This repository contains the text to speech material created at the Tutorial Factory at the #GLAMhack2021 for the Wikidata tutorials for the Sum Of All Swiss GLAMs project.
Project Team (and authors of the text in the SSML-files):
- Annina Clara Engel
- Alicia Fagerving
- Sarah Fuchs
- Valérie Hashimoto
- Oliver Waddell
- Nicolai Wenger
The generated audio will be published as spoken commentary to the presentations on the website https://tutorials.schoolofdata.ch/.
I put this up so you can easily create speech audio for your projects. Namely, feel free to grab any of the tutorials on https://tutorials.schoolofdata.ch/ in the language that suits you, translate the slides and the slide notes and generate the audio to go with it in one of the languages Google's Cloud Text-to-Speech service supports. Then upload it back to https://tutorials.schoolofdata.ch/ to have it published on the website and support GLAMs with using Wikidata!
Some notes on how to get the audio:
- For good results you should mark up the text you want spoken with Speech Synthesis Markup Language (SSML), have a look at the files in the folder "ssml" to see some examples
- If you want to use Google Cloud with Python like we did, have a look at their documentation, especially the quickstart document on how to use the client libraries (don't forget to set the environment variable GOOGLE_APPLICATION_CREDENTIALS) and the code sample to read SSML (you might want to consider using Poetry to manage the required dependencies)
- If you want to do it the way we did, have a look at the file filenames_tts_python_calls.ods that generates the filenames and the calls to create the audio and the file synthesize_text.py, in which you can simply append the calls to create the actual audio.
Good luck!, Oliver