Digital Journal — So much music, so little time. It’s gotten to the point where you want what you want, whenever you want it and now, you can actually get it.
With music recommendation services — the next generation of Internet radio — the world has become your iPod. These sites suggest tunes based on music you already enjoy. Here’s a breakdown of the top online services that would like to rock your particular world.
The Product: Pandora (www.pandora.com)
The Pitch: “When was the last time you fell in love with a new artist or song?”
The Skinny: Of all the music recommendation services, Pandora might be the most promising because of its extremely clean, accessible interface. But Pandora also makes for some great listening, if you’re willing to train it a little. Pandora is entirely Flash-based, so you simply access it from your browser. It loads, asks you to name a song or artist, and shebang! — your customized radio station is born. Pandora is part of the Music Genome Project, an attempt to attach roughly 400 different attributes to more than 300,000 songs, based on the dispassionate analysis of a cadre of music geeks. As with any of these services, you’re bound to bump into some stinkers. But if you give any song the “thumbs down,” it shame-facedly promises to never play something like that again, then tweaks the algorithm to account for your reaction. Also, it’s cool to see how they came up with their selections, even when they’re off. For instance, the Red Hot Chili Peppers’ version of “Higher Ground” produced a playlist beginning with New Order’s “60 Miles an Hour.” Why? Both have “subtle use of vocal harmony, mild rhythmic syncopation, minor key tonality, electric rhythm guitars and a dynamic male vocalist.” Sometimes Pandora works, sometimes it fails, and sometimes it works because it’s so leftfield. According to founder and music lover Tim Westergren, that’s the point.
The Product: Last.fm (www.last.fm)
The Pitch: “It’s fun, it’s free, it’s all about the music”
The Skinny: This truly is the most fun you can have discovering new music. Last.fm’s serves up its recommendations with a healthy community vibe. Instead of breaking songs down to their core components, as with Pandora, Last.fm relies on its extensive user base to provide “collaborative filtering.” In other words, if you’re into x, and your buddy is into x, you both might be into y. Last.fm is built on the Audioscrobbler plugin, which scans your hard drive for music and creates a profile based on your habits. From there, it can turn your favourite MP3 player into a personalized radio. The site itself can be a bit slow and tricky to navigate (many just use Google to search through it), but its content is invaluable. You can determine the most popular songs by any artist (the top Elvis Costello song in this hip community is, justifiably, “Oliver’s Army”) and meet your musical neighbours. Its user-reliant schema is bound to create some problems. Metatags can be jumbled, misspelled or just wrong. Then there’s the Dave Matthews Syndrome — that dude ends up on just about every playlist, right next to Radiohead. But the system could become much smarter and stronger in the future. Unless, of course, it gets much worse.
The Product: Tapestry (www.allmediaguide.com/tapestry)
The Pitch: “Makes ‘my music’ mean so much more”
The Skinny: The new kid on the block — so new, in fact, that it’s not quite out yet. But it has incredible potential. Brought to you by the music junkies at the All Music Guide (allmusic.com — an essential destination for anyone who’s ever gave a damn about tunes), Tapestry is another human-based recommendation system. This one uses AMG’s extensive data to profile songs with over 6,200 detailed descriptors, such as styles, themes, tempo and even individual instruments. Currently, only a demo version is available, and it seems AMG is keen on snagging business-to-business customers at first. But the concept rocks. Combine any musical style (say, “Gangsta Rap”) with any mood or theme (such as “Nihilistic”) and Tapestry will immediately generate a smart, accurate playlist that’s sure to upset any faith-based group.
The Product: Launchcast (http://music.yahoo.com/launchcast)
The Pitch: “Music that listens to you”
The Skinny: I hear great things about this veteran offering. Its adherents swear by it, and some won’t even bother checking out the newbies. But since I’m a Firefox-loving Canadian, this was enough to make Launchcast essentially snub me. It was exceedingly difficult to give this service a fair shake. Oh well.
The Winner:
Pandora, because its technology is so smart. If Pandora can find a way to incorporate Last.fm-style community recommendations into its algorithm, this service could very well render a DJ’s job obsolete.
This article is part of Digital Journal’s national magazine edition. Pick up your copy of Digital Journal in bookstores across Canada and the United States! Or subscribe to Digital Journal now, and receive 8 issues for $29.95 GST ($48.95 USD)!
