A microphone and receiver for automatic accompaniment in which the microphone outputs a mixed audio signal of a selected melody and a song voice signal resulting from singing a song to the accompaniment of the selected melody, and image control data corresponding to the selected melody. The microphone stores, plays and outputs the mixed audio signal and/or song voice signal and the receiver receives the mixed audio signal and/or the song voice signal outputted from the microphone, and outputs the mixed audio signal and/or song voice signal to a television receiver set, and at the same time, generates a caption image correspond to song words of the selected melody and a background image according to the received image control data, for output to the television receiver set.