Indian Journal of Science and Technology

Abstract

Background/Objectives: Carnatic music is one of the oldest and most historic music systems in the world. Its significance in Indian culture and tradition cannot be overlooked. Over the years, with the growth of media and communication, Carnatic music has sparked the curiosity of thousands of music lovers around the world who wish to learn it. Therefore, it is important to ensure that this style of music is preserved and imparted to all aspiring musicians around the world. One of the most challenging milestones in the field of Carnatic music analysis is raga recognition. Ragas are considered the backbone of Indian Classical music. Each composition in Indian Classical music is structured using ragas. Hence, raga identification is very important to analyze and study compositions. Methods: This paper proposes a raga recognition approach that uses pitch determination, segmentation and a key note mapping technique to identify ragas in a song. Compositions of different ragas are analyzed and features were extracted to arrive at a rule for classification. A total of ten ragas were considered for the study and the results obtained were found to be highly encouraging. Findings/Results: The proposed system was tested with different classification algorithms which produced promising results. This method can be expanded to other ragas which were not considered for this paper. Application: Carnatic raga recognition can assist music aspirers in identifying ragas and also help in classifying songs according to their ragas.

Keywords: Carnatic Classical Music, Indian Classical Music, Raga Recognition