Thanks for the great feedback. I think we can still improve the result, it's only a v1 in beta. I use nothing very advanced for rendering, a stack and tools rather simple. :)

I believe it's using Microsoft TTS voices (at least for some of them)