A Survey on Musical Feature Extraction and Classification Methods

Identifying musical instrument is challenging task because of its multidimensional nature. Every particular instruments have its own characteristics and physical features like energy feature, rhythm feature, temporal feature, spectrum feature, harmony feature etc. In this paper toolboxes that are all publically available for extracting these features. & the perception of valance and arousal has been also discussed. This paper offers an overview of the set of upper features. Particular attention has been paid to design of a syntax that offers both simplicity of use & transparent addictiveness to a multiplicity of possible input also the same syntax can be used for analysis of signal audio files, batch files, series of audio segments multi banned signals. Also we have studied about the preprocessing and automatic speech recognition the preprocessing is done and voice speech is detected based on energy and zero crossing rates.