Page 1 of 1

Raspivoice project now availabe on GitHub

Posted: Sat Nov 21, 2015 10:52 am
by wernert
A while ago I launched the following KickStarter project : ... ivoice-v20

The project is was funded successfully and I decided to make it available for download via GitHub:

It is a text-to-voice project with a difference. It uses WAV files for all the words it speaks. This means it speaks in very clear almost human like voice, but unfortunately there is a small pause between each individual words. Still, it is extremely easy to use it to make voice project that just need to speak short sentences.

There is documentation and instructions on how to install it on any Raspberry Pi (works on model A, B+ or RPi2) with Raspbian or even on an Ubuntu Linux PC at

Please feel free to give it a try and let us use this forum to discuss any improvements/ suggestions you may have about it. Looking forward to hear what you think.

Re: Raspivoice project now availabe on GitHub

Posted: Sat Nov 21, 2015 2:01 pm
by hippy
Have installed it but haven't managed to hear it yet as I seem to be having problems getting sound via HDMI ( everything, not just raspivoice ) and I don't have an audio jack lead to hand. sounds okay over HDMI. Voices seem to be a little more robotic than when the .wav files copied over to PC and played via Media Player. More noticeable with the male voice.

Much better than other text to speech I tried previously ( excepting via the Google speech API ). Some words are a bit odd, for example the word "the" seems rather peculiar and unnatural in male and female, eg -
./raspivoice ^m feed the world
./raspivoice ^f feed the world

A few questions -

How to get new words added with the same voices used in the provided .wav files ? For example; "aluminium" / "aluminum" ?

How to select between US/Canadian/British/other pronunciations for words like "lieutenant", "buoy" and "z" ?

What is the license for use of the .wav files ?