The Null Device
Posts matching tags 'hit song science'
More news from the leaden age of music diversity: major record labels are now using statistical hit-prediction software to pre-screen demos before wasting A&R ear-time on them. Hit Song Science, the software used, predicts whether a song is likely to be successful by comparing it statistically against a database containing the past 30 years' worth of Billboard hit singles:
HSS's crucial design flaw is that it can only look at the past. Those "leftfield", illogical and grassroots-inspired departures from the norm, such as disco or drum and bass, could not have been predicted - but they shift the mainstream and provide the momentum any culture needs to remain fresh. As Smith says, "Art is the one area where people can, and should be able to, make radical statements. Anything that encourages safe, consensus-driven music should be used with caution."
It's all in the clusters, you see. Hit songs, typically, fall into one of a number of groupings - there are around 50 in the US and 60 in the UK where, traditionally, tastes have been more diverse. Belonging to the same cluster does not mean songs sound the same, though, more that they are mathematically similar. And the analysis has thrown up some very unlikely musical bedfellows: Some U2 songs are in the same cluster as Beethoven, while spandex ultra rocker Van Halen sits right alongside MOR piano babe Vanessa Carlton. It is for this reason that Polyphonic are confident their software won't homogenise our already stratified and similar sounding charts. They are already working with one radio station to expand their playlist without losing audience share by selecting songs with the correct mathematical rhythms. In a world where drearily repetitive playlists have become the norm this could be the answer to an oft-uttered prayer.
It's interesting to consider where the fine line between well-formedness and homogeneity is. On one hand, one could probably state with mathematical certainty that serialist or aleatoric composition is unlikely to ever have mass appeal; as such, an algorithm which analysed music for having some kind of structure and rated it on that would be a good predictor of whether or not a piece of music has the potential (however small) to be popular. On the other hand, I suspect that HSS may overspecify things, to the extent of excluding perfectly workable new approaches which have not been tried or accepted yet. (via DIG)
Recording companies using computer analysis to determine likely hit songs. The program known as Hit Song Science claims to detect the deep patterns which appeal to popular tastes; several labels are using it to help plan release dates, or decide whether to put money into releasing a song at all. Could this usher in a new era of conformism and stagnation that will make the late '90s seem like an explosion of creativity, and drive everybody looking for anything other than homogeneous bubblegum in genre drag into the MP3 underground?