What is StemRoller?
StemRoller offers a user-friendly method to isolate audio components within any song. Users can initiate the process by simply entering a song title into the search interface. The application then locates the specified song online and utilizes artificial intelligence to divide it into distinct stems: vocals, drums, bass, and the residual musical elements.
Complementing the individual stem separation, StemRoller automatically generates a comprehensive instrumental track by merging all the non-vocal stems. This tool is built upon Facebook's advanced Demucs AI research project, making sophisticated audio source separation accessible as a completely free and open-source desktop application. It facilitates the creation of karaoke versions or the extraction of specific instrument tracks for various audio projects.
Features
- Song Search: Locates songs online using just the title.
- AI Stem Separation: Utilizes Demucs AI to split songs into vocals, drums, bass, and other stems.
- Instrumental Track Generation: Automatically creates an instrumental version by combining non-vocal stems.
- Multiple Stem Models: Offers choices including 6-stem separation and faster processing options.
- Free and Open-Source: Available at no cost and based on open-source technology.
Use Cases
- Creating custom karaoke tracks.
- Isolating vocals for acapella versions or remixes.
- Extracting specific instrument parts (drums, bass) for musical practice or sampling.
- Producing instrumental versions of songs for background music or performance.
- Assisting DJs in obtaining track elements for mixing.
- Supporting music producers and remix artists with source separation.
