Hi, I want to make new voice font with cmu_us_slt_arctic in our
training system. Although our training system’s spectrum analysis module uses difference
method from festival’s, we intend to combine the same HTS Token Parameters as
festival’s(
really festvox) and our specific spectrum coefficients . How can I make new HTS voice font from cmu_us_slt_arctic (*.utt
) in festival project, and is there tool or documents for it? And I know that festvox doesn’t make synthetic analysis and complex part-of-speech
analysis, is it true? Why doesn’t festfox make synthetic analysis in English? Just for computational
efficiency? Or don’t it be needed in English TTS? Thanks. YoungHo Han / Engineer / 선임연구원 Core Lab. INFINITY TELECOM Gurogu, Gurodong, Seoul, KOREA O)+82-2-565-8808 F)+82-2-6675-8811 M) +82-10-3936-2191 Pursuing
Infinite Innovations |