This paper presents an approach to distinguishing the zeros representing a sound source from those representing the transfer function on the basis of Lyon's residue-sign model. In machinery noise diagnostics, the source signature must be separated from observation records under reverberant path conditions. In numerical examples and an experimental piano-string vibration analysis, the modal responses could be synthesized by using clustered line-spectrum modeling. The modeling error represented the source signature subject to the source characteristics being given by a finite impulse response. The modeling error can be interpreted as a remainder function necessary for the zeros representing the source signature.
The ability to control the sound image to the rearward direction using front or lateral loudspeakers is an important step to popularize multichannel 3D audio systems in the home environment. Rear “directional bands” were found for narrow-band signals which are perceived only in the rearward direction, but any kind of directional bands for wide-band signals, which are perceived only in the rearward direction, have not been reported. It is known that spectral notches in head-related transfer functions are important cues for a sound image localization in the median plane. In the present study, the possibility of sound image control to the rearward direction was investigated using wide-band noise of which frequency components adjacent to the rear directional bands were eliminated. A total of 56 kinds of band-stop filters were used to eliminate one to the ten consecutive 1/3-octave bands with center frequencies of 1.6 to 12.5 kHz. In-phase stimuli were emitted from a pair of loudspeakers located at azimuth angles of ±90° in an anechoic room. The results showed that the probability of perceiving a sound image to be in the rearward direction was 95% when consecutive frequency bands, which were customized to individual subject, were eliminated.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.