xcLee001 /SonicVale

About SonicVale

SonicVale is a desktop app that turns novels and scripts into AI voice-overs. It splits dialogue by character and lets you pick a voice and emotion per line. The interface docs come only in Chinese, and the license blocks commercial use without written permission. You also have to download ffmpeg yourself before it runs.

Best for: Dubbing teams that need many characters and emotions for novels or short dramas.

• Imports novel, screenplay or video text and splits lines automatically
• Keeps a character library so voices stay the same across chapters
• Sets an emotion type and intensity for each line
• Edits dialogue in a table: text, character, emotion, intensity, auto-saved
• Cuts audio segments or inserts silence after generation
• Lets you point it at your own OpenAI-compatible LLM and prompts

1. 1
   Clone the repo: the command below then cd SonicVale

   git clone https://github.com/xcLee001/SonicVale.git

   Copy
2. 2
   Download ffmpeg.exe and put it in app/core/ffmpeg/
3. 3
   Install Python packages: the command below

   pip install -r requirements.txt

   Copy
4. 4
   Start the backend: the command below

   uvicorn app.main:app --reload --port 8200

   Copy
5. 5
   In sonicvale-front, run npm install then npm run start