Advanced acoustic analysis of historical theaters: empirical evaluation and predictive methodologies.

This dissertation investigates the acoustic properties of historical theaters, focusing on ancient open-air theaters and 24 Italian indoor theaters. By integrating advanced measurement techniques with data-driven modeling, key acoustic parameters—including reverberation time (RT), early decay time (EDT), clarity (C80, C50), definition (D50), sound strength (G), and interaural cross-correlation (IACC)—are analyzed across diverse architectural spaces. The study begins with an analysis of ancient Greek and Roman theaters, examining how structural and geographical features influence sound propagation. It then shifts to Italian indoor theaters, comparing their acoustic behavior with other iconic European theaters from the Baroque and Classical periods. Measurements in both occupied and unoccupied conditions reveal the impact of architectural design, materials, and environmental factors on acoustic quality. A key contribution is the application of machine learning, particularly Support Vector Regression (SVR), to enhance acoustic parameter predictions. The study demonstrates how stable parameters like RT and EDT can predict more variable ones such as C80 and D50, reducing errors in traditional averaging methods. This research advances theater acoustics by combining empirical measurement with predictive modeling, offering new tools for analyzing, restoring, and optimizing historical theaters while improving efficiency in scenarios where extensive on-site measurements are impractical.