Absolute Pitch: Automatic Sheet Music Navigation Based on Real-Time Music Recognition

Music has been the universal language of the world for centuries. However, for all these years, musicians have faced a challenge that remains unresolved: the task of turning pages while playing their instruments. The sound of page turning during performances often disturbs the music and can cause abrupt pauses, and on some occasions, pages can fall out, ruining entire performances. The goal of this research is to solve this issue and to develop an app that digitizes sheet music and automatically adjusts the displayed music at the rate of the musician's playing. Using the user's MusicXML file, the music is displayed on the device with the Android graphics API. After activating the microphone, the algorithm uses the properties of sound to analyze the audio signal, employing tools like the Fourier Transform. By converting audio signals into frequency components, the algorithm identifies harmonics in order to calculate the dominant pitch. Instead of just computing the frequency with the highest intensity, we calculate the average power of each note across all octaves to find the fundamental frequency. Once the pitch of the note is found, the app scrolls through the music and adjusts to the performer's playing. After testing the algorithm on different examples of sheet music, we find that the program is accurate in detecting pitch and displaying music in real-time.