By Beth Gilomen
We all like to think that we are unique little snowflakes. We assume that we formulate opinions of our own will and make our own decisions about our likes and dislikes. That may be true, but to a certain extent, one company has proven that what we like may not be as subjective as we think.
Platinum Blue, a research company, has developed something they call the "Music X-Ray." Basically, by analyzing hit songs, ranging from classical compositions to current pop music, Platinum Blue has been able to detect upwards of 60 mathematical patterns that popular songs fit. These algorithms are not detectable by the human ear, but they somehow appear to influence how we feel about the things we hear.
The company claims that using their analytical software, a record label can increase the chance of picking a hit from 10 percent to roughly 80 percent. They claim that they were able to predict the success of Gnarls Barkley's single "Crazy" by testing it against these mathematical patterns, and apparently, the program's predecessor predicted Norah Jones' success before she "was on anyone's radar," according to the Web site.
So great, right? Record labels won't have to spend millions of dollars promoting singles that, mathematically speaking, don't stand a chance. That's great for their budgets, but it sets a frightening precedent.
The company's CEO Mike McCready says that analyzing already-written songs will not stifle artists' creativity, but on National Public Radio's Talk of the Nation: Science Friday, McCready said that the software can help artists and producers see where changes could be made that would increase chances of success.
Arguably, the program is only doing, with more precision, what the artist and the producer are hoping to do themselves. But what's to keep independent labels, one's that already don't have any money to waste, from using the software, thereby thwarting the release of music that, by definition, is not meant for mass market release. If label's become overly consumed with possibility of creating hits, music that is narrowly tailored to suit specific audiences may loose a venue in the independent label market.
Obviously, the hope is that the labels putting out independent music are in it for more than money, but the chance to predict a successful, lucrative single could turn well-intentioned indie producers into Eve in Eden. And though the top of the line results can be pricey, a bare-bones analysis starts at $100 per song for labels, or just $10 per song for independent musicians, so if you've written a song you think could be a success, the temptation is high to check and see if you're right.
The whole situation begs the question of art versus technology. If art could become a predictable science, then what exactly is art? It's a line that artists may find themselves toeing more and more in the coming years.
For more information on Platinum Blue's Music X-Ray, or a $10 song analysis, visit www.platinumblueinc.com/.
digg •
del.icio.us •
sphere •
google
•
Comments (1)
Artists have a history of using technology for their own purposes, even technology that threatens to subvert their expression such as this one. This technology is just another channel for their creativity. Imagine someone using it to purposefully make a song that the algorithm ranks as very low. What would that sound like? How many different types of songs could you make that way? What does that say about what we like to listen to? This is not really about art v.s. technology, as there is no "v.s." there for anything to be about. It's just another tool for marketing music. Marketing people will use it to make money and artists will use it to make art. Perhaps other marketing people will use that art to make money, which would be an interesting twist.
Posted by dave | May 31, 2007 04:29 PM