Turkish Music Education: An Artificial Intelligence Based Performance Analysis Design

An Artificial Intelligence Based Musical Performance

Authors

Abstract

The purpose of this study is to develop a software system, using the Python programming language, capable of analyzing the pitch and usul (rhythmic pattern) structures in performances of Turkish music. To achieve this goal, the research was conducted within the framework of the Design and Development Research Model (DDRM) and was completed in three stages. In the first stage, a comprehensive literature review was carried out to identify relevant publications that could provide insight into the subject; both domestic and international written sources, as well as existing applications, were examined in detail. In the second stage, based on the main objective, the needs were identified, and a software design was created to perform pitch and usul analysis on performance recordings of Turkish music. The design, developed through the iterative cycle of DDRM, analyzes frequencies using Python libraries and custom algorithms, while storing and visualizing the results through a Firebase based system. Users can track analysis results across specific time frames and access detailed visual data for each analysis, including spectral representations, error curves, and color coded notation. In the third stage, the software design was tested through an application process based on the analysis of performance recordings. For this purpose, a study group consisting of performers and experts was formed. The performer group included two students with basic Turkish music training who could play instruments (Ud and Ney) at a beginner level, while the expert group consisted of three academics specialized in Turkish classical music and Turkish folk music. Before the application process, the students performed an etude prepared in consultation with the researcher and expert group: a 56 pitch study in the Bayati maqam and in Sofyan usul. The recorded performances were evaluated in terms of pitch and usul elements by both the expert group and the software. Subsequently, the analyses of the expert group and the software were comparatively examined, and the results were discussed. It was determined that the software could technically analyze performance recordings with a higher degree of precision than anticipated for pitch and usul elements, achieving a high level of consistency with the expert evaluations. The research was then documented and reported.

Downloads

Published

2025-09-01