Abstract (EN):
Acoustic feedback limits the maximum stable gain (MSG) of a public address (PA) system and thereby may make the system unstable. Acoustic feedback cancellation methods use an adaptive filter to identify the acoustic feedback path and remove its influence from the system. However, if the traditional adaptive filtering algorithms are used, a bias is introduced in the adaptive filter coefficients. Several methods have been proposed to overcome the bias problem but their performances are generally evaluated in PA systems with only one microphone and one loudspeaker. This work evaluates the performance of some state-of-art methods in a PA system with one microphone and four loudspeakers, which is a more practical configuration of such a system, with respect to the increase in MSG, estimate of the feedback path and sound quality. Simulation results demonstrated that, when the source signal is speech, the method based on the cepstrum of the error signal has the best overall performance and can increase in 30.6 dB the MSG of the PA system.
Idioma:
Inglês
Tipo (Avaliação Docente):
Científica
Nº de páginas:
23