How to use Stem Separation in Ableton Live 12

Stem Separation is a built-in Ableton Live 12 tool that splits a finished, mixed audio file into four stems: Vocals, Drums, Bass and Others. You right-click a clip, choose Separate Stems to New Audio Tracks, pick which stems you want and a quality mode, and Live puts each stem on its own audio track. It runs entirely on your computer and arrived in Live 12.3, in Live 12 Suite and on Push 3 Standalone.

This guide walks through the process, explains the two quality modes, and shows a sampling workflow: chopping a bass line and a vocal hook from the separated stems in Simpler.

What you need

If you own Live Standard, the menu item won't appear. Upgrading to Suite is the only built-in way to get it. Ableton confirmed the Suite and Push 3 requirement when it announced Live 12.3.

How to separate stems in Ableton Live

1. Drop the track into your Set. Drag the audio file onto an audio track in Arrangement or Session View. You can also start from the Browser.
2. Open the context menu. Right-click the clip (Ctrl-click on a Mac trackpad) and choose Separate Stems to New Audio Tracks. The same command is also in the Create menu.
3. Choose your stems. The dialog lists Vocals, Drums, Bass and Others. Untick any you don't need. If you only want the vocal, there's no reason to generate the rest.
4. Pick a quality mode. Choose High Quality when the stems will be heard in a mix and High Speed when you're auditioning ideas (see the next section).
5. Click Separate. Live processes the audio and adds one new audio track per stem, lined up with the original clip.
6. Solo each stem and listen. Check for bleed before you commit to a sound. More on that below.

To separate only part of a song in Arrangement View, make a time selection across the clip, right-click and choose Separate Stems for Time Selection. It's quicker than processing a six-minute track for eight bars of vocal.

High Quality vs High Speed

The two modes aren't just faster or slower versions of the same process. They work differently:

• High Speed separates all four stems in a single pass. Use it to decide whether a track is worth sampling at all.
• High Quality gives Vocals, Drums and Bass a dedicated pass each. The Others stem is built from whatever audio is left after those three. It takes longer, and the stems are noticeably cleaner.

A reasonable habit: use High Speed while you're digging, then re-run the section you've chosen in High Quality before you build a track on it.

Where Ableton saves the stems

Separated stems are written into your current project folder, under Samples/Processed/Stems. They're stored at 44.1 kHz with the same bit depth as the source file, according to the Stem Separation chapter of the Live 12 manual.

Because they're inside the project folder, the stems travel with the Live Set. If you move or share the project folder, the stems go with it. If you lose the folder, you lose them too, along with every chop you've made from them. If you're moving a Set to another computer, run Collect All and Save first so every sample it uses is inside the folder too.

Sampling a bass line from the stem

The video's most useful idea is to treat the separated bass as a finished, sound-designed patch you can play instead of building one from scratch.

1. Load an empty Simpler on a new MIDI track and drag the bass stem onto it.
2. Switch Simpler to Slice mode. It cuts the stem at each note so every slice sits on its own key.
3. Play the slices from your keyboard and find the one you like.
4. Tighten the slice. Adjust its start point and the fade-in and fade-out so it doesn't click.
5. Add character. In the video the chain is a delay for stereo width, Roar for drive and Utility for a gain boost.
6. Record a part, then clean it up. Select all notes (Cmd+A / Ctrl+A), quantize (Cmd+U / Ctrl+U) and push the velocities to maximum so every hit lands evenly.
7. Retune it to your track. Transpose the slice until it sits in key. In the video it took +7 semitones.

Making vocal chops from the vocal stem

The same Simpler approach works for vocals, with one important setting. For more ways to slice, warp and process a vocal, see the full vocal chops guide.

1. Drag the vocal stem into an empty Simpler and switch to Slice mode.
2. Lower the Sensitivity. At the default, Simpler cuts a vocal into tiny fragments. Turn Sensitivity down until each slice holds a whole word or phrase.
3. Shape the slices. Move slice start points so a phrase splits where you want it, then trim the fades.
4. Build an effects chain. A starting chain from the video: Echo for delay, a pitch shift down 5 semitones, Auto Pan-Tremolo for tremolo movement and Auto Filter to remove low end so the chop sits above the mix.
5. Record, quantize and duplicate the phrase so it fills the section.

For a second texture, drop the same stem into another Simpler and leave Sensitivity at 100. You'll get short, single-word stabs that work as rhythmic hits over the longer chop.

Getting cleaner stems

Separation is very good, but not perfect. Frequencies sometimes land in the "wrong" stem, and Ableton's manual says so directly. Typical examples are the top end of a synth bass drifting into Others, or reverb tails from the vocal leaking into the other stems.

• Use High Quality for anything that will end up in a finished track.
• Separate a shorter section. A time selection around the part you need is faster and easier to judge.
• Check the Others stem before deleting it. The missing top end of your bass may be in there. Layer it back quietly if the bass sounds dull.
• Filter the bleed. A high-pass on a vocal stem or a low-pass on a bass stem removes most leftovers.
• Commit early. Once a chop works, resample or freeze it so you're not relying on the full stem track.

Other ways to use Stem Separation

• Reference mixing. Solo a pro track's drums or bass and compare them with yours, level for level.
• Studying sound design. Hear a lead or bass with nothing else around it.
• Remixing and edits. Pull an acapella or instrumental from a track you have the rights to use.
• Practice and transcription. Mute the part you're learning and play along.
• Audio to MIDI. Run Live's audio to MIDI conversion on an isolated stem to get a cleaner transcription of the melody, harmony or drums.

A quick note on rights: separating stems is a technical process, not a license. Releasing music built on stems from someone else's recording still needs clearance.

Frequently asked questions

Is Stem Separation available in Ableton Live Standard?

No. Stem Separation is included in Live 12 Suite, from version 12.3, and on Push 3 Standalone. Intro, Lite and Standard don't have it.

Does Ableton's stem separation need an internet connection?

No. Everything is processed locally on your computer, and you don't need to install any other tools. The technology comes from Music AI, the company behind Moises, and runs inside Live.

Where does Ableton save separated stems?

In your current project folder, under Samples/Processed/Stems, at 44.1 kHz with the source file's bit depth.

Which stems can Ableton separate?

Four: Vocals, Drums, Bass and Others. Others holds everything that isn't classified as one of the first three, such as synths, keys, guitars and effects.

What's the difference between High Quality and High Speed?

High Speed separates all stems in one pass. High Quality runs a separate pass each for vocals, drums and bass, then builds Others from what's left. High Quality is slower but produces cleaner stems.

Can I separate just part of a clip?

Yes. In Arrangement View, make a time selection, right-click and choose Separate Stems for Time Selection.

Why is some of my bass or vocal in the Others stem?

The algorithm sometimes puts frequencies in a different stem than you'd expect, depending on the source material. Try High Quality mode, and check the Others stem for the missing part before you delete it.