Abstract: A sequence of performance data to be loop-reproduced is read out repetitively for at least two cycles, and audio data repeating for at least two cycles are created on the basis of the read-out performance data. Audio data of a second or subsequent cycle of the created audio data have mixed therein a residual portion (release portion or reverberation portion) of a tone generated in a preceding cycle. The audio data of the second or subsequent cycle are stored as audio data for loop reproduction, and thus, the stored audio data for loop reproduction will include a residual portion of a tone generated in a preceding cycle. Thus, if the stored audio data for loop reproduction are reproduced repetitively, it is possible to achieve loop reproduction reflecting therein a residual portion of a tone generated in a preceding cycle.