I agree, you can do a lot of this with the digitone so it is likely possible with the ST ( many of their machines use 4op synthesis). I’m really into speech synthesis. Though a simplified speech machine might quickly grow overused.