Cvoc is a console binary that produces text-to-speech audio files.
It is written for Linux.
It requires an OpenAI authorization key set to an environment
variable called GPTKEY.
Cvoc uses the OpenAI audio.speech API with several speech parameters
to convert text into speech in the form of an audio file. The quality
of the speech produced is more lifelike than that of previous
text-to-speech technologies.
Explanation of the API parameters
Cvoc uses the Linux utility 'play' (SoX) to play back the audio file.
Cvoc uses the 'curl' utility to create/stream the audio file.
Cvoc uses a required cvoc.ini
file to hold default values for the
text-to-speech parameters (~/.config/cvoc.ini
). Edit this file to
change the defaults.
example:
output=speech
voice=nova
format=mp3
speed=0.95
model=tts-1
Note: the output file name does not have its extension.
This is because it will use the format for its extention.
Two ways to execute cvoc
:
cvoc
Executing without an argument prompts you for all of the parameters. Press Enter to accept the default value of any parameter. The text is the last parameter requested.
cvoc the cow jumped over the moon.
Executing with text produces an audio file based on default parameters (cvoc.ini).