A unit selection text-to-speech-and-singing synthesis framework from neutral speech: proof of concept | Publicación