Data Auralization - Listening to the Dow

I was pleased to see this posted in the Alumni Notes of my Alma Mater recently---Justin Joque, a Data Librarian at the University of Michigan, put together a Data "Auralization" of the Dow from 1928 to 2011.  He uses tick data to create two audio bands, one with closing price, and one with trade volume.  Here's the Vimeo Link:

I've done a couple of experiments with turning stock data into sound before, and met with similar problems---it's exceedingly dull (as he points out, until we hit higher frequency trading), and generally a slower input medium than visualization.  That is, a glance at a line graph conveys as much information much more quickly.  However, there's something very cool about this and auralization techniques in general, namely that it has an ambient quality.  Yes, you can take visual data in faster, but you can take audio data in passively, maybe most usefully, when your eyes are busy working on something else.  And there's something else he captures very effectively: you feel the buzz of the market quite literally here---something that a vol chart can't convey so viscerally.  While visual data can be useful as a distanced summary, this technique might convey more of the phenomenology of the market. 

As returns have periodic tendencies---sound transformations have such a natural applicability.  And whether or not we're actually putting them to music, it's definitely the guys in signal processing who are making the big advances.  As Kalman filters and HMM have entered into modern econometrics, moonlighting on projects like this is going to be increasingly insightful for both what we hear, and what we see in finance.