I’ve been playing with Spotify for a while now, mostly using it to listen to music at home when I wanted to listen to a song I didn’t own. Recently, thanks to Owen Stephens (via Twitter), I found out that the service also contains a substantial collection of classical music and so could be of use academically to students at my conservatoire.
Spotify is a service which allows you to listen to music for free online through its application which you have to download and install (available for Windows, Mac and Linux via Wine). Its advantage over services like Last.fm radio is that you can search for specific tracks, and as long as they’re licensed to the service you can listen to them in full. It’s funded by advertising, so you’ll occasionally get ads appearing on the app or radio-style audio ads between tracks – the latter are a bit annoying if you’re in the middle of an album or something atmospheric, but a price worth paying in my opinion.
At the time of writing anyone in the UK can sign up for a free account, but people elsewhere must be invited by current users in order to control growth of demand. This may start again in the UK if demand grows too quickly, so beware!
The below is all my personal opinion after trying out the app for a little while. I’ve split it into sections to hopefully aid readability.
Classical coverage is decent but not comprehensive, gauged by searching for upcoming performance pieces at my college. There’s a fair smattering of ‘big name’ performers thanks to EMI Classics, and coverage is likely to improve with yesterday’s announcement of a deal with Naxos (depending what figures in that press release you use, that could be 10-25% of Naxos Music Library available free on Spotify!). Something that attracts me is the availability of musical theatre recordings which is a weak point in all our subscription resources. There also appears to be decent jazz coverage – though this by no means my specialist subject, a few cursory searches threw up plenty of results.
The metadata for classical music sucks. The ‘artist’ can either be the performer (one or more of conductor, ensemble, soloist etc) or the composer. If the composer isn’t the artist, they end up shoved somewhere in the track or album title. Occasionally there’s no indication who the performer is at all! As a commenter on the Spotify blog said, they should be consistent – a good start would be to follow the MusicBrainz Classical Style Guide. This is the end result of shoehorning classical music into an album-based system, when the work rather than the album or song is the key organising concept in classical music. The new iteration of the Classical Music Library is another example of this, and one I may come back to another day.
Thankfully the search function on Spotify works very well, so you can forgive some of the metadata messiness as you do tend to find what you want in the end. Browsing ability is impaired by inconsistency, but works very well for pop, jazz etc – click on an artist name to see everything available by them; click an album name to see the full track listing plus reviews from Allmusic.com; use advanced search features if you want to. Double-click a song title to listen, drag it to your play queue to enqueue it or a playlist name to add it there. All pretty intuitive.
Sound quality is fine – not CD quality but adequate and comparable with paid services like Naxos Music Library and Classical Music Library. I think there’s a bit of compression/limiting kicking in on pieces with a large dynamic range but not serious enough to damage enjoyment. Apparently bandwidth is around 64Kb/s which shouldn’t cause anyone on broadband/institution connections much trouble.
Correction: Spotify’s FAQ says: “We use the Ogg Vorbis q5 codec which streams at approximately 160kb/s.” so it’s a much higher bitrate than I’d heard elsewhere (possibly on a Guardian podcast, not sure). I have to say I didn’t think it sounded *that* good but the higher figure does make sense. I still hope our institutional connection could cope with many users concurrently.
The biggest downside is requiring the application to be installed – the one thing stopping me rolling it out immediately across the student computers in the library.
Give it a go! I’m going to discuss with my boss whether we should be asking IT to install the app on all the library PCs – my only concern at the moment is whether the app will be able to update itself afterwards, because I’ve noticed it seems to have quite frequent update requirements. There’s a lot I’ve not touched on here with the social side of Spotify – sharing links to tracks and playlists using spotify: URLs or third-party apps like Mixifier – as I’ve tried to focus on why libraries should be interested in Spotify now, and need to think a bit more about this side of things. Have a play, see what you think, and let me know if you agree/disagree with my comments!
There’s been a lot of buzz on the internet recently about the EU’s plans to extend the term of recorded music copyright, which I mentioned last year. The Open Rights Group has created the following cartoon video to explain what the plans are and why they’re bad. I think the number of statistics (and jumps from euros to pounds) confuses the explanation somewhat, but it’s still well worth watching.
Tony Hirst of OUseful.info made an interesting observation about the whole affair: the record companies have built their whole business model around exploiting an ever-growing copyrighted back catalogue, and they’re scared of watching this reliable income stream evaporate, especially given the growth in illegal music copying made ever simpler by technological developments.
Unfortunately, the (recorded) music companies don’t seem to allow much in the way of fair or non-commercial use. For example Tame The Web recently reported that a parody of Madonna’s Ray of Light publicising a library on Youtube had been taken down. The library in question tried to do the right thing and contacted Warner Brothers to request permission to use the song, but ‘no one would give them the time of day’. If the companies won’t help people who are trying to do the right thing (see also DRM which is thankfully dying off), people are inevitably going to start wondering why they should bother even trying. I should point out now that I write this in a personal capacity, and in my job I do take copyright infringement seriously and stop anyone I catch in the act!
Incidentally, I make a distinction between recorded music companies and printed music publishers because some of the latter are a little more enlightened, e.g. giving permission when asked for students to photocopy a piece for the purposes of examination as long as they’re destroyed afterwards (examiners usually need a copy of the sheet music to refer to during the performance, and sheet music isn’t covered by any academic copying licences).
Finally, I thought this was a good time to re-blog this excellent video explaining copyright law using Disney film clips, because Disney has been the driving force behind multiple US term extensions (we can’t have Mickey Mouse in the public domain, can we?).
Librarians have to worry about copyright. It’s an awkward thing, trying to balance the legitimate right of our users to make use of intellectual propery with the equally legitimate right of the creators of that property to be rewarded for their effort. In my opinion it’s particularly awkward in the realm of music, lumbered as we are with recorded music and printed music on top of the usual written word. Of course, the law is a bit different for each one, and there’s no easy licences we can buy to make copying music a bit more straightforward (the CLA offer licences for books and journals).
Unfortunately, the EU seems fairly determined to make life even more difficult for those who study music – they plan to nearly double the current 50-year copyright term for music recordings to 95 years, despite the academic report the EU commissioned advising against term extension.
This would be a disastrous move, maiming pioneering research into recorded music taking place at present for 50 years or so while they wait (again) for recordings to fall out of copyright. For example CHARM has been producing an online discography of 78rpm recordings with some accompanying professionally digitised sound files. The term extension would certainly mean taking the sound files offline – could it even require their destruction?
The end result of a term extension would be that those who care about intellectual property will obey the law and useful work and study using recorded music will be halted (who would willingly go through the nightmare of trying to track copyright holders of 90-year-old recordings?), while those who ignore copyright will continue to do so with impunity.
(Thanks to Ag who posted the ars technica link to the IAML mailing list)
Wordle made this summary – it is such a cool app, go and have a play with it! You can paste in any amount of text (some people have done whole books) or give it a del.icio.us URL and it’ll analyse the tags.
So what’s this Wordle cloud about? As part of my library qualification I wrote a dissertation about… well… you can get quite a decent idea of it from this visualisation! It looked at the possibility of using distributed classification (i.e. tagging) to describe the affective (mood and emotion) content of pop music. For it to be a feasible method I felt it had to have a degree of consistency, hence that word’s size in the summary. To massively over-simplify, I found that there was a degree of consistency: enough to merit further work.
Unfortunately I haven’t yet disseminated these results widely because my supervisor was keen for me to edit the dissertation into a journal article and get published ‘properly’ (under my own name – she wasn’t pushing me for the sake of her own RAE score!). Unfortunately I am not very motivated to do this (I have no ambitions of a academic career) with the result that I’ve been sitting on it for the best part of two years. Perhaps I should just release the PDF onto the web and let the fruits of my endeavours roam free…