This paper presents results concerning variations in impulse response functions measured in different rooms. The variation of these functions for changing source and receiver positions and orientation was examined by assessing the change in energies of the direct and reflected signal components. The various source/receiver positions and orientation examined during these tests correspond to typical positions occupied by human speakers in rooms. It was found that the response functions changed drastically with the variation of these recording parameters. The feasibility of inverting (deconvolving) such functions by using digital signal processing techniques was also examined. It was found that when the exact response function was employed, then such an operation was feasible, leading to enhancement and reduction in the reverberant energy present in signals (e.g., speech) recorded under similar conditions. However, when a response recorded at a different position in the same enclosure was employed, then, in general, the inverse filtering operation increased the distortions present in the signals. This indicates that signal dereverberation by impulse response deconvolution has a limited scope in most practical situations.
[1]
John Mourjopoulos,et al.
A comparative study of least-squares and homomorphic techniques for the inversion of mixed phase signals
,
1982,
ICASSP.
[2]
D. W. Farnsworth,et al.
Exploration of Pressure Field Around the Human Head During Speech
,
1938
.
[3]
Heinrich Kuttruff,et al.
Room acoustics
,
1973
.
[4]
Jont B. Allen,et al.
Image method for efficiently simulating small‐room acoustics
,
1976
.
[5]
Jont B. Allen,et al.
Invertibility of a room impulse response
,
1979
.
[6]
Vern O. Knudsen.
THE HEARING OF SPEECH IN AUDITORIUMS
,
1929
.
[7]
Marinus M. Boone,et al.
A new method to acquire impulse responses in concert halls
,
1980
.