Does anyone have experience making Festival voices and recording their own prompts?
By David Jones •
One of my acquaintances on a forum recently posted a list of things to check off and I decided, since I have a virtual machine with audio input and output, to give this a shot.
The first thing I did was read the checklist provided.
Then I wrote a little tutorial after a friend of mine helped me figure out how to build and install each of these components.
After that was done, I downloaded a series of voices and balanced prompts from the arctic databases. I extracted some of the voices and did a
~/festival/festival$ make testand got an introduction message, followed by this message:
voice script status: FAILED
test modes (script)
text modes
SIOD ERROR: unbound variable : voice_cmu_us_rms_cg
festival: text modes, caught error and tidying up
Segmentation fault (core dumped)
modes script status: FAILED
test parse (script)
parse script completed
parse script status: CORRECT
test scherr (script)
scherr script completed
10c10
Festival Speech Synthesis System 2.5.0:current February 2017
Festival Speech Synthesis System 2.4:beta December 2014
14c14
clustergen_engine: Copyright (C) Carnegie Mellon University 2005-2017
clustergen_engine: Copyright (C) Carnegie Mellon University 2005-2014
scherr script status: INCORRECTHere's the post I'm talking about:
How can I fix this?
5 Reset to default
