To provide a silence compression transmission device capable of processing a speech in real time.
A speech signal (c) is inputted from a signal source 1 to a silence compression transmission device 3 through a transmission path 2, and a reference signal a generated from a reference signal generation function 5 for generating a stable reference signal is varied by a varying function 6 which varies according to components of the input speech signal (c), and it is judged whether the signal (c) is voiced or unvoiced, by comparing a difference between the varied signal and the reference signal (a) by a comparison judging function 7, and the result of the judgment is outputted (d). On the other hand, the input speech signal (c) is outputted via the signal sending function 8. The signal sending function 8 outputs (e) in a form in which the silence part is removed from the input speech signal (c)by the output (d) of the comparison judging function 7.
Next Patent: RECORDER FOR AUDIO SIGNAL AND RECORDING METHOD THEREOF