To use Shazam [www.shazam.com] dial 2580 on your mobile phone, then hold the phone up to a music source, like a loudspeaker in a pub, which is playing some music you'd like to identify. The source needs to be loud and clear enough for 20 seconds. A few moments later you receive a text message to the same phone telling you the name of the song [or telling you it can't identify it].
It costs 50p to be told what the song is, or 9p if it can't - either because it doesn't recognise it or because you didn't get near enough to the source.
It works by taking a sample wave form and matching against a database of millions of songs. So how can such advanced audio-matching techniques be use in other settings? Perhaps it's voice-recognition that knows it's you calling, as the basis for personalisation services?
[NB I have no interest in Shazam, I just think it's very clever...]
Comments