Microphone wind noise reduction using singular spectrum analysis techniques

Eldwaik, O; Li, FF

Abstract

Wind noise is a known problem that contaminates microphone signals in many field measurement and audio recording scenarios. A recently completed EPSRC project has led to the development of a tool to detect such noise; this paper takes a step further by proposing the use of singular spectrum analysis (SSA) techniques to reduce such noise. One of the advantages of the SSA method is that it has the potential to retain wanted signals with less distortions when compared with other known signal processing techniques for wind noise reduction. The SSA decomposes signals in eigen-spaces, selects and groups the principal components according to their contributions and eventually reconstructs the wanted components back to the time domain. Following a brief review of the wind noise problem and existing solutions, this paper outlines the principle of SSA method, discusses group techniques used in the SSA procedures for wind noise removal, and presents the results.

Citation

Eldwaik, O., & Li, F. (2017, November). Microphone wind noise reduction using singular spectrum analysis techniques. Presented at 33nd Annual Conference And Exhibition. Reproduced Sound 2017 : Sound Quality By Design, Nottingham, UK

• Presentation Conference Type: Other
• Conference Name: 33nd Annual Conference And Exhibition. Reproduced Sound 2017 : Sound Quality By Design
• Conference Location: Nottingham, UK
• Start Date: Nov 21, 2017
• End Date: Nov 23, 2017
• Acceptance Date: Oct 16, 2017
• Online Publication Date: Nov 21, 2017
• Publication Date: Nov 21, 2017
• Deposit Date: Jan 2, 2018
• Publicly Available Date: Jan 2, 2018
• Publisher: Institute of Acoustics
• Related Public URLs: https://ioa.org.uk/civicrm/event/info?reset=1&id=256
• Additional Information: Additional Information : Proceedings ISBN: 978-1-906913-28-1; Event Type : Conference