During the Boston Music Hack Day, 30 or 40 music hacks were produced. One phenomenal hack was Rob Ochshorn’s Outlier FM. Rob’s goal for the weekend was to utilize technology to search for beauty and surprise in even the most overproduced popular music. He approached this problem by searching musical content for the audio that “exists outside of a song’s constructed and statistical conventions”.
With Outlier FM rob can deconstruct a song into musical atoms, filter away the most common elements, leaving behind the non-conformist bits of music. This yields strange, unpredictable minimal techno-sounding music.
So how does it work? Well, first Outlier FM uses the Echo Nest analyzer to break a song down into the smallest segments. You can then visualize these segments using numerous filters and layout schemes to give you an idea of what the unusual audio segments are:
Next, you can filter out clusters of self-similar segments, leaving just the outliers:
Finally you can order, visualize and render that segments to yield interesting music:
Here’s an example of Outlier FM applied to Here’ Comes the Sun:
Rob’s hack was an amazing weekend effort, he combined music analysis and visualization into a tool that can be used to make interesting sounds. Outlier.fm was voted the best hack for the music hack day weekend. Rob chose as his prize the Sun Ultra 24 workstation with flat panel display donated by Sun Microsystems Startup Essentials. Here’s Rob receiving his prize from Sun.
Congrats to Rob for a well done hack!
Music Machinery 