This software contains an integration of the WORLD vocoder (https://github.com/mmorise/World) and hts_engine API-1.10 (http://hts-engine.sourceforge.net). It also uses functions from the Speech Signal Processing Toolkit (SPTK) (http://sp-tk.sourceforge.net/).
The main modification can be found in lib/synthworld.cpp.
Follow the classic sequence:
autoreconf --install
./configure
make
at the end, the executable bin/hts_engine
will be created
You have to call hts_engine with the option -w to use the WORLD vocoder, otherwise the default MLSA vocoder will be used.
@Misc{pucher2018,
author = "Michael Pucher",
title = "Hts-engine-world: An integration of the WORLD vocoder and hts_engine API-1.10",
year = "2018",
publisher = "GitHub",
journal = "GitHub repository",
howpublished = "\url{https://github.com/mipuc/hts-engine-world}",
}
Michael Pucher michael.pucher@oeaw.ac.at