hello Ozone,I have spent the last couple of weeks expanding, updating and redesigning a very complete and seamless sound-analysis patch.This patch can extract many features from which the following would be of significance for ozone:

• spectral flatness (noisiness/peakiness).. useful for detecting specific soundevents within/from a soundmass
• tempo, beat/rhythm detection… tempo spectrum is very exciting (lots of micro tempos within signal are detected simultaneously)
• event detection (onset of pitched/unpitched events)… pitched onset detection is useful for detecting human voice and not detecting for example footsteps
• auditory-based loudness and spectral energy
• etc

all of these parameters and detections perform dynamic self calibration and juicy outputs between 0.0-1.0 are ready to be feed into Ozone.Besides analyzing environmental sounds, it would be interesting to have the ozone system listen to its own output and readjust itself and therefore from a non-linear dynamic system (cheap implementation of autopoesis.)