Skip to content
News

Opening Pandora.com’s Box

By: Jessica Grose

The music my parents’ generation heard was limited to what was played on the radio or, for the extra-eager few, what could be unearthed through many trips foraging in well-stocked record stores. Nowadays, even though MTV plays approximately six artists, any college kid with a broadband Internet connection can have immediate and unfettered access to millions upon millions of free MP3s and artists so obscure that they’ve only performed once in a shack in New Hampshire for an audience of seven. With this overwhelming media inundation, how can you go about finding music you feel passionate about? Pandora, the website created by the Music Genome Project, seeks to help music lovers find the proverbial white truffle buried in the mountains of Internet music scat.

The Music Genome Project, founded in 2000 by journeyman musician Tim Westergren, started by breaking down songs into nearly 400 specific musical components or “genes,” i.e., the different elements of vocals, instrumentation, melodies, harmonies, and several other “primary colors,” as Westergren calls them. Over the past five years, Westergren and a legion of programmers have analyzed the songs of over 10,000 artists and compiled them into a database, which is where the website Pandora.com comes along.

The way Pandora works is you type a band or a song into the interface. For instance, I put in the Shins. The website then scans its database for music with similar genomes to the Shins and creates a streaming radio station; the Shins station ended up including artists from Robbie Williams to American Hi-Fi to Fugazi. I had the best results with my Joanna Newsom station. Newsom, whose voice could be described as sounding like a preteen in a mental institution — childlike and unhinged — led me to the discovery of several other folksy bands with unique and strong-willed singers of whom I had never heard, like Diane Chick and the Out Crowd.

Pandora’s radio stations are free, as long as you’re willing to put up with ad banners littering your computer screen. A paid subscription to Pandora — which provides advertising-free usage — is $36 for a year and $12 for three months.

Along the way, there is the possibility for user interaction. With each song Pandora offers, the site gives you a chance to rate what they’ve generated. If you click on the thumbs-up icon that reads, “I really like this song – play more like it!” Pandora will respond with a genial response: “Cool. We’re glad you like it. We’ll be sure to play songs with similar musical qualities.” If you click on the thumbs-down icon, Pandora is effusively apologetic. “Sorry about that,” it tells you. “We’ll never play it again on this station.” The pop-up menu also includes links to purchase the music from iTunes or Amazon.com.

On the phone, Tim Westergren is as polite as his interface. He sounds like a relic from the days before Silicon Valley’s bubble burst: aarnest, and completely convinced that his website will change the world. He got the idea for Pandora through his own thwarted music career, and sees it primarily as a tool for introducing listeners to new independent artists. “I worked for ten years after college trying to become a rock star,” Westergren said. “Having had firsthand experience with the challenges independent artists face, this website can be the salvation for them…I’m really hoping we can start a sea change in what it’s like to be a musician. “

When I first heard about Pandora, it sounded like a Godsend: A free Internet radio station that introduces me to new music that sounds like the music I already love? Sign me up! After having mixed results with the Pandora stations I created (my boyfriend and I typed in Interpol and ended up with a Hoobastank song, much to our collective chagrin), I decided to email friends and colleagues so that they could test-drive Pandora and I could get a better sense of how well the music genome project accomplishes its mission to “capture the essence of music at the most fundamental level,” as Westergren says in a letter on Pandora.com.

Andrea Neustein — manager of the band MGMT, the purveyor of shitparade.com, and a contributor to Flavorpill — had a mostly positive reaction towards Pandora. “Quite a few recommendations were new to me, which was a surprise,” she said. “I work with a music distribution company and eMusic, so I’m constantly exposed to new and obscure music. A lot of grime I had never seen before, some bands, and then some remixes I had never heard.” She also noted, “It’s pretty awesome for a radio station to apologize for bad taste.”

The editors here at Spin were, in general, less impressed than Andrea was. Senior Associate Editor Jon Dolan thought that Pandora’s criteria, which appeared unspecific enough to allow broad swaths of comparison, seemed more constrictive of black music. “I put in ‘Wig Wam Bam’ by the Sweet and pretty soon I was listening to something I’d never heard of that sounded like Rick Astley collabing with David Gray,” Dolan reported. “But whenever I put in an R&B or soul artist it proceeds almost too predictably…does this somehow assume that black music is less genomically expansive to move beyond certain, uh, ghettos?”

Racial conspiracies aside, how did those outside the indie rock stronghold of New York City respond to Pandora? My older brother, Jacob, who’s working on his PhD in physics and is a trained musician, was less than thrilled with the responses Pandora churned out. “I entered in the Red Hot Chili Peppers (since they are one of the few bands of the last 15 years that I like), and, of the six or so songs they played that were not by the Chilis themselves, they ranged from mediocre to gawd-awful,” he wrote. “I think they would need additional categories such as ‘lyrics were not written at the level of 16-year-olds’ or ‘drums were not played by a drum-machine’ to have a chance at being effective.”

When I asked Jacob about Pandora’s genome designations, he was similarly unconvinced by Pandora’s efficacy. “I feel that fundamentally they miss the essence of the Peppers,” Jacob wrote. “They focus too much on superficialities (vocal harmony, mixed acoustic and electric instrumentation), get some things completely wrong (a vocal-centric aesthetic), and don’t include many aspects which are vital to why I like them (funk influence, virtuosic performers, intelligent lyrics).”

Tim Westergren refutes my brother’s criticism about the genomes on the site being off-point, because he says the phrases that appear on the interface (phrases Jacob referred to like “vocal harmony, mixed acoustic and electric instrumentation”) are not genomes — they’re “focus traits — complilations of genes.” And even with some negative feedback, Westergren is still thrilled with the project and sees it continuing to evolve. “This is the first time in all these years,” Westergren says, “I’m really starting to think that the original vision for this could come true. We could change in a meaningful way the basic structure of the music business. I hope that [in the future] being a musician will be like being a teacher or being a business man or whatever — you’ll find your 50,000 fans, and you don’t need the million dollar advance from the label to find your audience.”

And if that happens? “Then I can go back to being a rock star,” Westergren chuckles.