Abstract
Recorded speech or music signals experience considerable degradation when reproduced by a loudspeaker in a room. This work aims to differentiate between the effects of the normal-mode structure of the room and the loudspeaker. We have produced a tape of speech and music samples as they would be recorded in a room if reproduced through an ideal pulsating sphere. We have developed a spherical speaker with the property that speech and music samples reproduced through it in a room are subjectively indistinguishable from those that would be reproduced through the ideal pulsating sphere. The latter samples are generated by a digital convolution involving the Green's function of the room, which is obtained by a linear operation on signals generated by a spark discharge. This research indicates that a transducer can be designed that is subjectively equivalent to the ideal pulsating sphere; for this trandsucer, the normal-mode structure of the room—not that of the transducer—is the limiting factor in reproduction of speech and music signals. [Work supported ill part by the U. S. Army, U. S. Navy, and U. S. Air Force under contract DA06-039-AMC-03200(E); National Science Foundation (grant GP2495); National Institutes of Health, U. S. Department of Health, Education, and Welfare (grant MH-04737-04); and National Aeronautics and Space Administration (grant Xs(;-496).]
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have