broadcast to mobile - it's interesting, the trial worked brilliantly. All the papers said how wonderful the BT trial was, all the papers said how great it was, but somehow the business case isn't there yet and there are no plans to extend the service.
Another one, FLO, forward link only, another broadcast to mobile technology, designed specifically by the Qualcomm company for the distribution of multimedia content to millions of wireless subscribers simultaneously.
FLO is already in use in the USA and it's actually using UHF channel 55 in our TV bands for this and it offers multiple live TV channels with full access to Internet and interactive services.
It's interesting to see how it's done. The TV services are received on hand-held mobiles but the controlled and interactive services take place via the normal 3G mobile communication system using completely different bands. So, in the UHF band you have the TV programmes being transmitted, to control those services and to select and to provide your interactivity we use the 3G signals.
They have shown that typically 32 mobile TV channels can be transmitted within one standard UHF television channel.
Something of interest here. Although that's been going on in the States, Qualcomm bid for and was awarded a chunk of UK spectrum in the 1452 to 1492 MHz range, so that's above our current UHF TV transmissions.
It paid eight and a bit million pounds for the use of the band. It hasn't yet said what it's going to do with them.
So rumours suggest that Qualcomm has planned to import its mobile TV service - it calls it media FLO - into the UK and to use this band width. It is known that last year Qualcomm tested media FLO services with BSkyB, further stoking the rumours as to what may or may not happen.
But actually having spectrum available now, they have got it, bought it, could be a huge advantage for media FLO over other potential mobile TV services because the other services may not have sufficient spectrum available until the digital switchover takes place or is completed in 2012.
Which brings us on to the other main competitor, DVB-H. Digital video broadcasting group that set the standards for digital broadcasting throughout the world or most of the world and the H stands for "hand-held".
DVBH is technically based on the existing DVB-T terrestrial system, what you know as Freeview, and the DVB-T transmitter structure, the same transmitters can be used to carry DVBH at the same time. P modifications of the transmitters will be needed and more spectrum and frequency space will be required but up to 20 hand-held TV programmes can be carried in a standard channel. So a channel that holds a standard digital multiplex now could carry 20 TV channels.
It's standardised with an ETSI standard for DVB-H agreed worldwide. Handsets have been developed which provide normal mobile phone services.
In Oxford in 2006, the mobile phone company O 2 and broadcast transmission company Arqiva ran a six-month trial. It had 375 customers using 16-channel mobile trial.
Eight DVBH transmitters, eight little mobile phone masts, using UHF channel 31, and they only did that because they were given a special licence from Ofcom that allowed them to do test and development work.
They put these transmitters so that the signals were concentrated within the Oxford ring-road. It's interesting that the system worked within the Oxford ring road. You could get reliable TV reception on a hand-held phone like the one on the screen, at street level, outdoors, indoors surprisingly enough, and even inside moving vehicles. You could within the ring road actually drive around and somebody in the back of a car hopefully could be watching the TV programmes.
There was an electronic service guide provided which gave the information needed to the receiver to know what programmes were available and what to do with them, and also it helped the viewer to be able to select a, from a list of programmes.
The services were really liked by those taking parts. You can read the results of the research and everybody was positive. Even things that we had been worried about, like would the batteries last very long because mobile phones you have always got the inherent problem of how do you get enough power. Even they worked well.
So it's interesting, everybody liked it. As yet there are two problems. There is no spectrum available to carry the broadcast regularly and much work still remains to be done before a commercial business plan for such services can be developed.
In fact as one who looks at these things it's difficult to see how enough spectrum could be found for DVB signals to be transmitted countrywide until the full switchover is complete.
The Crystal Palace transmitter is not due to switch finally until April 2012, and when that happens it will free up spectrum space that could be used for DVB-H. Could this be done in time to watch hand-held pictures of the Games?
Let's look at the standards. With a plethera of the potential systems available it's difficult to predict what will happen but the European Commission or the Commissioner for information, media and society as she is, has stressed the importance self-Europe having a single standard for mobile TV. When you take the mobile phone, you go on the Channel tunnel and you get to the other side and it's automatically changed over. You can use the same mobile phone throughout Europe.
The same would obviously be very sensible if you could do it for TV. So Europe says this is a good thing. The European Commissioner has therefore recently thrown its weight behind DVB-H, the system based on the European digital video broadcasting system.
But of course that's all fine, except that we've got this problem with frequencies, and the proponents of the other systems aren't going to give up without a fight.
They are arguing that a good reason for adopting their systems is that they can work NOW, whereas frequencies with the full DVBH coverage aren't likely to be available until 2012.
But that's the technical stuff. We have overcome many engineering challenges, many ways of going on, many ways of sorting out the technical problems.
The technical standards, they exist. We can cope with them, deal with them. Manufacturers could start to make equipment as soon as they liked.
But coming to the purpose of today, there's still a great deal of work to be done, not only for a sound business case to be made for mobile TV - although it's attractive to many people undoubtedly, it would need to be pretty low cost before it takes off widely.
There are loads of unanswered questions. What's the best type of programme for watching on a tiny display. Would we want to watch normal programmes or will specially developed short snippets be more popular with commuters who might want to watch a half-hour programme while standing on the train.
But more than that, bringing the focus sharply to today's topics, what needs to be done to make the mobile TV programmes and the data truly accessible to all viewers? Accessibility is one of those issues we are all familiar with, we are all in favour of, say yes it is a good thing.
But I am an engineer. We need to apply a dose of common sense. It won't be possible for a visually impaired person to see too much detail on a two-inch screen but many normally sighted people are likely to find this difficult too. If great care isn't taken to make the displays as clear as possible, with sufficient brightness to watch pictures in reasonably bright outdoor conditions, then nobody will be happy.
The technology will come to our aid. New light emitting display technologies may prove better than the liquid crystal displays and already they are being used in some mobile phones.
The real problem is that, unless we use seminars like today's to highlight the potential problems of accessibility, nobody in the business will even give the slightest thought to the difficulties that users might encounter.
So I want us now to take an in-advance look at how mobile TV needs to be accessible and I want to highlight to designers and manufacturers of equipment and systems just what it is that we all need.
Screens. We have already mentioned they need to be of sufficient resolution to make the picture as clear as possible with easily adjustable brightness to cope with different viewing conditions indoors and out.
Many digital cameras, these days the ones with big LCD screens on the back, have a zoom facility which allows you to examine a small portion of the image in detail.
I think the same facility should be there on every mobile phone in order to allow you to magnify the portion of the picture you want to see and enable everybody to get a better visual image of the particular parts of the picture they want to see. It doesn't apply just to those with impaired vision.
Subtitles. It has taken around 30 years to get UK terrestrial TV to a situation where subtitles are available on most TV programmes. We need to act now in order to get mobile TV providers to ensure that ways are found of incorporating subtitles as an integral part of mobile television services right from the start.
Now, again, I am a practical guy. It may not be easy. You certainly can't use existing subtitling systems since the characters would be too small when displayed on a two-inch screen for anybody to read.
Maybe one answer would be to provide perhaps a half-inch deep ribbon across the bottom of the TV picture which could carry a stream of large Tiresias characters showing the subtitles. Digital broadcasters are already carrying such ribbons on their news pages so the technology is there. We just need to get mobile TV developers to realise that the service is needed and to incorporate it at the beginning.
Signing for the hard of hearing. Nobody knows more than I do how difficult this can be, and on a two-inch screen this really could be difficult since a tiny screen would mean an inset picture of a signer might be too small to be useful at all.
The presentation of the signer on the display screen needs to be of sufficient size and resolution to show all the movements, expressions and the gestures. I can't see immediately how I could achieve that on a screen of this size.
Maybe one idea would be to split the screen diagonally with half the screen reserved for the signer and the other half for the picture.
Already on digital terrestrial TV you may have noticed sometimes the visual image is reduced by perhaps 25% in order to provide a larger area to display the interpreter. But research needs to be done on these much smaller mobile TV screens to see just what can be usefully achieved.
Audio description. Audio description is a spoken description of the visual aspects of a television scene, inserted into the gaps between the dialogue. It helps visually impaired people to more fully understand what's happening on the screen.
It's been introduced on a wide range of TV programmes in recent years and proved extremely popular and extremely useful, and not only with those who have difficulty in seeing.
Now audio description needs an extra audio channel along with the standard mobile TV audio, but an audio channel like that requires only a very modest amount of data compared to the data being used to provide the TV services. Once again, we need mobile TV developers to appreciate the need and to build AD into the system right from the start.
If the levels of the AD could be adjusted by the viewer and if the service could be easily switched on and off, then so much the better. But it's important that we let the developers know our requirements as early as possible, so that ways can be found of incorporating them in the new mobile services.
Since mobile phone TVs always incorporate the possibility of two-way communication, it is a mobile phone after all, it might also be desirable for such future services to be able to receive subtitling, signing or audio description, not via the normal broadcast channel but via an Internet protocol data service, quite independently of the broadcast programmes but synchronised with them. It's one method that could work and has been demonstrated.
Another access service, clean audio. Many people with hearing impairments have problems in understanding the speech when there is a background sound on a television programme, and other normally hearing people can find background sound distracting. A possibility for a future access service would be to provide a clean audio channel which provides just the speech without any background music or other sounds. This, like audio description, will require another audio channel and a modest amount of data capacity reserved for its transmission.
It's only modest and the cost of providing it would be extremely small compared with the benefits to large numbers of people who would be helped by such a facility. It would be good, again, if mobile TV could accept this and incorporate this facility right from the start.
Hearing aid compatibility. Many hearing aid users find that they can hear phone conversations much better if they use a phone with a built-in inductive coupler and their hearing aid set to the T telecoil setting.
I hadn't come across a mobile phone with a T setting, many portable phones do incorporate inductive couplers so they don't have to be expensive. It might be good to ensure that users mobile phones could buy one with a modern equivalent of an inductive coupler. You have a Bluetooth adapter on your hearing aid which gives a much better audio signal. That could be important, because listening to good quality audio becomes even more important if you are listening to a half-hour TV programme than just listening to a very short phone call.
Speech input. No, I haven't gone off into the realms of cloud cuckoo land. Let's consider making speech recognition a primary requirement for mobile TV phones. There are already mobile units which respond to dial commands and respond to spoken commands. So let's make it a user requirement and insist that mobile TVs can be controlled by a user's voice. It would be a particular help for those with motor disabilities that make it difficult to select and press the correct button. Using speech as a method of selecting channels or interactive TV options would also help a much wider range of people who might not consider themselves "disabled" - elderly people, those with arthritis.
Speech output. Again, this isn't an impractical suggestion. Already the lady in my car's sat nav has no difficulty in nagging me for mile after mile with a range of different instructions, so the technology obviously exists for a mobile TV to read out its electronic programme guide or its on-screen information in audio form. This would be invaluable to those who find reading small print on a screen difficult, and at least one set-top box provides this facility already. It would be useful if most mobile phone TVs could incorporate it from the start.
Once you have this new generation of mobile phone that allows you to watch television, you will be able to, if those planning the future facilities now will get on with it - and that's in bold in my script, if those now planning the service will get on with it - you will be able to use the same mobile phone for a whole range of new facilities. Push to talk allows you to use a mobile phone like a walky-talky connecting you to a group of friends, as individuals or as a group. It makes it simpler for those with disabilities to keep in touch with those with whom they came to the Games without having to make lots of different phone calls. Finding your mates isn't easy if you can't see.
The same mobile phone could be used for ticketing. Before you go, your mobile phone could be displayed a barricade and it acts as your ticket when you get there.
With a bit of forward planning the same phone could be used to bring a whole range of text and video based services to those sitting in the stadium. Lists of medal winners and league tables could be made available, and video and audio clips of the various performances, allowing spectators to know what's happening everywhere and to feel they really are taking part in the games, chatting to friends and really enjoying the atmosphere of really being there.
So let's start now. Our history with television in this country and in most countries is that TV access services have had to be added to existing TV transmissions. That's often meant that there were technical difficulties and problems involved in modifying existing equipment. In contrast, we are now at a unique time in the development of new mobile TV services. Everything is still flexible since no irrevocable major business or technical decisions have yet been made and we certainly don't have any services that have been widely rolled out.
So before these services start in earnest, it's important that the telecoms regulatory bodies ensure that an agreed list of user requirements is drawn up in consultation with groups representing disabled people. Those companies intending to provide mobile TV services can then be persuaded, compelled if necessary, to incorporate the additional features needed to make mobile TV services truly accessible to all.
Time is short, as competing manufacturers and service providers thrash out their business plans for new mobile TV services. So the message today is that we really must seize the opportunity and start now. Work with the regulator to draw up a comprehensive list of user requirements to ensure that mobile TV, whether it comes, can be used by everyone.
And then, when we sit in that Olympic Park in 2012, we will know that, as a direct result of this seminar in 2008, we will all be making use of one of the world's best and most accessible mobile TV services. (applause)
JOHN GILL: Thank you very much. A question in the back row.
FROM THE FLOOR: Hello, my name I Philip I was involved in the, {inaudible} of one of these, and two things 2 years ago I would probably have agreed with you there is a huge problem with TV, but seeing all the trials around the world in Italy and Korea, and Germany, in the UK basically failing. People do like mobile TV but they are not willing to pay for it. We have seen now in other markets that handset manufacturers are coming out with mobile phones which support DVB T for example, which in other European countries you can re-see now with your mobile phone. So, I would say we'll probably end up with the normal DB T on mobile phones and not have extra bandwidth for dedicated mobile phones services. Another point you were mentioning subtitles, there are different ways of delivering subtitles where you can have as many languages as you want.
JIM SLATER: I think the real problem is the business case nobody has yet made a business case for mobile phones that works.
Please note: There is a more detailed technical paper available for this presentation.
JOHN GILL: Thank you very much Jim we now move on to Julian Jones who is representing British standards institution, he’s a member of committee on bio electrics, smart cars the accessibility of those he's been on SEN committees, he’s an expert on standardisation in this area of accessibility in technology.
| Previous |
Last updated: 20.11.2009 © Copyright reserved Website design: Digital Accessibility Team