Sunday, November 30, 2008
Rocky Top is little different from filtered noise?
After a detailed study, I have revealed what many have suspected for years - that Rocky Top is little different from noise. See the figure below for the damning proof.
|
Background: I needed to create a randomly generated "audio" source for a project. Spectrally, audio can be loosely approximated as filtered white noise (well actually, the PSD of just about any signal can be be approximated as filtered white noise, though the filter complexity changes).
As I was putting together this routine during the UT - UK game last night, I decided to use a wav of rockytop to verify my filter design (a first order IIR with the output resampled from 8 kHz to 30 kHz). And then I was struck by how else the graph could be taken and decided to poke a little fun at myself.
|
