Finding the Rhythm

Shortly after it launched a few years ago, I signed-up for Pandora. What’s not to like about a service that purports to know more about the music you like than you do?

The concept is actually pretty cool (on paper). Using their “Music Genome Project,” individual songs are analyzed by a musician for a specific list of attributes, such that each song attains its own “gene” that embodies all those attributes in a coded, computer-digestible manner. Algorithms can then be used to feed you, the listener, music that is “genetically compatible” with a specific song or songs that you tell Pandora when you sign-up for an account.

The problem is—it just doesn’t work very well, at least not for dance music.

I’m uncertain whether it’s an issue of the Music Genome Project not having enough of the right attributes to properly and thoroughly assess dance and electronia, or if it’s an issue of Pandora, as a service, having sufficient inventory of dance music that’s been analyzed.

But when I tried Pandora a few years ago, I found that the music it provided me was one dimensional (I don’t think I heard a single male vocalist), old (many tracks are quite old), and… This is hard to articulate… Very narrow in sound, pretty much all 135 BPM or higher which is where most trance and techno tends to sit. While it wasn’t all trance and techno, it was all vocal (probably “right” for me), and had “inspiration” from techno and/or trance, let’s say. Artist examples include Cascada (LOTS of Casada; ugh) and Ian Van Dahl; there’s nothing wrong with the music, it’s just that I don’t want to hear only that type of dance music, all, day, long…

The other problem is that the mixes programmed struck me as very odd choices, and were often quite incompatible with one another. While I understand it’s not a non-stop mixshow, if Pandora is supposed to be genius about song picks, it should probably know better than to pair a dark, edgy, harsh mix of one song next to a more mainstream, pop-friendly mix of another. You don’t see goths and jocks hanging out with each other at the average high school, and it doesn’t work any better with dance music.

So, earlier this week, I received an e-mail from Pandora announcing that they had just released a mobile version for BlackBerry. (Why should iPhone have all the fun?) It had been months since I’d logged into Pandora, so I figured I’d check it out anew and see if anything had improved. I promptly downloaded the application, logged-in, and listened for a bit while driving around yesterday.

Let’s just say that it’s the same old Pandora that I found wanting in so many ways the first time I used it. Nothing has changed. It still sounds like somebody’s marginal, and overly small dance music collection stuffed into an iPod and set to shuffle. You’ve heard it all before a hundred times. Your friends have heard it a hundred times when you’ve provided music for the party. Someone desperately needs to hand you an iTunes gift card so you can get something good. Something new. And hurry. Please.

My nephew Ted is a trance-head. We talk about music all the time, and I asked him if he’d ever tried it. Let me just cut to the chase and say his impression of Pandora was the same as mine… Too narrow, too old, too uninspired. Neither Ted, nor I, listen to dance music this way, or for this reason.

I’ve often wondered if there can’t be something like a Music Genome Project for dance music so I can objectively assess one song vs. another; know for sure that this track is “tribal house” and not “progressive house,” for example,and be able to do better, more compatible choices when mixing a set. But alas, at least with dance music, I’m honestly thinking it’s best to leave the analysis and choosing to the humans, whether that’s yourself (and what you put on your iPod), or the PD’s or DJ’s at radio stations (who certainly aren’t perfect either, but at least have access to new music).