VoxNovel is an ongoing project that generates audiobooks from a book input and utilizes different actors for each character in the book. This allows for a more immersive and engaging listening experience.
Listen to a sample of an audiobook generated with VoxNovel-Tortoise using a small passage from "Harry Potter and the Sorcerer's Stone": Tortoise Demo MP3 file.
Listen to a sample test file for audiobooks generated with VoxNovel using Bark TTS: BARK Demo MP3 file.
To get started with VoxNovel, follow these simple steps:
- Use bookNLP to extract the necessary HTML metadata from your desired book/text file in a txt format. A working Google Colab demo for BookNLP_EXTRACT_DEMO can be found here: BookNLP_EXTRACT_DEMO.
- Run the BookNLP_Demo output "TTS_Input.txt" file in the VoxNovel Google Colab Demo found here: VoxNovel tortoise Google Colab Demo.
For higher quality voices, check out our Bark TTS demo (although it may have more hallucinations than Tortoise and be less easy to clone other voices). A working Google Colab demo for Bark TTS can be found here: Bark TTS Demo.
If speed is more important to you, we recommend our Coqui TTS demo, although it has more Siri-like voices. A working Google Colab demo for Coqui TTS can be found here: Coqui TTS Demo.
If you have any feedback or suggestions, please join our Official Discord server: VoxNovel Official Discord Server. We would love to hear from you!