On the basis of performance event information, at least two notes to be sounded in succession or in an overlapping relation to each other are detected, and a tone pitch difference between the detected at least two notes is detected. Tone pitch difference limitation ranges are set in corresponding relation to various rendition styles. Rendition style to be imparted to the notes to be sounded in succession or in an overlapping relation is designated, and a comparison is made between a tone pitch difference limitation range corresponding to the designated rendition style and the detected tone pitch difference, so as determine whether the designated rendition style is applicable or not. In this way, whether the designated rendition style is to be applied or not is automatically controlled in accordance with the tone pitch difference of the at least two notes to be imparted with the rendition style. Further, pitch range limitation ranges are set in corresponding relation to various rendition styles, and the applicability of a designated rendition style is controlled depending on whether or not a tone to be imparted with the designated rendition style is within the pitch range limitation range corresponding to the designated rendition style.