IVONA TTS reader

Simple python3 application that reads aloud every line you type in STDIN using IVONA Speech Cloud. It also provides a simple spelling fixer using aspell.

I tested it under Python 3.5. It works fine for English, Polish and should also work for many other languages (but I didn't tested).

Why?

I was ill (and bored) and I couldn't speak because of a problem with vocal cords so I created this simple application to help me communicate more easily with people around me in home. I tried to use eSpeak but I wasn't satisfied with the quality of a generated sound.

Setup

You need to install gstreamer python3 bindings in your system. The spelling fixer requires libaspell installed. For example in Ubuntu or debian you can install them with:

$ sudo apt install python3-gst-1.0 libaspell-dev

Then set up python3 virtual env and install dependencies.

$ python3 -m venv --system-site-packages env
$ source env/bin/activate
(env) $ pip install -r requirements.txt

Usage

First you need to create IVONA developer account, generate credentials and and set them in environment variables:

$ export IVONA_SECRET_KEY="secretsecretsecretsecretsecret"
$ export IVONA_ACCESS_KEY="asdasdasdasdasdasdasd"

You can also set nearest region by setting IVONA_REGION env variable. Available ones:

eu-west-1: EU, Dublin (default one)
us-east-1: US East, N. Virginia
us-west-2: US West, Oregon

Then in the virtual env run

(env) $ python3 ivona_speech.py --help

to get all runtime options.

Tips

By placing ! at the beginning of line you disable spelling fixer for that line.

Example