Synthesized voices are legitimately a great way to read more and give your eyes a break. I personally prefer just converting a page or book to an audiobook myself locally.
The new piper TTS models are easy to run locally and work very well. I made a simple CLI application and some other folks here liked it so figured I post it.
I got Piper TTS running in a Docker container (I found that the issue is related to Python version and "phenomenize" library). If you're curious / interested in getting this to work, happy to help out & share the code. My contact is in my profile.
https://github.com/C-Loftus/QuickPiperAudiobook