A performance apparatus prevents a player's unintentional motion from being detected as a playing, thereby generating no sound, when the player plays a virtual musical instrument. The performance apparatus has a player operated stick, a ROM for storing layout information, which correlates plural areas set on a virtual plane to tone colors, respectively, an image sensor for continuously taking pickup images of a subject including the stick, and a CPU for calculating a difference image between a first pickup image and a second pickup image taken prior to the first pickup image and for detecting an operating position of the stick based on the difference image. A CPU refers to the ROM to specify a tone color correlated to the area corresponding to the operating position detected by the CPU. And a sound source generates a tone of the tone color specified by the CPU.