Co-integration of acoustic simulation software and GIS for speech intelligibility analysis in complex multi-source acoustic environments. Application to Toledo's Cathedral

Abstract

The speech intelligibility in complex multi-source acoustic environments depends on a variety of factors such as speech level, background noise level, reverberation time, as well as psychoacoustic effects. Since these factors can change for each source-receiver combination, in order to integrate information from various sources, efficient tools and techniques are required to determine the speech intelligibility at every listener position. In drawing up this study, two types of tools are used: (i) an acoustic simulation software (ODEON) and (ii) spatial analysis tools of a Geographic Information System (ArcGIS). To determine the speech intelligibility, the Speech Transmission Index (STI) has been calculated. Sound Pressure Level and Reverberation Time have been calculated at points of a grid at intervals of 1 meter. An automated workflow using ArcGis Modelbuilder has been created in order to obtain STI. Also, Arcmap has been used to represent and analyze the results over the complex geometry of the space. The proposed procedure has been calibrated using information obtained in the Toledo Cathedral (Spain) and the results for acoustic situations related to different celebrations of the liturgy are presented.