A method and electronic device for obtaining clear playback sound that is faithful to the original sound, in which data and audio data are played back under control of a CPU, and in accordance with the played-back audio data, a timer, which generates a CPU interrupt signal, is controlled and said interrupt signal is dynamically altered. The sound data obtained in accordance with the CPU interrupt signal is emitted to a speaker, thereby causing the timing of the second data and the timing of the CPU interrupt signal to agree with each other, thus reducing the burden on the CPU, and generating a clear playback sound from the speaker.