naztar0 / AudioSplitterBot

Telegram bot for separating songs into stems

Audio Splitter Bot

A Telegram bot for separating songs into components such as vocals, instrumental, guitar, or bass.

Basic technical info

• Based on aiogram python module and ffmpeg for audio processing
• Uses the lalal.ai API for separating audio tracks

How it works

Lala.ai is a service that allows you to separate audio tracks into components such as vocals, instrumental, guitar, or bass. But free usage on the official site is limited to 1 minute of audio.

This bot allows you to split audio tracks without any restrictions. When you send an audio file to the bot, it splits the file into pieces one minute at a time, sends it to the server, and then combines it back into one file and sends the result to the user.

Deployment

1. Create a database with the tables from the models.sql file
2. Fill out the .env file with the necessary data
3. Install dependencies from requirements.txt
4. Compile localization files with pybabel compile -d locales -D bot
5. Run the bot with python -m app
6. Run the updater daemon with python daemons/updater.py

License

This project is licensed under the CC BY-NC-SA 4.0 License - see the LICENSE file for details.