Percussa micro super signal processor

Here’s a brief video snippet I discovered someone took at a talk I did at this year’s South by Southwest, with interaction design pioneer Joy Mountford (formerly Yahoo, Apple). We were talking about the idea of “data as art”, which happened to coincide neatly with the Design and the Elastic Mind show at MOMA, featuring several works from Joy’s recently-disbanded Design Innovation Group team at Yahoo.

The audience response to the work Joy showed was really overwhelming, as search activity danced around the globe and photos came to life in three dimensions. And it was nice to be able to show them the tool used to create these projects, Processing, and encourage people to try it out for free, even if they hadn’t tried programming before.

But I was surprised by how people reacted to a quick musical demo I closed with. Using Java, I wrote a simple program that checked my Gmail account using IMAP, then translated the time spam messages arrived into MIDI notes. I’m still developing a more advanced real-time version, so I threw the resulting SMF file into Ableton Live.

I’ll actualy be showing a newer version of this for Internet Week at an event sponsored by Make Magazine; more on that in a few days. (I’ll also use that as an opportunity to post some updated code.)

We spend so much time talking about how visualization can make data more expressive that we sometimes overlook other media. The spam “musicification” made sense to people partly because even the untrained ear is sensitive to musical timing, I think. Sonification of data isn’t always the right choice; the results can be abstract, though perhaps there’s value in that, too. But it’s worth remembering that people are sensitive to sound as they are to visuals. Since it’s not an either/or choice, necessarily, it’s too bad that so often designers neglect aspects of sound and timing while focusing only on what something looks like. It’s a challenge, certainly — there’s a reason most of us mute annoying sound feedback on computer interfaces — but I think it’s an area in which we’ll see a lot more discussion.

Now, data in smell-o-vision — that’s a story for another day.