Automatic Chord Recognition Technique for a Music Visualizer Application

Stage lighting is a powerful tool to draw the audience into the music engagement in any music or concert performances. The traditional stage lighting control often requires professional operators and lots of time-consuming preparations. This paper presents an automatic chord recognition technique for a music visualizer application. The proposed technique is based on chroma vector and chord template fitting algorithms. The detected chord can apply to control the proper light color on stage for audience best experience. The experiments done by utilizing the GarageBand software application to mimic a variety of musical instruments, a microchip microcontroller performs signal processing function, and the detected chord displays on a LCD. Experiment results have indicated that our proposed algorithm can accurately detect chords for all instruments and in both major and minor chords. This algorithm presents a simple and ease of use design since it does not require complicate machine learning or statistical process techniques. However, there are few detection errors appear in complex chords due to hardware limitation deployed in the experiment and the algorithm needs an improvement.