Complements to the articulatory synthesis of formant targeted sounds

We present the recent development as a complement to the previous articulatory synthesis of formant targeted sounds with parameters derived from the inverse solution of speech production which has been proposed previously. The interpolation formula involved in the inverse solution is modified to guarantee the good nature of the interpolated zeros, i.e., the zeros must reside in the gap of their neighborhood poles. Comparison between the modified interpolation method to the old one is given. Also, a timing procedure of the pitch-synchronous reflection-type line analog synthesizer that is controlled by the frame-by-frame segment-based vocal tract (VT) area function together with variant vocal-tract length is proposed.