The Cassiopaean TTS Project

Thraaros

A Disturbance in the Force
Do you prefer listening to reading?
Do you know someone who'd be interested in reading The Wave, but they don't like reading too much?
Do you want something closer to a recording of the sessions than a transcript?
Then this is for you!

On July 21st 2025, I started creating a text-to-speech script that utilises the app @Voice Aloud Reader and a RegEx-based Readability parser to assign individualized digital voices to each speaker in the transcript (matching their genders if known), examples attached to this post. You can listen to the transcripts in the app at a customizable speed. It also shows images from the chapters. I used as many unique voices as I could.

The reading app can be downloaded:
- from the Play Store
- directly from the website (for use in Android emulators, and more)
There is no iOS or desktop version at the time of writing.

The script is available under this permanent link: Dropbox
The text file displays the current version (19, as of writing), and I'll be updating it as "time" goes on. It can technically be used for any chapter beyond the latest, but new speakers who occasionally appear in later chapters will sound like the previous voice. The file contains comments that explain how the code functions in case someone wants to improve it.

Instructions:
  1. Download the app and the script (the .json text file).
  2. Share the file to @Voice Aloud Reader on the same device. You don't have to open it directly.
  3. Follow the instructions included on the script's opening page. (Share chapters of The Wave from a web browser to the app, but don't copy and paste the text; use the share function in your browser's menu.)
  4. Check the link periodically for updates.
  • Please tell me if the instructions are simple enough or if screenshots or screen recordings are required.
I'm not affiliated with the app, but since Greg, the app's developer, spent many hours teaching me the functions of the parser in an email conversation, the least I can do for him is ask you, if you find the app useful, to consider buying the premium ad-free license. Greg also said he opposes subscription models, so you keep it forever; however, this project is free to use and improve (paid for with time and attention, the spiritual currency, if you will).

Trivia: I deduced the genders whenever possible to match the voices. Some names (Laura, Terry, Jan, Frank, Ark) or just pronouns (V, the most anonymous of the group) are used at least once somewhere in the first two volumes, so I also used those names in the script. Please let me know if I'm mistaken. I'm a fan of the voice I gave the Cassiopaeans, so I hope you like it: check the attached example recording from chapter 2 (in two file formats).
 

Attachments

@Thraaros , thank you, but we are actually in the process of finalizing our own audiobooks with Laura's voice. The first one should be released this month. It sounds way more natural than your demo, because we cloned her voice and did a lot of work to tweak the audio. AI is great, text-to-speech are great, but for this type of book, we wanted to make something just a little more professional.

It's always better to post here first before launching into a project such as this (and to get the author's authorization). If it hasn't been done, it's usually "already cooking".

Take your time to get familiarized with the forum. I see you just joined but said you wouldn't be around much, so no rush.
 
(I forgot to mention in my post that version 19 means it's fully compatible with 19 chapters, and slightly compatible with later chapters).

Hi @Chu, does that audiobook include voices for speakers other than Laura? If not, I might just stick with my script and keep improving it. I'm not sure yet. And on the topic of sounding robotic... (By the way, my project doesn't use AI at the moment.)

I've just found an AI TTS website that can be used to generate dialogues with different voices using a slightly modified version of the parsing script that I could easily make TTSReader V3.9.12 - TTSReader's Text to Speech Player. I've attached a manually-created mockup. While it requires a premium subscription, it allows downloading sound files and can be used to publish MP3s, so there'd be no need to run scripts.
 

Attachments

Back
Top Bottom