ASA 130th Meeting - St. Louis, MO - 1995 Nov 27 .. Dec 01

4pMU8. Automatic recognition of musical notes.

Jimmy H. Kapadia

John F. Hemdal

Dept. of Elec. Eng. and Comput. Sci., Univ. of Toledo, Toledo, OH 43606

A system for the automatic recognition of musical notes was designed and implemented on a IBM-PC using commonly available hardware. The input to the system was a recorded piece of monophonic music while the output was its musical score. The recognition was carried out in the time as well as the frequency domains. Different characteristics of the note, like its frequency (or its name) and its time value, were found. The approach for this project was primarily from a digital signal processing (DSP) point of view. Various DSP fundamentals, like comb filter analysis and cepstrum analysis, were used for achieving the ends. The emphasis was more on trying out different approaches than on achieving a desirable result. Finally, an attempt was made to recognize a chord.