8 March 02004

Doubts about models for listening to music in the future

I've been reflecting more on my claim that online radio is the model for listening to music in the future, helped by a range of exchanges with others.

Being sceptical I've so far come up with four types of reasons why my bold conjecture might come unstitched:

  • Socio-Cultural;
  • Economics of Technologies;
  • Economics of Rights Management;
  • Aesthetics.

Of these I think the first and last are most interesting (but then I'm not a techie or a lawyer, so no surprise there).


In his presentation to the ISP Forum a few weeks back, Stephen Gale, CTO of BT Online Services, described a future of wirelessly networked devices, including stuff that looks like a funky old Roberts radio, but has a broadband connection built in. Clearly he's on the same wavelength as Paul Sanders' 'ubiquitous music' premise which I started from before. Stephen wanted these user-friendly devices to enable listeners to choose to buy an album or track if they liked what they heard on the netcast — as Last.FM offers now, but without all that complicated web browser interface.

I discussed this with Stephen afterwards (our paths crossed ten years ago when we were both working in the Human-Computer Interaction area), and of course BT Online Services knows all about Last.FM, Live365.com and all those offerings. Stephen suggested that the success of services based on these models depends critically on finding the kind of user interface that users will really take to.

But this isn't the kind of user interface design where you build a prototype from first principles based on a functional spec and then iterate it with a sample of user representatives. We are not starting with a blank slate when it comes to listening to music: the 'users' already have a lifetime of expectations about how to select and listen to what they want. We need long term and longitudinal ethnographic studies that chart how habits change in response to changes in format and economics (the shift from vinyl to CD to digital file, the falling unit cost of recorded music, and the development of ever more broadcast and netcast channels).

To give an example of what I mean by this, here are some examples of changes based not on research but just on reflection on my own 25+ years of buying and listening to music (each of these may or may not be generalisable beyond music fans in their late 30s!).

  • Your relationship to the music your collection is qualitatively different when you have 1,000-2,000 albums compared with when you had 50-100.
  • Your spend a higher proportion of time 'grazing' on music, and less on focused listening.
  • Your new CDs are less treasured, more disposable items than your old LPs.
  • You select music to complement your mood: it has to fit in with the cultural ergonomics of your lifestyle, rather than take centre stage.

A stack of CDs is an 'interface' to music just as the Last.FM web page is. It has certain advantages over Last.FM that we shouldn't underestimate:

  • convenience and familiarity;
  • reasonable portability;
  • manageable (not overwhelming) choice;
  • ease of browsing;
  • ease of arrangement for future browsing/selection;
  • a certain object fetishism (though jewel cases are so much less lovable than 12" vinyl).

I created a mirror of my original article on the Ecademy site, and the comments made in response to this suggest that some people see the idea a music collection as directly transferable to digital storage and almost infinitely extensible. It doesn't matter whether 'experts' think that the concept of a 'collection' will become history in a future of ubiquitous music, where there are no practical boundaries to the music you can access. If people have significant parts of their lives and their identities tied up in their collections, they won't give them up lightly.

Economics of Technologies

My conjecture concerning the online radio model was based on Paul Sanders' conjecture about ubiquitous music. I guess that in turn depends on assumptions about ever increasing bandwidth and ever increasing storage capacity etc. As I've said in the Ecademy comments, these developments would trigger a point when listeners may not know whether one track is being played from their hard disk, from a server on the other side of the world or from some local/regional cache. They won't know and, as long as the sound quality is even, why should they care?

But these assumptions could be wrong. There's an argument that we will never have 'enough' bandwidth. Certainly as bandwidth increases, it would be good to switch to a format with better sound quality than mp3s, which may mean bigger files and higher data rates. Perhaps then — as in Stephen Gale's scenario — we will use the online radio model mainly for discovering new music that we like, and then buying our favourites on super duper surround-sound enhanced SACD format.

Economics of Rights Management

If you come up with a good idea, there's always a fair chance that some lawyers will find a way to block it. Often they do this by cultivating unnecessary fear among their clients.

The nice thing about the online radio model is that it reduces the need for you to own the music: you get the music when you want it and don't have to have a permanent copy. Thus the Digital Rights Management for the music is not directly your problem, but is handled by the service(s) that provide you with access.

If rights owners see the new models as a threat to their revenue streams, they may take steps to prevent the service providers licensing their material at a rate that would make the new models viable.


Over the last few weeks Last.FM has provided me with a lot of interesting and satisfying listening. But I've noticed that the majority of their listeners do not seem to have developed a long-term Last.FM habit, and few have listened to more than 600 tracks. It may be that the experience of listening to peer-based music profiles (amateur virtual DJs) begins to pall after a certain time. Meanwhile the costs of providing a professional DJ service sufficiently finely tailored to individual tastes may mean that this can be only an occasional habit.

Those are open questions that I'm trying to answer by listening to a lot of online radio.

Posted by David Jennings in section(s) Future of Music, Ideas and Essays, Music and Multimedia, Radio on 8 March 02004 | TrackBack
Post a comment

Remember personal info?