For a club disc jockey, keeping the crowd dancing is essential. It’s vital to have a good sense of music and play the right records to avoid the embarrassment of an empty dance floor. Now a team of 15 researchers at Universitat Pompeu Fabra in Barcelona, Spain, has developed a technology that allows even the tone deaf to pick music with the right rhythm and tempo to keep the mass swaying.
In December the university spun out a new company, Barcelona Music and Audio Technology (BMAT) to commercialise the technology, which automatically analyses and annotates the musical properties of tracks, allowing them to be organised based on the similarities between them, within a music surfer programme.
Users can organise a music collection based on different genres without even knowing the title of any songs, or the names of any artists.
"BMAT's technology will go through all of your music collection and analyse it to identify things like the tempo and the rhythm, and will add labels to them," said Xavier Serra, head of the engineering department at the university. "On top of that the application adds the functionality to search and to organise your collection. With this you could organise your music according to the style or the tempo, or you can generate your play list based on other criteria."
Another component of the system is a music 'recommender' that will search online stores for tracks that are likely to appeal to a user’s tastes. This feature, called FoaFing, draws on the musical characteristics of songs, the user’s profile, past purchasing history and what has been written about the songs in website news and reviews, to recommend similar songs. It is based on the Friend of a Friend (FoaF) concept that draws on information from thousands of machine-readable Web pages via RSS feeds.
The BMAT technology can be installed on a Web-based interface such as iTunes, or hardware such as MP3 players.
The technology has already generated a buzz among media content distributors, technology providers, record companies and hardware manufacturers, Serra said.
The BMAT technology was developed in an European Commission-funded project, SIMAC (Semantic Interaction with Music Audio Contents), established in 2004. The project, which received a €1 million grant, carries on until March this year. Other participants for the project include Philips Electronics, Queen Mary’s College London and the Austrian Research Institute for Artificial Intelligence.
Serra said the technology is ready to license and that BMAT is seeking partners and investors.