Opens in a new tab

AES E-Library

← Back to search

Convention Paper

Perceptually-Informed Features for Room Reverberation Modeling and Prediction

Authors: Kawczinski, Kimberly; Abel, Jonathan S.

AES Convention 149 · Paper 10446 · October 2020

Abstract

A room reverberation model based on normalized echo density and bandwise energy decay is presented. The model produces psychoacoustically transparent and statistically independent impulse response copies by sampling the parameters at various rates. These parameters are further considered as a feature set for prediction within an acoustic space and tested using a standard data set with three measured spaces. The numerical error introduced in the prediction procedure is evaluated, revealing space-dependent trends.

Details

Published in
AES Convention 149
AES Convention
149
Paper number
10446
Publication date
October 6, 2020
Session subject
Audio Processing
Affiliation
Center for Computer Research in Music and Acoustics, Stanford University, CA, USA (See document for exact affiliation information.)
Type
Convention Paper