Crater EP available on Pandora

Good news for Pandora users today! The Crater EP is in rotation now, which means you can add it to existing stations, create new stations from it, and hopefully hear tracks from it come up when you're in the listening neighborhood of acoustic- and piano-based music.

I've also created a shared station based off of The Crater and Canasta's album, We Were Set Up, and so far it's been a nonstop hit parade of indie piano-influenced bands. You can get to the station via my Facebook profile.

Touching/corny side story: I wrote a sappy cover letter to Pandora when I sent in my CD. It was about how when I was still back in Boston and had little to no clue what I was doing music-wise, I went to one of Pandora's very first town hall meetings at MIT and met one of its founders, Tim, and gave him a very, very handmade demo in a cheap crappy jewel case. I knew it wouldn't get cataloged, but Tim was emphatic about how important it is to get your music out into the world, and he seemed like just a genuinely nice and supportive guy. It's not often you can make a personal connection to a website the way you can with a band or a writer on a book tour, but Tim did exactly that, and I think it's helped make Pandora the relative powerhouse in the internet radio world that it is.

Pandora Part Two

I'm starting with one of the questions I posed in Part One, which is what characteristics of music do most listeners base their musical choices on? Which is clearly a very broad question, so I'd love to get everyone's input on what you, yes you, on what makes you like or dislike music.

More specifically, one of the characteristics that, for me, determines whether or not music has any staying power with me is the lyrics. I pay a lot of attention to the lyrical subtlety and prosody of a song, and even of an entire album, but I'm pretty sure I'm in the minority in this. Anecdotally speaking it seems most people's taste in music can be pegged by the characteristic described by the Music Genome Project -- the approximate 400 attributes each song in Pandora is rated on individually. So what happens to the lyrics in the Genome?

I asked Pandora's Tim about how lyrics were handled, and he acknowledged that they are a weak spot in Pandora, and in the Genome Project in general. Tim had an interesting perspective on which lyrics elements might determine a person's taste in music, based on his explanation of lyrics in the Genome. Commenting on how subjective they are (agreed), Tim said "one person's happy can be another person's sad." And in fact, lyrical "genes" were once included in the Project, but as the Pandora team came to decide on which genes were useful and which ones weren't, lyrics were kicked off the list because they were too difficult to measure objectively.

While I agree that the emotional content or interpretation of lyrics is far too subjective for something as taxonomical as the Genome project to measure, it did seem to imply that emotional content was all there was to lyrics. Another Pandora listener commented alter that evening that it would serve Pandora well to measure lyrics by more literary attributes, for instance, verse/chorus structure, use of a bridge, etc. The Genome does use basic literary traits such as these, but doesn't measure (as of yet) more advanced critical concepts that may actually indicate a lot about a listener's preferences. These concepts could be things like paradoxical language (typical of Modest Mouse), third person narration (typical of Ben Folds), or adherence to a strict rhyme scheme (any showtune you can name).

So here are my questions to you (the royal you): do these more advanced lyrical traits play a part in your musical preferences? Does much of your music collection have lyrical commonalities on this level? If you looked at the lyric sheets alone, would there be more similarities than if you looked at the genres represented by your iPod?

Geeks and Musicians Unite at Pandora Town Hall Meeting

Part One

Over the next several days I'll be posting reports on the Pandora Town Hall Meeting that took place last night at MIT Building 26 (MIT folks: you people need to name your buildings already, what is this, a dystopian futurist sci-fi novel?). While many of you may still be unfamiliar with Pandora, if you're at all interested in music or technology I urge you to at least give it a try and tell me what you think. Instead of simply writing an extended, serialistic promotion of the service, Pandora and the Music Genome Project open up (like a box, see?) a whole range of questions, many of which were addresses last night by one of Pandora's founders, Tim. And while I'd like to think you would just take my word on that, it's only fair to entice you with a preview:

Does digital music distribution give musicians (independent or just indie-label) a leg up against big record labels?
What characteristics of music do most listeners base their musical choices on?
How can small media outlet companies like Pandora remain neutral of the influence of Major Media?

All of which I find fascinating, relevant, and hell, I spent two and a half hours with the rest of the MIT nerdosphere last night -- I might as well make good use of it. Check back soon for the first part.

In the meantime, go to my show tonight.