Back to list

Improvement and evaluation of an entropy-based piano tuning algorithm

Authors Menrath, A.
Year 2016
Thesis Type Audio Engineering project
Topic Audio Signal Processing
Keywords psychoacoustics, acoustics, signal processing, Music Information Retrieval
Abstract In 2015 at the University of Würzburg Christoph Wick and Haye Hinrichsen developed the Entropy Piano Tuner, an open-source piano-tuning program, which idea is that the sound of a well-tuned piano has a lower entropy compared to the same piano out of tune. The Entropy Piano Tuner needs a recording of each piano note. Via digital signal processing an algorhytm tries to change the pitches in a way so that the entropy gets lower. The simulation of the pitch changes is based on the assumption that the inharmonicity of a string is independent from its tension. Acoustic measurements should be done to find out how the variation of the string tension during the tuning process affects a strings inharmonicity. It is of interest to evaluate, if the tuning result becomes better if this effect is considered.
Supervisors Sontacchi, A.