Geometrical parameter combinations that correlate with early interaural cross-correlation coefficients in a...

Area of Science:

• Acoustics
• Architectural Acoustics
• Psychoacoustics

Background:

• Accurate prediction of acoustic parameters like early interaural cross-correlation coefficients (IACCs) is crucial for performance hall design.
• Previous methods for predicting IACCs, including neural networks, required significant measurement effort and parameters.
• Re-analysis of existing binaural data provides an opportunity to develop more efficient prediction frameworks.

Purpose of the Study:

• To establish a framework for predicting early interaural cross-correlation coefficients (IACCs) with minimal measurement effort and parameters.
• To improve the accuracy of IACC predictions compared to previous methods.
• To investigate the influence of geometrical parameters on IACC prediction.

Main Methods:

• Re-analysis of previously collected binaural data from two multi-purpose performance halls.
• Development and application of regression models using linear combinations of polynomials of geometrical parameters.
• Comparison of regression model predictions with results from a previous neural network approach.

Main Results:

• Regression models, combined with specific measurement schemes, sufficiently predict IACCs within engineering tolerance.
• The developed regression models outperform the authors' previous neural network predictions for IACCs.
• The relative importance of various geometrical parameters in predicting IACCs was successfully investigated.

Conclusions:

• A regression-based framework offers an efficient and accurate method for predicting early IACCs in performance halls.
• Geometrical parameters play a significant role in IACC prediction, enabling simpler and more effective acoustic modeling.
• This approach reduces the measurement effort required for acoustic analysis in architectural design.