Text to speech

wblock@

wblock@

Developer
Is there a simple (small) text to speech program in ports? So far, I've seen audio/festival, which was unable to make a sound due to "no default voice". The man page is not helpful about that, and the web page is willing to to great depths to help create new voices, but neglects mentioning how to set one. They really love Lisp, though. Really. A lot.

audio/espeak requires audio/pulseaudio. No thanks.

audio/praat appears to be an audio speech research tool written by by a gang of Dutch mad scientists, evidently as some sort of world domination effort, or possibly to communicate with aliens, dolphins, or people from the South.

All I want is some announcements, possibly with numbers. Is recording my own the easiest way? Are there prerecorded samples in a port that could be used?
 
audio/pulseaudio is optional for audio/espeak. It won't be used, if you configure espeak with PortAudio library support. pulseaudio will only be installed when selecting one of the other two configure options. I have it installed using:

Code: 
(*) PORTAUDIO   PortAudio library support
( ) PULSEAUDIO  PulseAudio sound server support
( ) RUNTIME     Sound interface selected at runtime

And it works fine without pulseaudio.
 
rolfheinrich said:
audio/pulseaudio is optional for audio/espeak. It won't be used, if you configure espeak with PortAudio library support. pulseaudio will only be installed when selecting one of the other two configure options. I have it installed using:

Code: 
(*) PORTAUDIO   PortAudio library support
( ) PULSEAUDIO  PulseAudio sound server support
( ) RUNTIME     Sound interface selected at runtime

And it works fine without pulseaudio.
Click to expand...

Aha. I had turned off PORTAUDIO, and then it wanted to pull in audio/pulseaudio.
 
It's another (complex) daemon running and in the past it didn't work all that well (on FreeBSD anyway). I haven't tried it for a long time. All this for questionable benefit given FreeBSD's sound system is pretty darn good. It does add some functionality (networking stuff) and some things (e.g. webrtc) depend on it.
 
Why the text to speech is necessary, and how you wish to configure it :-D? I mean the voice will probably suck unlike some of the proprietary ones... let me know!!
 
audio/espeak is nice :) The default package also does not depend on audio/pulseaudio. I've bound the following script to a key if I want someone to read the currently selected text to me (also requires x11/xsel-conrad). This works even in PDF readers! It does not matter that the voice is not as good as it could be:
Code: 
#!/bin/sh
xsel -p | tr '\n' ' ' | espeak -s 300 -a 150 -v en-us
I've also bound pkill espeak to a key to stop it before it finishes.
 
You must log in or register to reply here.
Back
Top