2aAA11. Relationships between speech intelligibility and objective acoustical parameters for architectural features in Catholic churches.

Session: Tuesday Morning, June 17


Author: Antonio P. Carvalho
Location: Acoust. Lab., Dept. of Civil Eng., College of Eng., Univ. of Porto, R. Bragas, P-4099 Porto Codex, Portugal, carvalho@fe.up.pt
Author: Antonio E. Morgado
Location: Acoust. Lab., Dept. of Civil Eng., College of Eng., Univ. of Porto, R. Bragas, P-4099 Porto Codex, Portugal, carvalho@fe.up.pt
Author: Luis Henrique
Location: Polytech. Inst. of Porto, P-4000 Porto, Portugal

Abstract:

This study reports on subjective and objective acoustical field measurements made in a major survey of 36 Roman Catholic churches in Portugal built in the last 14 centuries. Monaural acoustical measurements (RT, EDT, C80, D50, TS, L, and RASTI) were taken at several source/receiver locations in each church. A group of college students was asked to judge the intelligibility of speech by evaluating live speech at similar locations in each room. This paper complements those presented at the 1996 Indianapolis and Honolulu ASA Meetings and concentrates exclusively on the relationships of the speech intelligibility averaged values with the objective room acoustics measures and with some architectural features of the churches. The averaged results by church are graphed and analyzedby comparisons. Correlation analyses and statistical modeling identified some relationships among the measures. For instance, squared cor-relation coefficients (R[sup 2]) of 0.67 were found for the relationships: SPEECH-RT AND SPEECH-TS. Between SPEECH and RASTI only a maximum R[sup 2] of 0.50 was found. Regarding the churches' architectural features, the maximum R[sup 2] found was 0.52 between SPEECH and NAVE HEIGHT. A general linear model including several architectural features increased the R[sup 2] to 0.72. [Work supported by FEUP and ESMAE/IPP--Portugal.] [See NOISE-CON Proceedings for full paper.]


ASA 133rd meeting - Penn State, June 1997