The subject of thesis „Application supporting learning to play the instruments by analysis and visualisation of musical signals” is an application which may be found helpful in improving such skills as sense of rhythm and playing the musical instrument, especially percussion. The application, written in Python, contains a sound recorder and a simple metronome – a kit of tools commonly used by musicians of any level of advancement.

The main element of the application is an analyzer which allows the user to verify their recording in terms of playing the musical phrase in the correct, even tempo. It detects the moments in the recording when new notes have been played. The algorithm consists in performing a short-time Fourier transform to create a spectrogram of the audio data and compare its content to the content of a reference recording.

The following chapters include both mathematical and programming description of this algorithm of music processing.