Abstract *

Introduction *

Background *

Evaluation *

Conclusion *

Bibliography *

This essay examines whether television broadcasting will migrate to the Internet or continue to be broadcast via dedicated mechanisms such as cable and satellite. This is important as this would dramatically change the nature of the medium from a one-to-many relationship to a one-to-one/many-to-many relationship. It first examines the current methods used to broadcast television and how the Internet can technically be used to replace these. It examines the literature available on criteria which will affect its take-up – image/audio quality, content, choice, client hardware, interactivity and availability. It concludes that the migration is inevitable in the long term (ten to twenty years) but in the shorter term digital transmission over cable and satellite will continue to grow and that there is a need for more trials and research.

The most important and revolutionary new form of communication since the introduction of the telephone and television is the Internet. Over the next few years it will affect many industries especially media as the Internet can be used as the delivery mechanism. Television is the most popular form of media – the average UK viewer watches 25 hours of television a week (Mulcahy 1998). It is currently making the transition from digital to analogue and while a lot of attention is being paid to this there is relatively little concern as to how this will be affected by the Internet. This is even though the Internet has far more potential to change television than digitalisation – it has the power to fundamentally change the nature of the medium from a one-to-many to a many-to-many/one-to-one relationship . Most research was carried out on the Internet as it is comparatively up-to-date and researchers active in this area tend to be heavy users of the Internet so they place a lot of information online.

Television is generally broadcast using three methods - terrestrial broadcasts using VHF (Very High Frequency) or UHF (Ultra High Frequency) radio waves, satellite broadcasts using microwaves and cable broadcasts using coaxial cable. A new method called point to multipoint radio is also being used where microwaves are transmitted from a transmitter to receivers within its line of site. During 1998 it was launched in Atlanta, Los Angeles, New Orleans and Montreal (Riga 1998). In addition there are new methods currently in development. Powerline transmission transmits data to the viewer through the power grid and xDSL (Digital Subscriber Line) technology allows television pictures to be digitally transmitted down telephone lines. A form called ASDL (Asynchronous Digital Subscriber Line) was used in an interactive TV trial in Ipswich and Colchester by British Telecom three years ago , but it is not being used commercially as equipment that must be installed in the telephone exchange is currently too expensive. (OFTEL 1998)

Cable, satellite and terrestrial are currently undergoing a transition from analogue to digital. Terrestrially digital transmission allows either six slightly higher than normal quality channels or one HDTV (high definition television) channel to be placed in the space of one analogue channel. (Jaroker 1998)

Table 1 below shows the popularity of terrestrial, satellite and cable in the UK and US:

Terrestrial television is much more popular in the UK than the US. This is partly because it is much harder to receive good quality broadcast signals in the US and because the cable market is much more mature.

In the UK terrestrial digital transmission started in November 1998 from a company called onDigital and contains about 30 channels (onDigital 1998). Satellite digital transmission started in October 1998 from Sky Television and cable transmission will start during 1999.

In the US the FCC (Federal Communications Commission) has announced that all broadcasters are required to broadcast at least one channel digitally by 2006. They are also encouraging the adoption of HDTV for selected channels. The first broadcasters started transmitting digitally in late 1998.

The Internet is a network of networks spanning the globe. NUA Internet Surveys (Nua Internet 1998) estimated that 151 million users were connected to the Internet as of December 1998. It began in the early 1970s as a network intended for the military to be able to communicate in case of a nuclear attack. It could achieve this using a technology called packet-switching. When two computers communicated, instead of forming a direct dedicated connection (circuit-switching - as with the current telephone system), they would send the data in little packets which would be routed through the network whatever way possible to arrive at there destination. If any intermediate connections were destroyed the data would be routed around.

In the 1980s the National Science Foundation expanded their network (NSFNET), which was based on this technology, to allow campuses and research centres in the United States to use their supercomputers. The network started to be used for inter-university communication, especially e-mail.

The Internet’s killer applications were e-mail and the world-wide web. The web was developed at CERN (the European Laboratory for Particle Physics) at the beginning of the 1990s. This enabled people to publish pages including pictures which could be viewed very simply by anyone around the world. It comprised HTTP (HyperText Transfer Protocol) and HTML (HyperText Mark-up Language). HTTP specified a protocol (a 'dialog' between the user's computer and the computer containing the pages) that allowed the transmission of the pages, and HTML specified a file format for the pages that allowed them to be viewed on any type of computer using software called a browser.

Video streaming is a method of sending video through the Internet. It is known as streaming as the video is displayed as the data arrives at the viewer's computer, rather than being received in its entirety and viewed afterwards. The main problems in video streaming relate to the Internet bandwidth (speed of transmission) and QoS (Quality of Service - reliability). Digitised video takes a lot of space - 1 minute of television quality video takes approximately 1,900 megabytes (million characters) of space per minute, while an average home connection to the Internet can transmit approximately 300 kilobytes (thousand characters) per minute and the Internet is already quite congested. Streamed video must also arrive at the destination reliably - if some packets of data do not arrive or arrive too late the player will 'stall' - it will not have any pictures to display on the screen. However, there are methods to eliminate both of these problems, which I shall examine later.

Whatever medium is used for transmitting digital video, the size must be reduced. This can be achieved by several methods. The size of the picture or the number of frames per second can be reduced. However this isn't a satisfactory solution if the aim of the video is to replace television - ideally the size of the picture should be increased. The most powerful solution is compression, of which there are two types. Lossless compression removes patterns of data and replaces them with symbols which can be substituted back on decompression, for example "195555555555874" could be represented by "19[10×5]874". The decompressed data is identical to the original data before it was compressed. Lossy compression involves removing data that wouldn't be missed by the viewer (or would be least missed by the viewer). For example ATRAC audio compression used in recording Minidiscs removes quieter sounds that are obscured by other louder sounds. This gives a greater scope for compression at the loss of some image quality. Standards for video compression are set by MPEG (Motion Picture Experts Group). Their standards are used in the recording of video on CDs (MPEG-1) and digital television transmission (MPEG-2), though these media have a much higher bandwidth than the Internet (approximately 9 megabytes per minute for CDs and 30 megabytes per minute for digital television). These use a form of compression called DCT (Discrete Cosine Transform). All the companies developing video streaming software (which will be mentioned later) have developed proprietary CODECs (Compressor-DECompressor algorithms) that enable video to be transmitted over a normal telephone line. The quality of these is very poor - the only real solution to the problem is to use a broadband (high bandwidth) connection to the Internet. (Hunter, et al. 1998)

There are a few ways to do this - primarily cable modems and xDSL (Digital Subscriber Line) technology. Cable modems allow computers to connect to the Internet through the cable system, though the cable company must upgrade their equipment for them to work. They typically allow a maximum of 15 megabytes per minute, which is more than good enough for television quality video. XDSL technology allows computers to connect to the Internet at a faster rate through its existing copper connection to the telephone system. It allows a maximum of 75 megabytes per minute, but requires costly upgrades at telephone exchanges and it requires the viewer to live within a certain distance of the exchange. There are other methods of connecting to the Internet in between these capacities and traditional modems, such as ISDN (Integrated Services Digital Network), but their capacity isn't high enough for television quality video.

If many people viewed video at this bandwidth the Internet itself would come to a standstill due to the amount of data - the main backbones (the high bandwidth connections that make up the Internet across the world) have a limited, albeit always increasing, capacity which already causes the Internet to slow down during peak hours. A solution is multicasting - broadcast of Internet packets from one source to many, selected destinations. This is as opposed to unicasting (from one source to one destination - how Internet transmission normally takes place) and broadcasting (from one source to every single destination). To work multicasting must be supported on routers (the devices that route information from one point to another on the Internet). Currently it is not supported on most routers, but this is rapidly changing. (Savetz, et al. 1996)

QoS can be improved by using different protocols to transmit the video. Most data is transmitted across the Internet using TCP/IP (Transmission Control Protocol/Internet Protocol). IP defines the structure of the packets and TCP is a method used to make sure they arrive at their destination (involving sending retransmission requests for packets that go missing). TCP is unsuitable for video streaming as by the time the packets are retransmitted they are of no use - it is better just to skip the missing information. As each packet is quite small a few missing packets shouldn't make much difference to the viewer's image. An alternative to TCP is UDP (User Datagram Protocol), which is used by video streaming software) - any missing packets are merely dropped.

A reservation protocol can be used to put a certain amount of bandwidth aside for streaming between the source and destination of the video. This provides a greater guarantee that the streamed video will get through reliably. It also keeps aside a certain amount of bandwidth for normal traffic - so if too many people try to open streams then they will get a 'busy' message rather than all of them getting a degraded connection. The main reservation protocol is called RSVP (Resource Reservation Protocol). The Internet is currently being upgraded from IP version 4 to version 6, which will also provide additional facilities for reservation and to improve QoS. (Hunter, et al. 1998)

There are many video streaming products available. The most popular are RealNetworks RealVideo, VDONet VDOLive and Microsoft NetShow. RealVideo uses fractal video coding technology called ClearVideo. Their audio streaming software is the de facto standard for audio streaming so it has large leverage in the video streaming market.

VDOLive uses the VDOWave wavelet CODEC. It is used by CBSNews, ABC, MSNBC and PBS among others. It is scalable, which means that more or less information is transmitted to the viewer depending on the size of the connection.

Netshow is Microsoft's contribution. Unlike the other products it can use any CODEC installed into Microsoft Windows - which and Microsoft have arranged licensing agreements so it includes all the CODECS used in the above packages, as well as its own. It is very powerful and includes multicast and interactive capabilities. It has been integrated into Microsoft Windows, which means it is included with almost all new PCs.

A number of sites have started offering streamed video. These include broadcast.com which is quoted on the NASDAQ exchange and claims to be ' the leading aggregator and broadcaster of streaming media on the Web.' (broadcast.com 1998) It was valued on the NASDAQ exchange at approximately $1.3 billion (source: Bloomberg). It broadcasts a wide selection of channels and programs including old films and live streams from the American Independent Network and many local channels. Another is FasTV which is trying to create a searchable archive of shows from every channel. They have created a system that allows them to instantly digitise and archive existing broadcasts (DeMocker 1998). InterneTV claims to be the first broadcast station on the Internet - they have been broadcasting video 24 hours a day since early 1995. They also have two affiliates - KVR-InterneTV and JAM-InterneTV. Most of their content is music, though they also have performing arts and some film shorts and trailers. (Ashcraft 1997). All of these are in their infancy and offer a very limited range of programming through generally unintuitive interfaces.

In comparing traditional methods of broadcasting with the Internet I shall use the following criteria.

• Image/Audio Quality
• Cost
• Content
• Choice
• Client Hardware
• Interactivity
• Availability

These represent the dimensions along which consumers and producers would evaluate the different platforms. They are similar to the criteria defined in the Digital Television Market Study (Jaroker 1998) but I have added receiver type, interactivity and availability as they are greatly affected by using Internet delivery.

The baseline for image quality is the quality of current broadcasts. These are transmitted using an encoding system called PAL in most of Europe and a different system called NTSC in America and Japan. These provide 525-625 lines with 25-30 interlaced frames per second. The interlacing means that 50-60 "fields" are transmitted per second, but each field only contains alternate lines of the picture, so it takes two fields to make up one frame. Sound is transmitted terrestrially using NICAM (Near Instantaneously Companded Audio Multiplexing) and using sub-carriers on cable and satellite, all of which are similar in quality to FM stereo. Digital television transmission methods use MPEG-2, which has higher quality pictures and CD quality sound. This will soon be ubiquitous as it is also used by DVDs (Digital Video Discs), which are the replacement for VHS videos and are already becoming popular. To reach the mass market the quality of Internet transmission would have to at least match it.

As stated above under Video Streaming (page 5) there are many ways available of encoding Internet video each of which can produce different quality results. As long as broadband connections become widely available and the QoS protocols are implemented, CODECs that meet the quality of MPEG-2 (or even MPEG-2 itself) will be available.

The cost of transmitting television programmes can be split into two - the cost of creating or buying the programmes themselves and the cost of distribution. These costs are spread between the viewers and advertisers. Over time viewers are paying more - in the US in 1987 viewers paid 27% ($10bn) and in 1997 they paid 42% ($32bn) (Unknown 1998b). Extra revenue streams would be possible with interactivity.

Terrestrial transmission are traditionally free-to-view, with certain exceptions such as Canal+ in France, which requires a monthly subscription. However with the advent of digital many of the new channels require a monthly subscription and there are also pay-per-view events and films. Cable and satellite both require a monthly subscription and additionally an decoder box (and a satellite dish in the case of satellite). They both also offer pay-per-view events. All these services also gain revenue from advertising. In all these cases the same companies handle programming and distribution.

In the case of using the Internet the distribution and creation of programming would be separated, which could potentially create much more competition. The viewer must have an Internet subscription, though these are gravitating towards lower charges subsidised by advertising (Freeserve, operated by Dixons, is among several companies in the UK offering free internet access on this model). Traditionally users in the UK and some parts of the US are charged per minute, though costs are going down every year. @Home, a US Internet provider that is one of the first companies to provide broadband cable modem access and may prove a model for further companies, has no access charges, just the monthly subscription/line rental charge. (Seidman 1998)

The advantage of the Internet is that once the user is online, payment structures would be completely flexible - for example there may be a charge per programme received, charges per channels or a charge that allows the viewing of any program within a programming library. Programming could be given away as to tempt viewers to watch more programmes within a series. There could even be tiered pricing for differing amounts of advertising or for enhanced audio quality. Satellite and cable have already proved that many people are prepared to pay extra for more choice, but there is no real experience so far of marginal pricing, ie the more television the viewer watches the more they pay.

Payment on the Internet for small transactions can be a problem. For larger purchases (ie five pounds or above) credit cards can be used, but for smaller transactions the administration charge makes this unfeasible. To have full freedom of choice on the Internet, ie to make it feasible to watch individual programmes made by independent programme makers, an alternative payment method must be used. There are various companies developing forms of 'digital money' which can be used to make 'micropayments'. These generally involve the payment of an amount of money into some Internet wallet software. They can then which they can spend this with companies who have subscribed. An example is CyberCoin from CyberCash which was introduced in September 1996 and is already being used by InterneTV for their video on demand. (CyberCash 1998)

Advertising becomes complicated when content crosses borders, but using scripting (a means of automating the video player) it would be possible to create a video stream that automatically switched the viewer to a localised advertising stream during breaks and then switched back to the main stream. (Glave 1998)

Interactivity is the ability of the viewer to interact with the television system. Basic interactivity would be a quiz show where viewers could join in or the ability to purchase items (including pay-per-view films and events) seen advertised by pushing a button. At its extreme the viewer would be able to decide what they would want to watch and when. There have been many attempts at interactive television going back to the 1970s, but until now none have been successful (Stewart 1998).

Most television broadcasting is currently one-way, thus offering no interactivity or mock interactivity (such as Teletext). A limited degree of interactivity can be added to satellite broadcasting requires by requiring the decoder to be plugged into a phone line. This has been implemented by DirecTV in the US and is mainly used for purchasing pay-per-view. (Unknown 1998a)

Cable and xDSL provide the bandwidth required for full interactivity and have been used as the basis for a number of interactive trials, though there are no companies are as yet using it for television transmission. According to Andrew Currey, Head of Interactive Television, Videotron UK (a cable company), "The limits to the system, in the analogue environment, have been the capital cost of the technology required in a network box, without an obvious revenue stream to underwrite it. In the digital environment, however, this changes immediately." (Curry 1996) Cable companies are only just making the transition to digital so its capabilities in use remain to be seen.

The Internet is inherently fully interactive. Due to the massive amount of investment and research that is being carried out many products have been created to make it very easy to create interactive Internet applications. This is being integrated within video streaming products - for example the VDOLive player allows hot spots within video adverts to link to another video or a web site (Naden 1997). Microsoft Netshow also has full interactive support including interaction between video and web content and the ability to introduce interactive elements into video streams which are being created live. (Microsoft 1998)

The most popular multimedia interactive content development system on the PC and Mac is currently Macromedia Director. Macromedia have developed a web version called Shockwave which has become the most popular web interactivity tool and has streaming abilities. It is currently limited to sound, but video will soon be added. This provides a wide base of multimedia developers already using this tool who will be able to create interactive content.

No matter how much choice is available or what the quality of the picture is like, no means of television distribution will be successful without programmes people want to watch. Already some of the major channels, for example Fox News, C-SPAN and M2 (MTV channel 2) are being broadcast 24 hours a day through the Internet, but no premium content such as new films or big sport events are.

Programmes are becoming increasingly like commodities. Satellite and cable companies have found that sport coverage of key teams is 'killer content' which alone is almost enough to make many people subscribe. To keep these rights secure Comcast and Cablevision (two American cable networks) have bought sports teams and News International (owner of the British Sky satellite network) are attempting to buy Manchester United Football Club. Mulcahy (Mulcahy 1998) terms other premium content ‘destination content’. These are programmes that are good enough to make people decide to watch them rather than programmes that are watched ‘because they are on’. Exclusive regional distribution rights of most of these programmes are sold to channels across the world. The prices being paid are spiralling upwards – NBC pay approximately $36 million per episode of ER. For these to be made available on the Internet there must be a means of charging in place to replace that revenue stream and their must be enough revenue available to make it worthwhile changing the distribution agreements. Until this happens the Internet can only be a companion to other transmission methods.

The key advantage of the Internet is choice. Currently the medium with the highest amount of choice is digital television from satellite and cable. The network with the most channels is DirecTV (Unknown 1998a). It has about 200 channels of which most are sports or pay-per-view (DirecTV 1998). With this explosion in the number of channels the EPG (Electronic Programme Guide) assumes key importance as the gateway to the service (OFTEL 1998). This is an on-screen menu system that shows what programmes are currently being showed by channel and by genre. The are still, however, arranged a normal channels - to view a specific programme the viewer must select the right channel at the right time.

Interactive TV, whether through the Internet or through a cable/xDSL service, revolutionises this. The idea of channels becomes redundant. The viewer can choose what they want to watch and what time they want to watch it. This is a complete paradigm shift. Interactive trials have been very artificial and not long enough to get conclusive results – as Mulcahy (Mulcahy 1998) states, "the interesting indicator is what actually happens when the service settles down. The question revolves around whether or not the service is firmly established as a ‘must have’ within the household when its novelty value disappears."

The Internet provides an additional advantage over normal interactive TV that it can be received from anywhere in the world rather than just local content. It also means that costs of broadcasting are drastically reduced - anyone can create and broadcast programming. This is termed 'narrowcasting'. Computers and the Internet has enabled anyone to create niche content and web sites that may only interest a small proportion of people but are still successful because of their world-wide reach. Video streaming allows the equivalent to happen to television programmes. An example is Westerns.com from which old Westerns which are not shown on television any more can be viewed (Bunn 1997). The Internet is simultaneously global and local - it is already possible to access the local news for many areas of the USA from anywhere in the world, perfect for ex-patriots. There are also enormous possibilities for community television and distance learning. The University of California at Berkeley, for example, are building up a video database of course lectures, seminars, etc. (Rowe, et al. 1996) To go even further it would be possible to show weddings live on the Internet so any family members and friends who couldn't be there can watch them.

Accompanying this is the constant reduction in the price of video production which parallels the reduction in the cost of audio production that has meant many records are created on 'project' studios which are built for thousands rather than millions of pounds. This doesn't replace the big studios who create expensive, popular programmes but supplements them. Under these circumstances the EPG becomes even more important. With large amounts of content it moves more towards being like search engines. Anyone could create an EPG, so users would have choice as to which to use. Some may rate programmes, some may only allow those considered to be above a certain quality threshold and some may have specialised uses. An example is FasTV, which is trying to create an archive of programming from every channel, tied together by a natural language search facility. (DeMocker 1998) So far companies in this segment, like FasTV and broadcast.com are not facing any competition from portals (pages with search facilities which are used as gateways to the rest of the Internet) they will ultimately try and move in if they see it as a profitable area.

However all this choice may not be wanted. Even with the number of channels we have now, Mulcahy (Mulcahy 1998) states that people tend to watch three to four channels from a multi-channel package and just watch those most of the time.

Different client hardware (ie equipment at the viewers home) is required for different television distribution formats. Terrestrial television requires a tuner, which is built in to the same unit as the screen. Cable also requires a tuner which may or may not be built into the television depending on the cable system and country. Satellite tuners are never built into the television. Satellite has the added disadvantage of requiring a satellite dish.

Whether the tuner is external or internal, they are extremely easy to use and provides instant access to programming currently being transmitted. Television units are available at many price levels with a variety of screen sizes.

Traditionally the Internet has used the PC as its client hardware. The PC, with its generally small but high quality monitor, has been adequate as it is normally used by one person at a time for active content. Television, however, is normally viewed by several people and is mostly passive content. PCs also far too complicated and difficult-to-use for the purpose of watching broadcasts. The solution is an Internet-enabled television or an Internet set-top box (similar to the tuner boxes used in cable and satellite).

Many companies are currently developing set-top boxes which integrate television and the Internet, though they are at a very early stage and have not yet had mass acceptance. One of the most prominent is WebTV, a division of Microsoft. WebTV has a keyboard which is linked to the box by infra-red which can be used to view the web or for e-mail. By pressing a button the viewer can switch between the Internet and television. Currently video streaming is not supported as it uses too much processing power, but as the boxes become more powerful, it will become standard. A newer product which shows a possible way is Sega Dreamcast - which is marketed as a computer games console but it runs Windows CE (a cut-down version of Microsoft Windows), can be directly connected to the Internet and is more than powerful enough to decompress and display digital video (Vink 1998). For television access a very easy to use interface must be developed that allows fast and direct access to the electronic programme guide. Consoles, such as the Sega, provide a good model as the games must start fast and easily they must have very intuitive and attractive user interfaces.

While Internet video delivery may become popular with set-top boxes, as cable and satellite have, to become universal, televisions must come with Internet facilities built-in. Decompressing streamed video is similar to the decompressing other forms of digital TV, so will become very cheap within a few years because of mass production and declining silicon prices. Other parts required to access the Internet such as CPUs and broadband modems must also undergo a substantial decrease in price. This is possible through Moore's first Law - Gordon Moore, the co-founder of Intel, stated that every 18 months chip prices halve and their power doubles.

As mentioned under Image/Audio quality, broadband internet access is required for television to be available over the Internet. This means that availibility will be restricted to those who can afford to pay for it and live in areas where it is available. Already availability issues are occuring with other forms of communication – for example cable television is only available in the areas which the cable companies think will be the most profitable. According to Mulcahy (Mulcahy 1998), it is unprofitable for cable companies to provide infrastructure in 25-30% of the UK. Many of these are rural areas where xDSL technologies can’t be used as they only work upto a certain distance from the exchange. This is a similar situation as with mobile phones, where it is unprofitable to provide coverage in some rural areas. These geographical problems will slowly disappear as wireless satellite bandwidth increases – at the end of 1998 Motorola release Iridium which is a mobile phone network that uses low oribiting satellites to provide 100% world coverage. It is currently very low bandwidth and very expensive to use, but bandwidth is increasing and competing services are being built which will make the services much cheaper. (including Teledesic which is partly owned by Bill Gates).

Cost is a bigger barrier – for television through the Internet to be as available as terrestrial television, the price of an Internet receiver, MPEG decoder and related electronics must come down to near the price of an analogue tuner for it to become practical to include the facilities in all televisions and thus for it to be available to all. As stated under Client Hardware above (page 14), this should happen over time. Another possibility that would limit the availability is if the industry moved to getting most of its revenue through viewing fees rather than advertising – ie only government run television would be available for free.

From the criteria above, there are a few pre-conditions before Internet streamed video can match television. There must be widespread broadband access to the Internet. Spencer Grimes of Salomon Smith Barney estimates that by the end of 2000 there will be 9 million homes with cable modems. (Altman 1998) This would have to accelerate over subsequent years. There must be one or more established providers of digital money. Recently DigiCash, one of the most prominent of these providers has gone bankrupt, but the technology does now exist and CyberCash is valued at $22 billion (source: Bloomberg). The innovative, distributed nature of the Internet should help it reach these goals.

Even if these are fulfilled, they are merely pre-conditions for the Internet to compete. Consumers would then make the final decision, based on whether they to find the advantages of the Internet (primarily true interactivity, wide choice and narrowcasting) compelling enough to make them buy set-top boxes. In the meanwhile the move towards digital terrestrial, satellite and cable broadcasting will continue.

Unfortunately there is very little research and about the use of streaming for broadcast quality television – most is about other ways of combining the Internet and television and about streaming low quality material to current modem users. It would be very useful if an interative television trial was to be carried out over a selection of broadband home users (ie cable modem users).

Altman, L. (1998). The Next Step. In Broadcast and Cable, Address: http://www.buyside.com/features/jun98/broadcast/companion01.html

Ashcraft, J. (1997). The first broadcast station on the Internet showcases videos Address: http://www.internetv.com/html/press.htm

broadcast.com (1998). About broadcast.com Address: http://www.broadcast.com/about/

Bunn, A. (1997). Hitching Nostalgia on Streaming Westerns. In Wired News, Address: http://www.wired.com/news/news/culture/story/7675.html

Curry, A. (1996). Key Marketing Issues In Attracting The Interactive User, Address: http://www.itvnews.com/research/curry.htm

CyberCash (1998). CyberCoin Information, Address: http://www.cybercash.com/cybercash/services/cybercoin.html

DeMocker, J. (1998). TV by the Bit. In Wired News, Address: http://www.wired.com/news/news/technology/story/14168.html

DirecTV (1998). DirecTV Channel Lineup, Address: http://www.directv.com/programming/index.html

Glave, J. (1998). SMIL Hopes to Weave the Streams In Wired News, Address: http://www.wired.com/news/news/technology/story/8843.html

Hunter, J. Witana, V. and Antoniades, M. (1998). A Review of Video Streaming over the Internet, Brisbane, The University of Queensland Distributed Systems Technology Centre.

Nua Internet (1998). Nua Internet How Many Online December 1998 Address: http://www.nua.ie/surveys/how_many_online/index.html

Jaroker, J. (1998). Digital Television Market Study Address: http://pw1.netcom.com/~jjaroker/dtv/RA6.html

Microsoft (1998). Microsoft Development Network Library, Microsoft Software

Mulcahy, M. (1998). The TV Viewer of the Future, London, Financial Times Media and Telecoms.

Naden, D. (1997). Can TV stations use Streaming Video to make $? In theAntenna June 1997, Address: http://www.theantenna.com/97/jun/bottomline/dnaden_vdo6_97.html

NIP BT Interactive Television Trial Address: http://www.tagish.co.uk/nip/nip/261a.htm

OFTEL (1998). OFTEL’s first submission to the Culture Media and Sports Select Committee Inquiry into audio-visual communications and the regulation of broadcasting.(OFTEL), London.

onDigital (1998). ONdigital launch weekend success Address: http://www.ondigital.co.uk/sitev2/who/press_main25.htm

Riga, A. (1998). Look TV wireless alternative to cable, satellite, In The Montreal Gazette, Wednesday 9th September.

Rowe, L. A., Berger, D. A. and Baldeschwieler, J. E. (1996). The Berkeley Distributed Video-on-Demand System, In Multimedia Computing - Proceedings of the Sixth NEC Research Symposium, Ed. T. Ishiguro, SIAM, 1996.

Savetz, K., Randall, N. and Lepage, Y. (1996). MBONE: Multicasting Tomorrow's Internet, California, IDG.

Seidman, R. (1998). AOL's Broadband Woes and Foes. In Seidman's Online Insider, 5(22), Address: http://www.onlineinsider.com/html/july_5__1998.html

Stewart, J. (1998). Interactive television at home: Television meets the Internet, Edinburgh, University of Edinburgh.

Unknown (1998a). Digital Satellite TV FAQ Address: http://www.cdnet.com/dishman/faq.html

Unknown (1998b). Technology and Entertainment Survey. The Economist, Issue 47, 1998.

Vink, G. (1998). Dreamcast Technical Pages Address: http://www.canadawired.com/~gvink/Sega/index.html [Concepts, 1997 #6]