Method and system for acoustic shock protection

摘要

The method involves performing a pattern analysis on an input signal, preferably using a feature extraction process, to identify a parameter space corresponding to a signal space of the input signal. A rule-based decision is then applied to the parameter space to detect an acoustic shock event, which is then removed. Gain control is performed as part of the removal process. Analysis of the input signal is preferably performed using a weighted overlap-add (WOLA) analysis technique, to identify a parameter space corresponding to the signal space of the input signal. A rule-based decision is applied to the identified parameter space to detect the acoustic shock event for removal. Alternatively, the analysis block (4) may use e.g. artificial neural networks, hidden Markov models, Bayesian estimation methods, or signal statistics. Gain control is applied, based on the shock detection and features extracted from the input signal, a calibrated gain being set to meet a predetermined safe output level. Preferably a feedforward configuration is used to provide a fast-acting and intelligent gain control system. A timer component (14) in the analysis block provides protection against level-over-time exposure. A logging block (10) provides on-line data collection of the key parameters of shock events and a calibration block (12) provides data required to account for differences in headphones, headsets, and other factors. An independent claim is also included for a system for providing protection against acoustic shock.