Stem Splitter

A stem is an isolated component, group, or part of a song. For example, a typical pop song could be split into stems for vocals, piano, drums, bass, and other instruments. Each stem is typically a separate audio file that can be played or remixed separately, giving producers granular control over each part of the music.

If you recorded each track of a song separately, then you already have separate stems and do not need a stem splitter. However, if you only have a full mix of a song and you want granular control over each part, then you need to split the stems. Splitting a song into stems allows you to do the following cool things:

• Practice: Mute one part and play along, or play only one part to help you transcribe it.
• Remix: Take just one isolated part like the vocals and remix it with other parts of your own creation. You can raise the volume of the vocals to hear the lyrics better, or you can crank up the bass if you want to make your neighbors mad.
• Perform: Use the backing track without vocals and sing to it during a live performance.

Our tool performs audio source separation using machine learning techniques. All you need to do is upload an audio file, and our tool will automatically split the various parts into stems. As soon as the process completes, you can download a zip file containing multiple audio files. Each audio file represents one stem or isolated part of the song. Then you can import each part (vocals, piano, drums, bass, etc.) into your favorite DAW and start mixing.

The tool is completely free during our open public beta. When the beta ends, we plan to enable a paid plan that will include premium features such as faster processing times, higher quality outputs, longer audio files, and more. However, we plan to always have a free version that is available to everyone.