In the field of music, beam notation is the standard for musical notation, albeit difficult to understand. It is easier for people to read numbered musical notation rather than beam notation. A song that already has a notation is easy to be arranged, but many songs are made spontaneously without notation. Therefore, a computation system is developed to automatically generate the numbered musical notation of a song. The system is using the Discrete Fourier Transform algorithm to convert the audio signal from time domain to frequency domain, without implementing noise filtering beforehand. Tones perceived from the digital sound will be converted into numbered musical notation. We have tested our system in two mode based on the input, i.e. human voice and music keyboard device. In general, our system can detect about 53% precisely for human voice, and 66% precisely for music keyboard device.
Copyrights © 2013