<%@ page language="java" contentType="text/html" %> <%-- Include common initialisation code --%> <%@ include file="/arch/common.jsp" %> <%-- The current tab --%> <% String currentTab = "Research"; %> <%-- Content of navigation pane --%> <%@ include file="nav.jsp" %> <% showCurrentLink=true; %> <%-- Current navigation location --%> <% String currentNav = "Reports and Theses"; %> <%-- Include the code for the document header --%> <%@ include file="/arch/header.jsp" %>

Research Report CS-RR-231

<%-- Include the code for the lines and navigation --%> <%@ include file="/arch/middle.jsp" %>

Hugh Scott and R.G. Wilson, A Comparison of Filters For Audio Signal Segmentation in Audio Restoration (October 1, 1992).

Abstract

This report is concerned with the choice of filters for the estimation and detection from a time-frequency representation of audio signals of partials - the approximately sinusoidal components, typically generated in harmonic series, of which many musical signals are composed. The report contains a detailed theoretical and experimental comparison between filters with Gaussian and bilateral exponential time envelopes, in terms of performance on noisy data and simplicity of implementation. It is concluded that the exponential filter is better on both counts. There follows a brief examination of the problem of onset detection, in which both additive and multiplicative approaches are examined. Some tentative conclusions are drawn about the most effective way to solve the problem of partial detection in noisy signals.

<%@ include file="hardcopy.html" %> <%-- Include the code for the document footer --%> <%@ include file="/arch/footer.jsp" %>