Copyright © The Echo Nest Corporation, 2005-2008

Echo Nest Technology

Unlike other automatic methods of acoustic feature extraction, which progress strictly on raw audio signal, the Echo Nest audio analyzer models how humans perceive music through proprietary machine listening techniques.

Based on fundamentals of psychoacoustics, music perception, and adaptive learning, the analyzer simulates the auditory path and cognitive process of the human hearing system (see Figure).

Powered by greatly optimized computation modules and data coding, the API outputs a compact XML-formatted perceptual description of the structural content of the music signal — computation takes about 1/50th of the duration of the track and the text output is about 1/20th of the size of its corresponding mp3.

audio analysis for new music applications