An Effective Feature Calculation For Analysis & Classification of Indian Musical Instruments Using Timbre Measurement

Abstract

Musical instrument recognition is significant field in the research of computer music which is related to the modelling of sounds. Analysing & synthesing the structure of musical note is of importance both for modelling music signals and their automatic computer-based recognition.Musical sound is produce by five dimensions: pitch, loudness, duration, spatialization, and timbre. First four parameter can be controlled but timbre remains difficult. Timbre then naturally became the main subject of this work. It is important property of sound that separate one music instrument from another and independent of pitch and volume. This work presents a system for identifying a specific musical instrument from monophonic recordings. The system proposed in this paper has been trained and tested with three Indian musical instruments samples. Instruments include flute, harmonium and sitar, which are most commonly used in Indian classical music. The Statistical and spectral parameter are used for the classification of the sounds in Indian Musical Instruments. The SVM classifier proves its ability in automatic and accurate classification of Indian Musical Instrument. Using separately recorded notes as test sets, we were able to achieve average accuracy as high as 88.88 % for SVM that deciding if a note is played by the sitar or others.