Robin Whittle - First Principles - research and expression

Copyright 1995 Robin Whittle      3 October 1995

Updated May 2006 to contain a zip file of the 350 page draft report Future Developments in the Music Market (1995).  This was in an old Word format which many modern Word programs can't read. 

Updated January 2011 to include a PDF of Future Developments in the Music Market, which has all the diagrams and should be readable on all computers.  See below: #fdmm

World Wide Web:


This is the paper verbatim from 1995.  I hope to update it in early 1999. (Oops - I never did. RW, 2011)
Please read it the context of the diagrams and other material at:
Robin Whittle  7 December 1998

This paper was prepared for the Communications Research Forum in Sydney 19 & 20 October 1995. The Forum is presented by the Bureau of Transport and Communication Economics (BTCE) and the Communications Law Centre (CLC). This is the same text as the original paper except for minor additions intended for readers outside Australia.


By the turn of the century many music consumers are expected to have fast network access, home computers and CD-R disc writers. This will enable them to purchase music via electronic delivery, rather than on physical media such as compact disks. Existing distribution channels and radio's stylistically restrictive music discovery process will be bypassed as artists and listeners engage in two way communication, without geographic restrictions. Radical changes to industry structure are expected. As amateur musicians share music electronically, folk music - withering in the age of mass media - may flourish in the 21st century, in a profusion of contemporary styles.


Introduction and key points

1   Production, discovery and delivery

2   Technologies

3   Marketing techniques

4   Industry structure

5   Protection of intellectual property

6   Themes of change



The Internet today provides new methods of communication between listeners and artists. Despite its low speed, it enables the discovery of music and the marketing of CDs whilst bypassing the traditional industry structures of radio and CD shops.

Within five years, many artists and listeners will have broadband links to the global network - links designed to support the high data rates required by digital video. Since music typically requires a tenth the data rate of video, it will be possible to sell music over the network, eliminating the need to press and distribute compact discs.

This paper explores the foreseeable features of this age of electronic music delivery. It is an extension of work done between October 1994 and April 1995 for the Bureau of Transport and Communication Economics on a draft paper Future Developments in the Music Market (FDMM). The BTCE is preparing an edited version of that draft for publication.  [Dec 1998: this did not happen.]

Predicting technical, social, aesthetic and economic developments is an inexact science, but even if only a subset of these projections eventuate, we can expect a more efficient, personalised and diverse recorded music industry within ten years.

As production technology comes within consumers' reach, the traditional distinctions between production and consumption will be challenged. Personal World Wide Web (WWW) sites today enable a resurgence in folk art. As bandwidth increases, WWW sites will increasingly be used for sharing the music of amateur musicians. The popular experience of music was until early this century largely the result of the efforts of friends and relatives. Mass media's dominance now makes it natural to think of music as an industry. The broadband global network will strengthen that industry and greatly facilitate the exchange of amateur music. While much of this will be derivative of professional music, a significant proportion will be folk music - produced for the enjoyment of friends and evolving independently of commercial considerations.

Beyond recorded music, this distinction between artist and listener will be challenged in another sense when input from both the composer and the "listener" is used by computer software to generate music in the home.

This paper focuses on new ways of marketing professional music, the likely changes to the structure of the industry and on the challenges of protecting intellectual property in the age we are now entering - the age of extreme information fluidity.


1.1 Production

Music production techniques are well known and increasingly coming within the reach of consumers - or at least musicians without external financial backing. DAT (Digital Audio Tape) recorders costing under A$2000 can record sound digitally with near perfect quality and the recordings can be transferred without degradation to CDs.

Advances in electronic musical instruments, MIDI sequencers and multi-track recorders, enable a great deal of music to be produced at home. These advances favour some styles of music rather than others and have been responsible for many stylistic developments in popular music in the last two decades.

While some kinds of music will always require expensive recording techniques, and the employment of skilled musicians, an increasing number of popular forms can be made with the artists' own resources - rather than depending on the assistance of outside investors. This paper does not focus on production, and for simplicity it will be assumed that the expenses involved in music production are not a significant issue.

Beyond financial and technical resources, the experience and motivation of the artists is a crucial factor in the quality of the music they produce.

The aesthetic and social milieu in which artists work is the primary factor affecting their productivity. The increasing use of the Internet, and the move to electronic delivery will change the conditions in which artists work - exposing them to a potentially vast range of music. The social phenomena of fashions, which are currently based on mass media, are likely to change and become more diffuse. New patterns of social and aesthetic development strongly affect both the listener's demand for music and the productivity of artists - by enriching their inspirational environment.

1.2 Discovery

Music discovery and delivery are the two major bottlenecks in today's music industry. Of these, discovery is probably the most restrictive.

Once a listener knows of some music - typically after hearing it, or reading about a new release of an artist they already like - they can usually purchase it. This may involve ordering and waiting for several weeks, but a motivated listener can usually purchase whatever music they like once they know what to ask for.

Very little music is discovered in CD shops. Highly motivated retailers may introduce keen listeners to new music - but this is the exception. Staff assistance is required to listen to any piece of music, and the music is usually heard on headphones, whilst a totally different style of music is emanating from nearby loudspeakers. Even people browsing CDs they are familiar with may be put off by the contrary aesthetics of the music playing in the shop.

The primary means of music discovery today is listening to the radio. Television, listening to friends' music and hearing music at dance clubs, cinemas and theatres are also important. This paper considers radio and new networked means of discovery.

Commercial radio, community/public radio and the ABC (Australian Broadcasting Corporation) radio stations play the primary role in introducing Australian listeners to new music. However there are strong stylistic biases in the music which commercial radio can play. The tracks must be not too long or short - to allow for advertising and other scheduled breaks. They must not be too gentle - otherwise the adverts will be jarring. They must not be too anarchic or controversial - so as not to upset advertisers or alienate listeners. The music chosen for commercial radio generally has prominent vocals - otherwise the advertisements, which must have demonstrative vocals, would seem out of place.

Since the notion of "popular" music is based on radio play and on sales, and since large sales typically only follow radio play, our entire conception of what is "popular" has been continuously skewed by the stylistic bias implicit in the spotlight of radio play. With a stylistically neutral discovery system, the notion of popular music, and the diversity of music sales would be much wider than it is today.

Today's stylistic restrictions and shortage of radio channels means that the variety and volume of music which could be discovered, even by an intent radio listener, is a small fraction of what is available.

Radio and the music industry are awkward partners at best. Radio needs music, but prefers to play what listeners already like - or which is fresh and reasonably familiar to them. The recorded music industry's primary desire of radio is for it to play new material - the music they are trying to sell at that time. The effects of radio's inherent conservatism flow back through the investment chain and make it hard for truly innovative artists to gain the backing they need for their music to be advertised or played widely on radio.

Listeners also discover fresh music through articles and reviews in printed magazines, and through advertisements. Video clips, advertisements, and prominence in magazines and radio are all nearly essential for success in a hit based business model - where the demand may be transient, and must be stimulated and satisfied quickly in order to sell enough units (as discs or cassettes are called) to pay for the production of the music and its promotion.

To achieve this breakthrough into public prominence - before the artist is well known - tremendous financial and organisational resources are needed. This co-ordinated manipulation of the discovery process is usually beyond the capabilities of artists and their management. It is the domain of record companies.

Today, listeners' dependence on mass media for music discovery means that artists whose music is not in the radio and television spotlight can expect at best a slow growth in sales through word of mouth.

The World Wide Web makes it possible to discover music which is not, and perhaps never will be, broadcast on radio or TV. Today, the process is slow - typically limited to one minute mono samples which take several minutes to download, but it is available globally, and the music is available 24 hours a day. The music can be searched for by a variety of means, and is associated with contextual information about the artist, typically including graphics. These sites are typically linked to a means for purchasing the music on CD, since the Internet today is too slow for electronic delivery.

A good example of global retailing of Australian music using the World Wide Web is: [Dec 1998: now defunct.] - where CDs from hundreds of artists are available, each with a one minute downloadable sample, short biography and CD cover image.

This is the start of a new means of music discovery and purchasing - WWW-order mail-delivery. As listeners gain faster links to the Internet the World Wide Web will be of increasing importance - for music discovery and later for electronic delivery.

1.3 Delivery

After starting with cylinders, the recorded music industry settled on discs as the delivery medium for recordings. The Compact Disc would be scarcely comprehensible a hundred years ago, but it retains the same problems of all fixed content physical media - they must be mass produced, distributed and retailed.

These steps involve investment and risk - since there is no certainty that the disc will find a buyer. The high costs of producing and handling these physical objects account for most of the discrepancy between the cost of recorded music to the consumer and the remuneration received by artists. The costs of promotion - artists and their record companies paying for access to the discovery processes - account for significant costs as well.

Electronic delivery promises to reduce costs and risks by eliminating the need for the mass production, distribution and retailing of physical objects, whilst improving the flexibility and responsiveness of the industry to listener's demand. Demand is often fickle and may fade if not immediately satisfied. Getting music to consumers within hours of its creation - instead of months with discs - is sufficient reason for electronic delivery to thrive even if it is no cheaper than using CDs.


This section briefly describes the many technologies which can be expected to assist in music marketing in the next ten years. See Future Developments in the Music Market for explanations and references. Information about many of these technologies, and links to many music sites, can be found with a free WWW search service such as InfoSeek - .

2.1 Digitisation and compression

Music for audio CDs is digitised in stereo, with 16 bit samples at a sampling rate of 44.1 kHz. This accuracy is adequate for most listening environments. It is equivalent to using 1mm graph paper to draw the sound waveform - where the paper is 65 metres high and 44.1 metres wide for each second. Systems offering multi-channel surround sound and higher resolution may also become significant in the future.

CD quality digital audio thus requires 1,411.2 k bits per second - 635 megabytes per hour. (The Bible is around 4 megabytes.) When CDs were developed, there were few techniques for data-reduction - or compression. Modern techniques enable truly lossless compression of music data to around 30 to 40% of its usual data rate. Quality adequate for many listening situations can be provided with compression to between 20% and 4% of the original data rate.

The latest MPEG Level 3 audio compression and decompression software (for MSDOS) can be obtained from Fraunhofer ISS at This enables remarkably good musical quality to be achieved at 128 kbit/sec for stereo. Music is still enjoyable and relatively free of degradation at 64 kbit/sec for stereo. This gives a frequency response up to around 8 kHz at only 4.5% of the original data rate.

This brings reasonable quality music delivery within reach or Basic Rate ISDN - which provides two 64 kbit/sec links. A promising approach for continuous sound (and video) at a variety of data rates is Xing Technologies Streamworks ( This enables reasonably intelligible speech, and recognisable but not enjoyable music over an Internet link using a 14.4 kbit/sec modem.

2.2 Mass storage

Hard disks are plummeting in price and rising in capacity. It can be expected that home PCs of 1998 will have two gigabyte drives, and so will have the capacity to store an entire audio CD's worth of data - 780 megabytes.

The most important development in optical discs is not the new high density pre-pressed discs, but the falling prices of an established technology - CD-R discs and the drives which write them. The drives look like a CD-ROM drive, and can function as one as well. They are part of a computer system and with appropriate software can be used to write blank CD-R discs. These discs can only be written once, and for an audio format disc must be written in one session. Some CD-ROM formats can be written in several sessions. The discs cannot be erased. They can be played on normal audio CD and CD-ROM players.

After continuing price drops in the professional market, CD-R drives are set to become consumer items with prices under A$800 late next year - just as CD-ROM drives were two years ago. Blank disc prices are likely to fall to A$5 or so by 1998. Consumer adoption of computers with 2 Gigabyte hard disks and CD-R drives is an essential precondition for the development of a mass market for music via electronic delivery.  [Dec 1998: CD-R blanks are now about AUD$2 or $2.40 with case.]

Improvements in CD-R capacity and the availability of re-writable discs (using phase-change and magneto-optical technology) can also be expected in the next four years.

Recent decisions in the development of pre-pressed discs indicate that a new standard for single or double sided discs will emerge. This is currently known as the DVD proposal and is planned for introduction in late 1996. Discs will be the same size as today's CDs and will contain 4.7 gigabytes per side, or 8.5 gigabytes per side if a dual layer of pits is used. So each dual layer side contains eleven times the raw capacity of today's CDs - whilst compression technology enables music to be stored three to ten times more efficiently. These discs are unlikely to be used for audio initially - their huge capacity is aimed at MPEG-2 compressed video - which at a typical rate of 4 megabits/sec uses 1.8 gigabytes per hour.

Within two years we can expect a single pre-pressed DVD CD to carry 17 gigabytes - enough for ten days of good quality music compressed to 12% of its usual data rate. These products use red light lasers and within five years a further doubling of capacity may be feasible with blue light systems. However pre-pressed CDs are still mass produced physical objects.

2.3 Networks

The Internet will grow in speed and ubiquity. As more consumers use it and as electronic funds transfer becomes common, the World Wide Web will become a valuable (and later essential) way of conducting personal and business activities.

Today, consumers access the Internet via dial-up modems at a maximum speed of 28.8 kbits per second. There are few if any prospects for this rate improving. This is a good rate for many of today's World Wide Web applications and most delays experienced now are due to the Internet itself or limitations of the server.

Another established network technology is Basic Rate ISDN (BR-ISDN) - which is provided by Telstra (Telstra is Australia's nationally owned telecommunications carrier). Optus (the second, privately owned carrier) announced they would not provide BR-ISDN and would progress straight to HFC cable modems (described below). BR-ISDN provides two 64 kbit/sec digital links over an existing twisted pair telephone wire, with certain distance restrictions. This service is currently priced beyond the reach of most consumers, but there are expectations of price reductions when Telstra completes its "Future Mode of Operation" upgrade to its national network later in 1996. BR-ISDN is barely fast enough for electronic music delivery.

 A mass market for electronic delivery depends on network access at broadband speeds - greater than 2 megabits/sec. Australia's introduction of Hybrid Fibre Coaxial cable (HFC) systems is the largest and fastest in the world - even if only one company rather than two was installing it (Whittle 1995).

While HFC may initially be used for distributive services such as pay TV, their real value lies in their ability to support broadband cable modems. Standards are still evolving and there is debate about demand for bandwidth and the HFC network's capacity to support high levels of usage. However, by 1998 it should be feasible to offer consumers a broadband cable modem with 10 megabit/sec downstream (into the home) capacity and 5 megabit/sec upstream. These will support multiple streams of communication at once, including Internet access which itself can support many concurrent activities.

It may take several years for prices to drop sufficiently for truly mass market adoption of broadband cable modems. Video, at rates between 3 and 8 megabits/sec (for movies to the more demanding fast action sports), will be accommodated by these modems, but it may be the turn of the century before video servers can be developed to give large numbers of consumers Video On Demand - instant access to personalised news, sports, movies, programs and other material.

From one point of view, Video On Demand is an evolution of video - which has traditionally been broadcast. However Video On Demand will also be a part of the evolution of the World Wide Web - as it speeds up and delivers MPEG-2 video in real-time. Whatever terminology is used, there are profound implications in this instant, personalised exchange of all forms of electronic communication, at video rates, with WWW hyperlinks.

Even with more modest cable modems of 500 kbit/sec (as are being developed by local manufacturer NetComm) electronic delivery of music will become feasible. Consumers will need a computer and CD-R drive and the cable modems must provide a cheap connection to the Internet and probably to the underlying ATM (Asynchronous Transport Mode) switching technology on which all future network developments will be based.

The HFC networks are the most important development in the communications infrastructure since the development of the telephone, radio and television. Another technique - ADSL - provides broadband digital links over the existing twisted pair copper telephone wires within a few kilometres of the exchange (Whittle 1994). This may be used where HFC cables are not installed. Telstra is currently trialing the technology to 300 Melbourne homes.

Optical technologies - fibre-to-the-curb and fibre-to-the-home - are too expensive for deployment at present. The HFC networks are flexible and can be reconfigured (at some expense) to provide greater bandwidth for each home. While there is debate about the long term adequacy of HFC - to support High Definition Television - it certainly provides adequate capacity for electronic music delivery.

These two way broadband network technologies are only likely to be available in built-up areas in cities and towns. Some rural and remote consumers may be able to gain network connections via microwave links. Satellites are unsuitable since they are primarily one-way links, which must cover the whole country.

2.4 Encryption and network based funds transfers

Existing technology provides effectively uncrackable encryption for those who want or need security from interception. The same techniques can be used for digital signatures - which authenticate files, proving that the file has not been altered and was signed with a particular person's private key. "Absolute" security is assured by the use of key lengths which necessitate super-computer searches lasting billions of years. Public key cryptography enables this security to be achieved without divulging the private key - so any two computers with a two way communication link can communicate with complete secrecy about the content of the messages. (RSA 1995).

These facilities are available today in a freely available program called "Pretty Good Privacy" and the principles are well understood and have not been found wanting by years of cryptographic research. For better or worse, governments seem unlikely to stop consumers, businesses and criminals from using these techniques.

Secure Internet based funds transfers are taking place today with several schemes based on existing credit cards. One of these is First Virtual ( Agreements between credit card operators and Netscape Communications will also allow credit card transactions with security and greater convenience. There are several proposals for digital cash - financial transactions which are not traceable and so ensure the consumer's privacy. The most advanced of these is David Chaum's DigiCash ( Personal computer software and industry standard protocols for secure, secret, authenticated communications and funds transfer are likely to be mature by early 1997.

2.5 Encryption and copying

The combination of encryption and the global broadband network means that we are entering an era of extreme information fluidity. Anyone can send any information to anyone else without detection - even if the transaction breaks copyright, privacy, security and censorship laws. We need new ways of thinking about and protecting intellectual property. FDMM, and section 5 of this paper take a functional overview of copying and revenue enhancement measures. An important essay on copyright was written by John Perry Barlow. (Barlow 1994).

Currently the physical difficulty of copying material - such as the sound on CDs - is the major deterrent against widespread consumer copying of music. However copying will soon be easy with CD-Rs and broadband networks.

Despite claims regarding digital watermarks, special codes and network copyright surveillance schemes, there can never be an effective technical means of preventing the copying of linear content. Linear content includes music, text, still images and video. Non-linear intellectual property - computer programs or any interactive application which relies on a program - can easily be copied, but there are some prospects for controlling the running of computer programs. There are no technical means of protecting linear intellectual property from copying or misuse.

2.6 Personal computers

Rapid advances in personal computers can be expected to continue. As the Internet becomes a common part of life, the home computer will become the ubiquitous information appliance. Home computer marketing now appeals to the desire for performance and status - using terms similar to those used for racing cars.

 Personal computers already have the power to decompress MPEG audio Level 2 and MPEG-1 video with the standard CPU - but with significant performance limitations in real-time. Specialised hardware and advances in software and CPU performance mean that good quality real-time audio decompression will be available without using all the machine's resources - so it will be possible to listen to music coming from BR-ISDN or a broadband modem whilst running other applications.

After the year 2000, home computers may be fast enough to synthesise music in real-time. This goes beyond recorded music and the scope of this paper. Such new forms of music are explored in FDMM and in a recent interview with Brian Eno (Kelly 1995).

2.7 Distributive technologies

Pay TV is a well understood technology - whether analogue or digital and whether delivered by cable, terrestrial microwave or satellite. It may expand the music discovery process by providing more channels for music videos.

A more significant development is the provision of multichannel "digital cable radio" - a subscription service which can be delivered by the same means - using the channels intended for television to carry dozens of channels of specialised audio programming. Services such as these - DMX and Music Choice in the USA - may be introduced in Australia on the HFC networks. These services have the potential to greatly enhance the music discovery process. However there is the possibility that listening to these services will satisfy listener's desire for music - rather than stimulate their desire to purchase recordings. These services also present a threat to artist's income due to home recording - so the programming schedules are not pre-announced. The programming is often automated - with no announcer or advertisements. Artist and music title details can be displayed on the decoder or on its remote control. Such services are only likely to be available at home, rather than in mobile situations.

A similar multi-channel distributive audio radio based system, also using MPEG Level 2 audio compression, has been developed and is planned for introduction in Europe. This is Digital Audio (or Sound) Broadcasting (DAB or DSB). The Department of Communications and the Arts Communications Lab has conducted trials in Australia. DAB signals flood entire areas from potentially multiple transmitters - both terrestrial and satellite. The signals are immune to the multi-path distortion which plagues FM radio. Since the music is digitised, it can be encrypted so that it is only available to paid subscribers - however it is envisaged that DAB will typically be free to air. Like "digital cable radio" ancillary data can be sent as well - such as graphics and information about the artist.


Networked communication provides benefits to both the artist and the listener which are impossible with today's physically based selling methods.

While is impossible to imagine all the uses of networked communications, it is clear that the next decade will see an unprecedented empowerment of individuals to communicate bi-directionally with individuals and groups, on a global basis.

It will be assumed that the cost of transferring hundreds of megabytes of music within a country will be just a few dollars. For international transactions, the higher telecommunications costs will not be a problem because a seller in Australia would establish buffer sites in major markets - the UK, the USA, Europe and Japan. Their music could be sent to the buffer site once, and delivered from there to listeners for browsing and purchasing. This does not constitute a distribution arrangement - the buffer site is just a computer which can receive and send data. All transactions take place between the listener and the artist's computer - which would typically be located in their studio or office.

Projecting into the year 2000, many artists and listeners will have broadband cable modems linking their computers to the Internet 24 hours a day. CD-R blank discs will cost a few dollars and will be used to make standard audio CDs, or CD-ROMs with many hours of compressed music. So the costs to consumers of storing the music they purchase will be comparable to today's audio cassettes.

3.1 Discovery - searching and browsing

The listener will have many techniques to find music of interest. A WWW search on an artist's name, or the name of a piece of music, will soon lead to the artist's site, and perhaps to sites containing discussions and other material relating to the artist. Searching for the name of a genre would lead to similar results for the music of many artists. Reviews and "recommended" lists will be popular guideposts. Discussion groups (usenet or mailing lists) will be a fruitful source of direction when searching for music, as will personal WWW sites expressing appreciation for music and artists.

One possibility is electronic music identification - where the listener plays a sample of an unknown piece into a search engine. Within a minute or so they receive links to the sites of artists or other sources of the music - or to music which sounds like the sample. This will be especially valuable when searching for a piece of instrumental music recorded from the radio in a DJ mix when the tracks have not been announced.

Once the listener finds the site, they can quickly view text and graphic information about the artist - while hearing their music or a spoken introduction. Video material may be practical as well. They proceed to a menu of tracks listed by name, with graphic identification, length and purchase price. By clicking on a track, they hear it, and can fast forward and rewind through it. Typically the whole track will not be played at the browsing stage - to reduce the chance of people recording it without paying.

Artists have few limitations on the music they make available. They may offer a dozen mixes of a track - some fans will buy them all. Artists may still wish to package their music as albums - and so listeners would only be able to buy a complete package of music, which could also contain graphics, text and other material.

The WWW server software will typically be able to identify the listener so that listeners who regularly visit the site can be shown only those tracks which they have not yet purchased. Browsers and purchasers can comment on the tracks - in text, with voice-mail or with video. This raises some fascinating possibilities for artists understanding their audience. When the artist checks their feedback they can view their most devout fans expressing their appreciation spontaneously in their own homes - with a microphone and a video camera. Lurid and demonstrative teenage video-mail may contribute to the artist's inspirational resources.

This feedback - effectively mini reviews from fans - can be made available to the browsing listener. In this scenario, the artist's server recycles the recommendations of fans - leading to the server being a global stage and archive for the fan's expression of appreciation both to the artist and to other fans.

The display of each track can indicate its sales volume. It would also be possible to display the names, or aliases, of people who purchased each track. When the number of purchasers is large, the listener could send a list of their friends and people whose tastes they respect, and the browser can display only the names of purchasers on that list. The names could be displayed with dates, in date order, so that the peer group social status of early purchasers would be enhanced - similar to the way high scoring arcade game players enter their signatures so the machine displays the signatures and scores to their friends. Fashion and social factors can be strong demand drivers. Personalised interactive sessions like this can enhance some social factors in ways which are not possible with mass media.

Purchasers could also receive email, voice-mail and video-mail from the artists - presumably in response to significant purchases or noteworthy feedback. These communications may become more highly sought after than the tracks themselves.

All material received in a browsing session can be shared amongst listeners - as can the purchased tracks. The social aspects of music purchasing can be maximised by several friends being in constant text/voice/video communication - from their homes all over the world - as they collectively browse an artist's site. They can talk amongst themselves, listen to the same browsing tracks and dance around their rooms - while a microphone and perhaps a video camera conveys their appreciation to the artist.

Each track within an artist's site would have a stable WWW address and would be available for years after its release. This enables people, when discussing the music in email, a discussion group or in a printed magazine, to include a link to the track so that readers can quickly browse the track being discussed.

There is great potential for the new discovery and purchasing arrangements to appeal to people who are reluctant to venture into CD shops.

3.2 Purchasing and enjoyment

Whilst listening to tracks, some are selected for purchase. When the listener is ready, a total price is presented, with a list of the tracks to be purchased. The listener can review their selections, and confirm the purchase by clicking a prominent and attractive icon. At this point, their WWW browser negotiates with the artist's server about funds transfer and a few seconds elapse while the transaction (digital cash, encrypted credit card information or something else) is completed. The artist's server confirms the transfer with the bank and the artist's bank balance is immediately credited.

Then the music tracks are delivered in a compressed form, while the listener waits for them, or does something else. Within minutes, or at most an hour, the tracks arrive on the listener's hard disk. From there they can be played and will typically soon be written to a CD-R for long term storage. At any stage, these tracks can be retrieved from the CD-R and written to another disc, or sent to anyone else with a fast network link. The tracks can also be combined by listeners - doing their own DJ mix in real-time to form new tracks. Software can automatically do on-the-beat mixing and so can generate a seamless party/aerobics mix from the listener's collection.

It is clear that many novel and enjoyable possibilities exist, and that the direct artist-to- listener communication can be so advantageous to both parties as to make the idea of intermediaries seem anachronistic.


The example above bypasses most of today's music industry structure. There is no need for record companies, distributors, retailers, radio, TV, magazines or advertising.

This example is for music produced by currently active artists with WWW facilities. For music from the past, and from other artists, the listener will be interacting with the server of a distribution or retailing company.

4.1 Radio and distributive media

Radio will still play a crucial role in the discovery process - since it is unlikely that WWW browsing will take place whilst driving and travelling on public transport. One notable possibility with the increased channel availability with Digital Audio Broadcasting and "digital cable radio" and the like is that there may be "stations" which exist purely to aid music discovery. These would specialise in particular genres, so hundreds of channels, perhaps with a global distribution, would be required to cover the breadth of musical tastes.

These "stations" would be free to listen to, and would be financed by the artists and their management. The music would be presented in a way designed to maximise the chances of listeners purchasing it on disc or over the network. So these would be 24 hour a day infomercial stations dedicated to particular styles of music.

This may seem uneconomic given the cost of producing traditional radio programs, and of paying for AM or FM channels. However, programs of this nature can be assembled by computer from contributions from the artists paying for their air time. The signals can easily be transported around the globe and made available on distributive channels, or via direct links to consumers in the broadband network. When these "infomercial" channels are available via the WWW, they need not be real-time signals, but would resemble a searchable archive of material which could be accessed at any time. Thus what starts off as a promotional "radio" exercise becomes a network accessible archive of lasting value to artists, listeners and historians.

4.2 Record companies

It may seem that record companies won't be needed in the future, but their investment, creative guidance (questioned by many artists) and promotional resources are likely to sustain existing companies even when pre-pressed discs are less important.

Many listeners will still want to purchase pre-pressed discs - even if they can purchase the music via electronic delivery. These discs may come from the artist themselves or via a record company which has the physical distribution and business connections necessary to get the product to consumers. Pre-pressed discs, with their specialised and authentic packaging will still be attractive to consumers - many tactile and visual aspects of packaging can never be delivered over networks.

There will still be many consumers without access to electronic delivery or the Internet, or who are not interested in using them, so record companies will remain in business. However record companies will no longer be the only way of selling music, and in many respects they will not be the best. The immediacy and personalisation of electronic delivery is ideally suited to marketing rapidly evolving material where diversity, passion and fashion are salient characteristics of the pattern of demand.

CDs of standard and higher capacities, as music CDs, CD-ROMs and MPEG-2 video discs will all be available in the same packaging, with the same distribution systems based on pre-pressed media. Perhaps record companies will merge with companies which produce and distribute electronic games, video tapes and discs and CD-ROMs.

4.3 Distributors

The prospects for today's CD distributors do not look promising. With record companies still holding the reins of mass market promotion, artists dealing directly with consumers and with consumers having network access to large cut-price WWW-order mail-delivery retailers, how many pre-pressed discs are going to be shipped to retailers without retailers dealing directly with record companies and artists?

Retailers will quickly adopt networked means of ordering CDs. Local distributors' main advantage over more distant sources may be the discs they stock and the speed with which they can deliver them to retailers. But these are expensive, risky activities.

4.4 Retailers

If CD retailers continue to sell only pre-pressed discs, then their prosperity will decline with the rise of WWW-order mail-delivery and electronic delivery. However retailers can use the new technology to provide valuable services to their customers.

Due to their location in business districts, and their access to capital and expertise, retailers can provide customers with music CD-R discs - with the customer browsing the music in the shop from an in-house system or a WWW session. Whether the music is sourced from pre-pressed audio CDs, high density compressed CD-ROMs, telecommunications links to record company servers or from the World Wide Web directly from artist's servers, retailers can offer the benefits of electronic delivery at an early stage, and without the listener investing in a substantial personal computer, CD-R drive and broadband network link. Another niche is where the listener has low speed access to the Internet, and can browse and purchase music, but have the music sent to a retailer who writes it onto a CD-R.

Since a great deal of music is sold to young people, and to those with bohemian lifestyles (and incomes to match), retailers who bring the new technology to their customers can expect to do a good business. For instance, it should be possible for a concert to take place in Germany, and be televised globally, and for an edited mix of that concert to be available on CD-R the next morning at retailers all over the world. Traditional pre-pressed CDs may also be made available several weeks later.

Retailers must make it easier to discover music in their shops. FDMM discusses some approaches to this. The existing situation, where mute discs and packages are browsed when the shop is awash with another kind of music is quite uninviting. Whether potential customers want to browse the music on discs or that which is available for electronic delivery, they need to do it in a conducive environment, and in a way which supports the social needs of groups of people shopping together.

4.5 Economics - purchase price and artist remuneration

For simplicity, this discussion assumes that the artist is both the composer and performer of the music. In an electronic delivery environment, royalty payments to composers who are not the artists selling the music are problematic. However, the searchable nature of the WWW (and artist's will want the tracks on their site to be visible to search engines) will make it easier for composers to search for anyone selling their compositions.

Most of the cost of a CD is a result of the costs and risks of physical manufacture and distribution. With electronic delivery, the operation of servers is automatic and virtually costless. The only fixed costs in the system are those of telecommunications costs for the delivery of the music, and for the blank CD-R disc for the listener to store the music on. The telecommunications costs whilst browsing will be fairly low, and are likely to be met by both the listener and the artist through their Internet access fees.

The transport of 200 megabytes of data (an hour of music compressed losslessly to a third of its normal size) will not be trivial for several years. However, the broadband network will be built to deliver gigabytes of video data, at least within cities, for a few dollars at most, so the telecommunications costs should not be a problem for music delivery by the year 2000. The costs of this transport of 200 megabytes may be borne by the artist, the purchaser or both, but ultimately the purchaser pays for it.

It can easily be imagined that an hour and a quarter of music - equivalent to a CD - could be delivered to a purchaser for four dollars or so. This would cover the cost of transport and, on average, cover the artist's expenses in supporting the browsing activities which result in such a sale. This enables the artist to sell a CD's worth of music to a listener for A$15, or perhaps A$10. The listener would need to spend around A$5 to store it on a CD-R, or less if compression is used, as prices drop or as higher capacity storage systems become cost-effective.

So it can be foreseen that a listener could purchase a CD's worth of music for A$15, of which A$3 may be for the CD-R, A$4 for telecommunications costs and A$8 gross profit for the artist. When these prices may be realised in equivalent 1995 dollars is hard to predict. Perhaps 2000 or perhaps a few years later. Equivalent cost reductions are not available with pre-pressed CDs. The cost of the disc and packaging is only about A$3, and the technologies of pressing and physical distribution are mature and dependent on the costs of labour, fuel and raw materials.

If music was available at half the present price, perhaps twice as much of it would be purchased. The electronic delivery model presented above delivers revenues to the artist which are around twice the best case royalties available with pre-pressed CDs. On these assumptions, a complete change to electronic delivery would quadruple artist's gross incomes from recordings.

With electronic delivery, the cash flow to artists would be immediate - rather than delayed by 6 month accounting cycles. The payments would be direct, rather than after substantial commissions taken by licensees and agents.

Most recording artists today make nothing from their recorded music sales - after allowing for their own or the record companies initial investment. Electronic delivery offers a marketing model with potentially higher returns, fewer risks and fewer business and accounting complications. The computer hardware and software and network links needed for artists to sell their music electronically will be consumer priced - dramatically lowering the entry barriers to artists wishing to globally market their own music.

The quantity of music purchased is affected by factors other than price. Most particularly it is affected by the ease of finding the music and making the purchase. The immediacy of delivery is also a significant factor. Rather than travelling to a CD shop during business hours, listeners can browse buy and take delivery at home at any hour. This increased ease and immediacy could easily account for a further doubling of the quantity of music purchased.

A further impetus to increased sales volume is the ability of artists to release an almost unlimited amount of music - very different from the "one CD a year" model commonly used today, a limit set primarily by record company promotional strategies.

Although it will take some time to dominate over sales of pre-pressed discs, electronic delivery is so attractive for artists and listeners, that its success seems assured.


Whether music is delivered over networks or on discs, it will be sold in an environment of extreme information fluidity. What is to stop one listener buying a track and then making perfect copies, via the Internet or CD-R, for a dozen friends? What is to stop them placing the track on a network site (perhaps their own home based WWW server) so anyone can copy it for free? Or for a fee? Technically, the answer to these questions is "nothing".

There are no technical or cost-effective legal means of stopping users doing almost anything they like with music or any other digital data which comes into their possession. While the prospect of electronic delivery excites artists, and deeply worries record companies, both are concerned about widespread abuse of their intellectual property in the near future, irrespective of how the music is delivered.

This section considers five kinds of copying, and four kinds of response to maximise artist's revenues - from legal, social and economic perspectives. It does not address moral rights or the payment of composers. Copying can be classified in five categories:

1 Purchaser copying
The purchaser making copies for their own use. This is likely to be revenue positive because the buyer will pay more for music when, for instance, they can make a copy for their car or portable cassette player. It can be revenue negative where the listener's ability to copy saves them from buying a copy in the alternate format. The creation of personal collections of tracks on cassette or CD-R is a form of copying which enables the listener to gain more value from the music they purchase. Whilst the music on alternate physical formats can be sold by the artist, artists cannot sell a listener a personalised collection - so the ability of music to be copied is the only way artists can ensure that listeners derive the fullest value from the music they purchase.



2 Listener sharing
A listener (the purchaser, or someone who received the music by other means), making copies to give to a friend - but in circumstances where the friend would not have bought the music anyway. In some instances, this action will cause the friend to become interested in the music and buy more from a legitimate source. (Revenue neutral or positive.) A second reason this may be revenue positive is when the purchaser is prepared to pay more for music if they are able to give a copy to a friend.



3 Listener theft
A listener giving away one or more copies of the music to people who otherwise would have purchased it - but not for financial gain. (Revenue negative, but the effect could be lessened if the purchaser would pay a higher price because they could share the music with friends.)



4 Listener piracy
Low volume copying for financial gain but not as a commercial business - by people who bought the product primarily for their own enjoyment or who obtained it by some other means. (Revenue negative.)



5 Commercial piracy
 Copying music or manufacture of products without the artist's authorisation for commercial gain. (Revenue negative!)



Copying cannot be stopped by technology. The four remaining approaches are:
Legal measures - laws and enforcement procedures
Legal approaches may deter some listeners from copying material, but it is not cost-effective, or good public relations, to pursue these measures to the point of prosecuting listeners.



Enhanced community attitudes
These cannot be changed in isolation. Conscious and unconscious attitudes to artists in general result from everyday practice - a function of technical, economic and perhaps legal factors. Widespread enhancement of listener's relationships with individual artists may lead to generalised improvements in listener's attitudes to the intellectual property of all artists.



Enhancing listener loyalty to the artist's interests
The mass media is not conducive to listeners identifying with artists' real need for their intellectual property to be respected - especially as artists are keen to give it away over radio. Networked interactions provide much better prospects for genuine relationships and for listeners to learn why they should protect artists' interests.



Making legitimate purchase easy, cheap and attractive compared to copying
There are excellent prospects of using networks to make legitimate purchase cheaper, more convenient and more personalised compared to today's approach which is based on CD shops.



The state may assist with the prosecution of commercial pirates. Search functions in the networked environment make it easier to detect such activities directly or via people discussing them. Costs and effectiveness of prosecution would depend on the legal system of the country in which they operated. Increasingly, governments of third-world countries are being pressured to provide effective intellectual property protection.

There is a crucial distinction between (2) listener sharing (revenue neutral or positive) and (3) listener theft (revenue negative). Listener sharing drives the music discovery process (something the artist desires and often pays for) - it is often revenue positive. Listener theft does not, and is revenue negative. The distinction between the two can only be made some time after the copying occurs. The copier's actions and intents may be identical. The deciding factor is how much of the artist's music is subsequently purchased legitimately by the friend.

The law cannot distinguish between these two and in addition, it is not cost-effective to enforce the law on listeners. A blunt approach by any means to stop listeners copying music at all could diminish (2) listener sharing - which makes a crucial contribution to music discovery. Listener sharing will be especially important with broadband connectivity since it will be easy to send a friend some music in the hope they will like it and buy some more of it legitimately. A blunt approach at stopping music copying in general is likely to backfire and diminish many listeners' respect for the law and for intellectual property in general.

The greatest problem lies with (3) listener theft and (4) listener piracy. While a strong and appropriate legal framework for intellectual property is essential, copyright law is ineffective with individuals. Enlightened community attitudes are a second approach - but these will tend to follow daily practice as affected by the other approaches.

There are two promising approaches to minimising copying which diminishes artist's revenues:

  • Enhancing loyalty from listeners to the artist
  • Making music increasingly easy and cheap to purchase
  • Electronic delivery can be used to achieve both these goals in a way which is hard to imagine now, since we are used to mass produced physical objects and purchasing being very remote from the artist. The direct communication, and the knowledge that most of the purchase price goes to the artist, are two of the reasons why listeners will be attracted to legitimate purchase.

    The industry has no choice but to fight revenue-negative copying by enhancing relationships with the customer, and by making legitimate purchasing convenient and cheap so it is attractive compared to illicit copying. Success with both these will change community attitudes for the better.

    The networked based transactions of electronic delivery are ideal for these two most promising approaches.

    A recent and important US Government policy document on intellectual property protection in the information age is available in several formats (IITF 1995). An Australian mailing list in which intellectual property and network policy discussions take place is the Link list (Link 1995).


    The last chapter of FDMM lists the challenges and opportunities facing the industry and some higher level observations - "themes of change". Some of these are:

    6.1 Many ways when there were few

  • Many new promotional, discovery, distribution and playback technologies.
  • The gap between playing recordings and performing live has already been bridged from both ends. New instrument technology will enable more sophisticated use of recordings and software generated music on stage.
  • Styles of control during listening - previously track to track from CD to CD. Later, track to track from any collection of music. Later, changes to the structure of the track while it is playing. Later, multi-variable controls for software generated music and music without a fixed duration.
  • 6.2 Introduction of new ways

    For each instance of the "many ways" trends in the music industry listed above:
  • The existing ways are still available (An exception is that vinyl records are no longer pressed in Australia).
  • The new ways arise primarily from technological innovation and the industry exploiting new niches which previously did not exist.
  • The significance of the existing ways is typically reduced because they no longer have the same dominance in the field.
  • The cost effectiveness of the existing ways may fall in absolute terms because of reduced usage - for instance broadcast TV as it loses market share to subscription TV and VOD.
  • The cost effectiveness of the existing ways may increase due to price competition. For instance, CD prices dropping because of competition from listener copying, overseas WWW based CD retailers and electronic delivery.
  • The cost effectiveness of the existing ways may increase because the old way uses new ways to achieve the same purpose. For instance retailers may be able to afford high speed telecommunications connections and CD-R writers 2 to 6 years before many of their customers. So the new flexible techniques give the old way - the CD shop - enhanced cost effectiveness.
  • Ten years is quite long enough for a previously dominant way of doing things to be completely outmoded - for instance CDs displacing vinyl records.

    Barlow 1994
    John Perry Barlow, The Economy of Ideas: A framework for rethinking patents and copyright in the Digital Age Wired, March 1994.
    Robin Whittle, Future Developments in the Music Market is a 350 page draft paper prepared for the Bureau of Transport and Communication Economics between October 1994 and April 1995. This was part of the Communications Futures Project, in the days of the Keating government.  Around a hundred copies were printed for people attending the Music Industry Summit in April. The BTCE is editing this draft for publication by the Australian Government Publishing Service. [Dec 1998: in the end, it was not published.  I have the Word file of the first draft of FDDM.  Here it is, in a .zip file, with a number of other items of potential interest, including a 1996 update:  FDMM.ZIP].  FDMM explains the many technologies to a non-technical reader, and explores most of the issues raised here in greater depth.  

    January 2011: here is a PDF of that 1995 draft:  FDMM-updated-2011.pdf with all the diagrams.  <<< This is the version I suggest reading first, but please see other material in the zip file if you are really interested.
    IITF 1995
    US Information Infrastructure Task Force, Chair Bruce A. Lehman, Intellectual Property and the National Information Infrastructure 260 pages of copyright law background, discussion and policy recommendations. Available in Word and Adobe PDF formats from This is also available in printed form from the US Patent and Trademark Office Box 4 Washington DC 20231 USA. ISBN 0-9648716-0-1
    Kelly 1995
    Kevin Kelly, Gossip is Philosophy - interview with Brian Eno Wired May 1995 pp146-151, 204-209.
    Link 1995
    The Link mailing list can be subscribed to by sending email to:, with no subject, and with the text "subscribe link" in the body of the message. A WWW archive of the list is available at .
    RSA 1995
    RSA Security, Frequently Asked Questions about Today's Cryptography .

      Whittle 1994

    Robin Whittle, ADSL - Bridging the Superhighway Gap? Australian Communications, May 1994, pp 81-90 Bringing ADSL Home - The Race is On, Australian Communications, June 1994, pp 87-95. [Dec 1998:  My articles are listed at ../telco/articles . ]
    Whittle 1995
    Robin Whittle, Australia's Hybrid Fibre Coaxial cable systems Australian Communications. In press for the December 1995 issue. [Dec 1998:  My articles are listed at ../telco/articles . ]


    The Bureau of Transport and Communications Economics (BTCE) [Dec 1998:  The BTCE has been restructured and the communication and arts people are now in a new organisation called the Communications Research Unit: . ]

    The Communications Law Centre (CLC) contact details are: CLC, The White House, University of New South Wales Sydney NSW 2052 Australia Ph +61 2 663 0551 Fax +61 2 662 6839.