Copyright (C) 2006-2012 - Darren Sessions
Portions Copyright (C) 2012, Cepstral LLC
Portions Copyright (C) 2013-2016 - Jeremy Kister
Portions Copyright (C) 2019 - dOpenSource
All rights reserved .
Provides a direct interface into the Cepstral text-to-speech engine for Asterisk eliminating the need to write audio output files to the filesystem and then stream them back out.
Also provides multi-digit dtmf recognition with a max digits and a wait-for-digit timer very simular to the AGI 'get data' command.
- Asterisk development header files
- Cepstral Swift Text-to-Speech engine (version 5 or 6)
./configure [options]
make
make install (as root)
make reload (as root)
grep 'swift' /etc/ld.so.conf
In /modules.conf uncomment EITHER:
;noload => chan_alsa.so
or ;noload => chan_oss.so
Load the module if not loaded:
asterisk -x 'module show like <your channel driver>'
asterisk -x 'module load <your channel driver>'
# rhel-based os
yum install -y alsa-lib alsa-lib-devel alsa-utils alsa-plugins-oss
# debian-based os
apt install -y libasound2 libasound2-dev alsa-base alsa-utils alsa-oss
Type "core show application swift" at the Asterisk CLI prompt.
in extensions.conf:
exten => s,1,Answer
exten => s,n,Swift(This is cepstral talking in the default voice from swift.conf)
exten => s,n,Set(SWIFT_VOICE=Callie-8kHz)
exten => s,n,Swift(This is cepstral talking in the Callie voice)
exten => s,n,Swift(Please enter three digits,5000,3)
exten => s,n,Swift(You entered ${SWIFT_DTMF}. Goodbye)
exten => s,n,Hangup