A voice controller controls a voice-controlled apparatus, such as a television set, and includes a receiver for receiving sound information from a corresponding transmitter, which is associated with at least one sound source, for example an amplifier or the television set itself. A processor coupled to a microphone and the receiver corrects the sound signal registered by the microphone in accordance with the sound information received by the receiver in order to eliminate a component of the sound signal corresponding to the sound information. The sound signal corrected in this way is subjected to voice recognition.