The Data-Driven DJ and the Dangers of Style

I'm focusing my design towards the sketch of a data-driven percussive instrument that I presented in class a week ago, and in my research I've come across an interesting data-driven musician I'd like to share with you.  He goes by the name the Data-Driven DJ, real name Brian Foo, and is a conceptual artist/programmer.  He has quite a few projects already, each focusing on a different data set, including one on income inequality along a NYC subway line.  Click to watch


His visualizations are beautiful, and well synchronized with the presentation of data.  However, I find that his musical mappings vary from imaginative to not interesting, and though he admits that he has formal musical training and is getting better with each project.  He intentionally restricts himself to a sound world within each project (though) and often relies on musical gestures and sounds for his mappings that together connote a distinct and pre-existing style (popular music, or in the case above, Steve Reich).  While this might make his work aurally accessible, the reliance on pre-existing styles makes his work predictable across the span of each composition, and reduces the drama within each data-set.  E.g. in the subway project above, the sounds signifying areas with high median income aren't really that different from lower median income areas.  If his purpose was to orient the listener towards the sharp contours of the dataset, I think relying on a pre-existing style took away the potential for shock at any particular datapoint.

I think his work will be very useful for prompting questions about the aptness of musical mappings to data, and most importantly for his very generous documentation, which he compiles for every project here.  He uses the programming language CHUCK for mapping audio parameters, and I think I would like to explore how CHUCK might compare to MAX-MSP for my purposes.  To close, I'd like to present my favorite project of his so far, documenting the travel of refugees from a UNHCR dataset (I think the musical mappings are more mature and nuanced in this one) Click to watch