Recent implementations of binaural synthesis have combined high-frequency pinna diffraction data with low-frequency acoustic models of the head and torso. This combination ensures that the salient cues required for directional localization in the horizontal plane are consistent with psychophysical expectations, regardless of the accuracy or match of the high-frequency cues, or the fidelity of experimental low-frequency information. This paper investigates the effect of a nonrigid boundary condition on the surface pressure and the resulting interaural cues used for horizontal localization. These are derived from an analytical single sphere diffraction model assuming a locally reacting and uniformly distributed impedance boundary condition. Decreasing the magnitude of a purely resistive surface impedance results in an overall decrease in the sphere surface pressure level, particularly in the posterior region. This produces nontrivial increases in both the interaural level and time difference, especially for sound source directions near the interaural axis. When the surface impedance contains a reactive component the interaural cues exhibit further changes. The basic impedance characteristics of human hair and their incorporation into the sphere diffraction model are also discussed.
Previous analytical and empirical studies of the human auditory system have shown that the cues used for localization are modified by the inclusion of nonrigid scattering surfaces (clothing, hair etc). This paper presents an investigation into the acoustic impedance properties of human hair. The legitimacy of a locally reactive surface assumption is investigated, and an appropriate boundary condition is formulated to account for the physiological composition of a human head with hair. This utilizes an equivalent impedance parameter to allow the scattering boundary to be defined at a reference plane coincident with the inner rigid surface of the head. Experimental examination of a representative synthetic hair material at oblique incidence is used to show that a locally reactive surface assumption is legitimate. Additional experimental analysis of a simple scattering problem illustrates that the equivalent impedance must be used in favor of the traditional surface impedance to yield physically correct pressure magnitudes. The equivalent acoustic impedance properties of a representative range of human hair samples are discussed, including trends with sample thickness, fiber diameter, bulk density, and mass.
A general analytical model is developed for the scattering of sound by a sphere with a nonuniform impedance boundary condition that is divided into two uniformly distributed hemispheres. In addition to the overall solution for the time harmonic pressure, the analytical result gives insight into the modal contributions and coupling for different cases of source incidence and boundary impedance. Modal cross coupling is shown to exist between incoming and scattered wave modes of equi-order and nonequal degree when the degrees are opposite in parity (odd-even or even-odd coupling). This cross coupling is strongest between modes of adjacent degree, and decreases as the degrees become dissimilar. The overall magnitude of the cross coupling is dependent on the extent of the impedance mismatch between the two surface hemispheres. Simulation and discussion are given for several specific cases of source incidence and impedance (each hemisphere is given a different constant impedance value). These results are consistent with expectations from the scattering of sound by a sphere with a uniformly distributed surface boundary. The broad scattering characteristics of the hemispherically divided sphere are shown to be analogous to connecting the appropriate sectors from the corresponding uniformly distributed spheres.
This paper presents the modeling and simulation results of active noise control (ANC) in a small room using the wave-based approach defined by particle velocities and sound pressure within the defined boundary conditions. The ANC system excitation is a single-frequency noise with an adaptive feedforward configuration. The Finite Difference Time Domain (FDTD) algorithm is used to model the room acoustics due to a boxed loudspeaker of single frequency. A control system based on the filtered-x least mean-squared (FxLMS) algorithm is utilized to synthesize a cancelling noise using a secondary loudspeaker. The single channel system is modified into a multichannel system and genetic algorithm (GA) is used to optimize the sensors and actuators placements simultaneously. Numerical results are plotted to demonstrate the performance of the control system. These show that the numerical modelling technique can be used to combine room acoustic simulation and FxLMS adaptive control. This provides a way for the optimum placement of the microphones and loudspeakers before being used in a practical complex enclosure.
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.