A Chord Recognition Method of Guitar Sound Using Its Constituent Tone Information

Abstract A simple method for identifying musical chords of audio sound played by a guitar is proposed herein. The chord recognition is achieved by comparing similarities between normalized note numbers extracted from the spectrum of the sound and those generated by a database. The database consists of normalized note numbers corresponding to the fundamental frequency of the chord notes, and the database of the chords with different root is shared. The database of five types of the chords were prepared; Major triad, Minor triad, Major seventh, Minor seventh, and Dominant seventh chords. The name of the chord is determined by the root played. Over 800 sounds of the chords played with electric guitars were tested, and the recognition rate showed good performance, which indicates the feasibility of this method. © 2021 Institute of Electrical Engineers of Japan. Published by Wiley Periodicals LLC.