I’m excited to share one of my latest projects. an audio separator that can separate the components of music into vocals, drums, and instruments. Not just that, I also built a voice separator that focuses only on human speech.
This project was created in Python, and I built it with the help of AI for guidance and troubleshooting. In other words, to facilitate the work. Which AI never fails to do.
What is the project?
The goal of this project is simple:
Take an audio file (a song, a podcast, or any recording).
Strip it into individual components like vocals, drums, guitar, etc.
Or extract only the human voice, filtering out background sounds. This is simple, no fancy equipment, nothing, just some Python codes.
This is useful for:
Musicians who want instrumentals for practice.
Podcasters who need cleaner speech. without background noises or sounds that could make the audio unclear.
And even developers who are experimenting with audio processing can benefit from it.
So as you see, it's not just one field tool. It's for all.
How it works
The project uses Python with libraries, which include:
Spleeter: for splitting music into parts (vocals, drums, bass, etc.).
Librosa: for analyzing and processing audio. It's like the mind of the code.
Sound device & sound file: for recording audio and then saving it.
Scipy: for signal filtering.
NumPy: for efficient mathematical operations.
Example: Music Separation
Feed in a song, then get separate audio files for vocals, drums, bass, and more.
Example: voice separation
Record speech if you don't have it ready or load an audio file. Extract only the spoken voice, and ignore the background sounds.
License
This project is licensed under the MIT License, so you are free to use and modify it as you wish without any backlash.
AI's role
This project was built with the help of AI assistance for:
Writing separation scripts.
Debugging dependency issues. Or make it more powerful, in other words.
Structuring the repository and documentation. to make it easier to scan.
It was an exciting way to collaborate with AI to build this project.
Final Thoughts
Working on this project showed me how powerful open-source tools and AI collaboration can be with the right mindset and the willingness to help. If you’re interested in audio, Python, or machine learning, I’d love for you to check it out and share feedback!
GitHub Repository: Audio Splitter Voice Music Separator
Top comments (0)