Audio Key Finding: Considerations in System Design and Case Studies on Chopin's 24 Preludes
- PDF / 1,484,324 Bytes
- 15 Pages / 600.03 x 792 pts Page_size
- 75 Downloads / 153 Views
Research Article Audio Key Finding: Considerations in System Design and Case Studies on Chopin’s 24 Preludes Ching-Hua Chuan1 and Elaine Chew2 1 Integrated
Media Systems Center, Department of Computer Science, USC Viterbi School of Engineering, University of Southern California, Los Angeles, CA 90089-0781, USA 2 Integrated Media Systems Center, Epstein Department of Industrial and Systems Engineering, USC Viterbi School of Engineering, University of Southern California, Los Angeles, CA 90089-0193, USA Received 8 December 2005; Revised 31 May 2006; Accepted 22 June 2006 Recommended by George Tzanetakis We systematically analyze audio key finding to determine factors important to system design, and the selection and evaluation of solutions. First, we present a basic system, fuzzy analysis spiral array center of effect generator algorithm, with three key determination policies: nearest-neighbor (NN), relative distance (RD), and average distance (AD). AD achieved a 79% accuracy rate in an evaluation on 410 classical pieces, more than 8% higher RD and NN. We show why audio key finding sometimes outperforms symbolic key finding. We next propose three extensions to the basic key finding system—the modified spiral array (mSA), fundamental frequency identification (F0), and post-weight balancing (PWB)—to improve performance, with evaluations using Chopin’s Preludes (Romantic repertoire was the most challenging). F0 provided the greatest improvement in the first 8 seconds, while mSA gave the best performance after 8 seconds. Case studies examine when all systems were correct, or all incorrect. Copyright © 2007 C.-H. Chuan and E. Chew. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
1.
INTRODUCTION
Our goal in this paper is to present a systematic analysis of audio key finding in order to determine the factors important to system design, and to explore the strategies for selecting and evaluating solutions. In this paper we present a basic audio key-finding system, the fuzzy analysis technique with the spiral array center of effect generator (CEG) algorithm [1, 2], also known as FACEG, first proposed in [3]. We propose three different policies, the nearest-neighbor (NN), the relative distance (RD), and the average distance (AD) policies, for key determination. Based on the evaluation of the basic system (FACEG), we provide three extensions at different stages of the system, the modified spiral array (mSA) model, fundamental frequency identification (F0), and post-weight balancing (PWB). Each extension is designed to improve the system from different aspects. Specifically, the modified spiral array model is built with the frequency features of audio, the fundamental frequency identification scheme emphasizes the bass line of the piece, and the post-weight balancing uses the knowledge of music theory to adjust the pitch-class distribution. In particular, we consider
Data Loading...
