| 摘要 |
The invention regards a method for noise reduction in an audio device whereby an electrical and/or digital signal which represents sound is routed simultaneously through:-a signal analysis path, and-a signal processing path wherein the signal amplification is individually controllable in specific frequency bands by attenuation values derived from the signal analysis path, whereby the signal in the signal analysis path is routed simultaneously through:-a first detector which identifies the presence of speech indicators in the overall signal, and-a second detector which in a predefined number of frequency bands detects the modulation amplitude, and-where attenuation values in each of the predefined frequency bands are calculated based on the combined results of the first detector and the modulation amplitude in the specific frequency band detected by the second detector,-where the attenuation values in the predefined number of frequency bands are routed to the signal processing path in order to attenuate the signal level in corresponding frequency bands. |
