Synthesizing singing

Abstract

An attempt to synthesize singing is presented in which a synthesizer containing five cascaded formants is complemented by two noise generators and fixed filters. The performance is controlled by computers reading the input notation and processing it with pronunciation and musical performance rules. Both solo and choral singing are synthesized. Various effects will be demonstrated, such as timing of pitch changes and consonants, coloratura, markato, timbre, legato, and multiple pitch singing. Also, some effects concerning musical expression will be presented.