![Open Access](https://i-scholar.in/lib/pkp/templates/images/icons/fulltextgreen.png)
![Restricted Access](https://i-scholar.in/lib/pkp/templates/images/icons/fulltextred.png)
![Open Access](https://i-scholar.in/lib/pkp/templates/images/icons/fulltextgreen.png)
![Open Access](https://i-scholar.in/lib/pkp/templates/images/icons/fulltext_open_medium.gif)
![Restricted Access](https://i-scholar.in/lib/pkp/templates/images/icons/fulltextred.png)
![Restricted Access](https://i-scholar.in/lib/pkp/templates/images/icons/fulltext_restricted_medium.gif)
Automatic Identification of Tabla Tempo and Transcription of Bols
Subscribe/Renew Journal
Automatically extracting music information is gaining importance as a way to structure and organize the increasingly large number of music files digitally available and has become an important part of multimedia research. This research becomes more interesting as well as challenging when the music analysis is done according to the instruments used in the making of music. One such very popular percussion instrument called Tabla widely used in accompanying Indian Classical music recitals is analyzed based on the collection of digital recordings. The database consists of popular taals commonly used in Indian classical music. Bols are the basic notes of Tabla. A taal is a predefined sequence of Tabla bols. Different such bol arrangements give rise to various taals. Based on the speed of repetition of bols, the taals are broadly classified into low (Vilambit), medium (Madhya) and fast (Drut) tempo. Thus a tempo represents the rhythmic information of a taal.
In this paper, an automatic system for identifying and transcribing Tabla bols of different tempos is explored. The transcription process is based on three main steps: firstly tempo of the audio clip is identified using autocorrelation technique. Secondly, the recorded clip is segmented where each segment represents a bol. In the third step, Mel-Frequency Cepstral Coefficients (MFCC) features are extracted from the separated bols to form templates for pattern classification. Two pattern classification techniques namely Dynamic Time Warping (DTW) and Vector Quantization (VQ) are analyzed and compared for evaluating the performance of bol identification. Overall bol identification accuracy of the system for tempo independent classification is 96.18%.
In this paper, an automatic system for identifying and transcribing Tabla bols of different tempos is explored. The transcription process is based on three main steps: firstly tempo of the audio clip is identified using autocorrelation technique. Secondly, the recorded clip is segmented where each segment represents a bol. In the third step, Mel-Frequency Cepstral Coefficients (MFCC) features are extracted from the separated bols to form templates for pattern classification. Two pattern classification techniques namely Dynamic Time Warping (DTW) and Vector Quantization (VQ) are analyzed and compared for evaluating the performance of bol identification. Overall bol identification accuracy of the system for tempo independent classification is 96.18%.
Keywords
Autocorrelation, Bol Segmentation, Dynamic Time Warping, Feature Extraction, Onset Detection, Vector Quantization.
User
Subscription
Login to verify subscription
Font Size
Information
![](https://i-scholar.in/public/site/images/abstractview.png)
Abstract Views: 228
![](https://i-scholar.in/public/site/images/pdfview.png)
PDF Views: 4