摘要 |
PURPOSE:To prevent great deterioration of the S/N of sending speech output due to noises, by varying automatically the gain frequency characteristics of a speech sending path in response to the noise level so that a proper S/N is always secured. CONSTITUTION:The speaker's voices are collected to a microphone 5 and amplified by a preamplifier 3 to be applied to a noise level detecting part 1 and a filter part 2. The audio signal is supplied to a rectifying and charging/discharging circuit 12 at the part 1. The circuit 12 is set so as to have a large charging time constant and a small discharging time constant. Thus the audio signal having its amplitude changing suddenly on the time axis is not charged; while the charging voltage corresponding to the noise level is delivered to a level deciding circuit 13 for the steady noise having its amplitude changing slowly on the time axis. Then the circuit 13 delivers a level deciding signal corresponding to the charging potential to a control circuit 14. The circuit 14 delivers a binary signal in response to the noise level and has the filter characteristics when the signal is set at ''1'' and has no filter characteristics when the signal is set at ''0'', respectively. |