.TH "open_jtalk" "1" .SH "NAME" open_jtalk \(em Japanese TTS system .SH "SYNOPSIS" .PP \fBopen_jtalk\fR [\fBoptions\fP] [\fBinfile\fP] .SH "DESCRIPTION" .PP This manual page documents briefly the \fBopen_jtalk\fR command. .PP This manual page was written for the \fBDebian\fP distribution because the original program does not have a manual page. Instead, it has documentation in the GNU \fBInfo\fP format; see below. .PP \fBopen_jtalk\fR is a program that synthesize speech waveform from Japanese texts. It uses HMMs trained by the HMM-based speech synthesis system (HTS). .SH "OPTIONS" .PP A summary of options is included below. .IP "\fB-x dir\fP" 10 dictionary directory .IP "\fB-td tree\fP" 10 decision tree files for state duration .IP "\fB-tm tree\fP" 10 Show version of program. .IP "\fB-tf tree\fP" 10 decision tree files for Log F0 .IP "\fB-tl tree\fP" 10 decision tree files for low-pass filter .IP "\fB-md pdf\fP" 10 model files for state duration .IP "\fB-mm pdf\fP" 10 model files for spectrum .IP "\fB-mf pdf\fP" 10 model files for Log F0 .IP "\fB-ml pdf\fP" 10 model files for low-pass filter .IP "\fB-dm win\fP" 10 window files for calculation delta of spectrum .IP "\fB-df win\fP" 10 window files for calculation delta of Log F0 .IP "\fB-dl win\fP" 10 window files for calculation delta of low-pass filter .IP "\fB-ow s\fP" 10 filename of output wav audio (generated speech) .IP "\fB-ot s\fP" 10 filename of output trace information .IP "\fB-s i\fP" 10 sampling frequency [16000][1\-\-48000] .IP "\fB-p i\fP" 10 frame period (point) [80][1\-\-] .IP "\fB-a f\fP" 10 all-pass constant [0.42][0.0\-\-1.0] .IP "\fB-g i\fP" 10 gamma = \-1 / i (if i=0 then gamma=0) [0][0\-\-] .IP "\fB-b f\fP" 10 postfiltering coefficient [0.0][\-0.8\-\-8.0] .IP "\fB-l\fP" 10 regard input as log gain and output linear one (LSP) .IP "\fB-u f\fP" 10 voiced/unvoiced threshold[0.5][0.0\-\-1.0] .IP "\fB-em tree\fP" 10 decision tree files for GV of spectrum .IP "\fB-ef tree\fP" 10 decision tree files for GV of Log F0 .IP "\fB-el tree\fP" 10 decision tree files for GV of low-pass filter .IP "\fB-cm pdf\fP" 10 filenames of GV for spectrum .IP "\fB-cf pdf\fP" 10 filenames of GV for Log F0 .IP "\fB-cl pdf\fP" 10 filenames of GV for low-pass filter .IP "\fB-jm f\fP" 10 weight of GV for spectrum [1.0][0.0\-\-2.0] .IP "\fB-jf f\fP" 10 weight of GV for Log F0 [1.0][0.0\-\-2.0] .IP "\fB-jl f\fP" 10 weight of GV for low-pass filter [1.0][0.0\-\-2.0] .IP "\fB-k tree\fP" 10 GV switch .IP "\fB-z i\fP" 10 audio buffer size [1600][0\-\-48000] .IP "\fBinfile\fP" 10 text file .PP option '\-d' may be repeated to use multiple delta parameters. generated spectrum, log F0, and low-pass filter coefficient sequences are saved in natural endian, binary (float) format. .SH "EXAMPLE" .PP If you installed hts-voice-nitech-jp-atr503-m001 in the current directory, the following command let you make a voice file from input.txt: .PP .RS .PP .nf % open_jtalk \-s 48000 \-p 240 \-a 0.55 \\ \-td tree-dur.inf \-tm tree-mgc.inf \-tf tree-lf0.inf \\ \-tl tree-lpf.inf \-md dur.pdf \-mm mgc.pdf \\ \-mf lf0.pdf \-ml lpf.pdf \-dm mgc.win1 \\ \-dm mgc.win2 \-dm mgc.win3 \-df lf0.win1 \\ \-df lf0.win2 \-df lf0.win3 \-dl lpf.win1 \\ \-em tree-gv-mgc.inf \-ef tree-gv-lf0.inf \-cm gv-mgc.pdf \\ \-cf gv-lf0.pdf \-k gv-switch.inf \-ow output.wav \\ \-x dic_dir input.txt .fi .PP .RE .SH "AUTHOR" .PP This manual page was written by Koichi Akabe vbkaisetsu@gmail.com for the \fBDebian\fP system (and may be used by others). Permission is granted to copy, distribute and/or modify this document under the terms of the GNU General Public License, Version 2 any later version published by the Free Software Foundation. .PP On Debian systems, the complete text of the GNU General Public License can be found in /usr/share/common-licenses/GPL. .\" created by instant / docbook-to-man, Tue 03 Jan 2012, 22:40