Thank you.
I recognize some of the links stated in the answers. The quality of espeak
and festival voices I have installed or tested, is not high enough. I am
impressed of the quality of windows voice anna. On the internet there was,
and maybe still is, some windows sapi5 voices. Fx english and french voices.
Likely not with an legal license. They are all of a high standard. There was
at some point an effort to make software, so you could install them on
gnulinux. But it stalled. Reverse engineering them to source code, that is
usable on gnulinux is not possible?
The cepstral voices I know of. I find the english david voice acceptable.
Cepstral mentions the cepstral voices in combination with orca. I downloaded
voice david and installed it. And got no error messages. But I could not get
orca to use voice david. I did not get voice david to make any sound. I am
not that skilled, that I can determine if it can be done on ubuntu 14.04.
Cepstral does not provide support. Cepstral linux voices are not free or open
source software?
From time to time gnulinux puzzles me. Fx to my knowledge there is no high
quality gnulinux video editor. Still pitivi cannot get their crowdfunding
funded fast. Maybe there are reasons that a voice of windows voice anna
quality is not a priority for gnulinux. I think I once heard, that people who
rely on text to speech voices, prefer voices like the orca voices. And often
they heavily speed up the voices. If I want to listen to a book converted to
an audio book via text to speech, I want the quality level of windows voice
anna. If the money came together, getting natural gnulinux voices should be
manageable?
I find some of the edinburgh voices acceptable. At some point I wrote
edinburgh and asked if I could get or buy their voices. They replied, that
they were using underlaying licenses limiting them from turning over the
software to anyone. I do not recall asking them who would benefit from their
research?
Ivona make acceptable voices. They make many languages. The software is not
freeware, nor open source. They only make windows versions. My knowledge is,
that they use one basic software technology. I guess algorithms is what it is
about. To make a voice, they record human beeings speaking thousand and
thousand of words and sentences. They output of the computation of the sound
files is voice software.
