In this guide
Open an audio clip's Playlist menu and select Extract stems from sample. Choose the parts you need, check how the original is muted, and audition the new tracks before exporting. Separated parts are estimates, not the original session tracks.
FL Studio stem separation extracts up to four parts from an audio clip: drums, bass, instruments and vocals. In the Playlist, open the clip menu and choose Extract stems from sample, select the parts you want and run the extraction. FL Studio places each selected part on a new Playlist track beneath the source track.
Image-Line's Playlist manual describes the command, its muting options and the first-use model download. Menus and features vary with version; check your installation and current product terms if the option is absent.
Extract stems from an audio clip
- Prepare the source. Import a recording you have permission to use as an audio clip in the Playlist. Keep the original so you can compare a quiet verse and busy chorus with the extracted parts.
- Open the clip menu. Click the menu icon at the upper left of the audio clip and choose Extract stems from sample. If prompted, allow the initial model download and wait for it to finish.
- Select the stems. Choose the needed combination of vocals, drums, bass and instruments. Each selected item creates a Playlist track; the instruments category can still contain several instruments together.
- Choose source muting. The dialog offers options to mute the source clip, mute the source track or do nothing. Muting the source helps prevent hearing it together with the extracted parts. Check whether the source track also contains other clips you need before muting the entire track.
- Extract and listen. Image-Line notes that long processing can take minutes and FL Studio may be unresponsive during extraction. After it completes, audition each part and the combined result against the original at a similar level.
How do I make a backing track or acapella?
For a vocal-reduced backing, extract the necessary parts, mute the original and audition the accompaniment without the vocal stem. For a vocal reference, audition the vocal stem separately. Listen to layered choruses, reverb tails and quiet phrases: vocals can remain in the accompaniment and wanted instrument details can be lost. Keep an authorized instrumental or original session tracks as a reference when available.
For a handoff, confirm the expected start time, format and processing. FL Studio's manual describes Playlist-track export options, including export from song start. Review the files in a new project and check their balance and alignment; see the stem delivery checklist.
What if stem extraction is missing or fails?
| Problem | What to check |
|---|---|
| Command is missing | Check the installed FL Studio version and edition, then open the menu of an audio clip in the Playlist. A pattern or automation clip has different menu options. |
| First use prompts for a download | The separation model is a downloadable component. Follow the prompt and allow it to complete before retrying. |
| Extraction crashes or sounds garbled | Image-Line documents a Safe mode in the extraction dialog. Test a short clip and consult its support instructions if the issue persists. |
| Audio sounds doubled | Check whether the original clip is playing alongside the extracted tracks. Compare the source and stems separately before combining them. |
For a damaged optional-feature download, Image-Line's troubleshooting guide documents its repair controls. Follow the instructions for your installed version; preserve your project and user data before changing or removing files. The Limit CPU usage option can leave more resources for other tasks, but the manual says FL Studio itself is unavailable for editing during the extraction.
FL Studio and Saturalabs: different ways to separate
FL Studio's documented workflow can create several selected categories in one extraction. Saturalabs instead takes a description of one audible source per pass and returns Isolated sound plus Background audio. This can be useful when you want to test a specific sound rather than a fixed category, but separate requests are not an automatic four-stem export.
To try it, upload a supported MP3, WAV, MP4 or MOV up to 50 MiB, describe one source such as “the electric bass guitar,” and audition both outputs. Choose the remaining mix for a bass-reduced practice track. AI WAV downloads require an account and sufficient credits, with the cost shown before download. Results can contain bleed, miss notes or alter effects; repeat from the original rather than stacking processing on an earlier result.
Are these the original production stems?
No. Both workflows estimate sound sources from a finished mix. They cannot guarantee the same tracks, effects or quality as the original project. If precise control is important, ask for the producer's exports. See the Logic Pro Stem Splitter guide for a different editor's workflow and hardware requirements.