Method and apparatus for preprocessing of an acoustic signal before speech coding.

摘要

The method involves input of an acoustic signal to a high pass filter. If the energy (E2) of the high-pass filter output exceeds a certain fraction of the energy (E1) of the unfiltered signal, a first state is selected. If it is less, a second state is selected where the signal is subjected to high frequency pre-emphasis. The modification does not occur if the filtered signal energy is less than a threshold. The state is changed if several successive signal frames indicate a state different to that previously determined. A counter (32) is incremented when a comparator (27) indicates a different state and is decremented towards zero when the same state is indicated. At a threshold value, the output state is changed and returns to zero.