This study presents an automatic transcription system for Turkish music for the first time in literature. We first discuss the characteristics of Turkish music that are taken into consideration in the design of the system. Then, the following signal processing components of the system are described briefly in relation to each other and explaining their function in the system: f0 estimation, automatic tonic detection and makam recognition based on pitch distributions, frequency and duration quantization.
Published in:
Signal Processing and Communications Applications (SIU), 2011 IEEE 19th Conference on
Date of Conference: 20-22 April 2011