Back to blog

How to Export Stems in Ableton, Logic Pro & FL Studio (2026): A Step-by-Step Guide

Short answer: to export stems, render each track or group separately as a 24-bit WAV at your session's sample rate, with every stem starting at bar 1 and ending at the same point, and with the master-bus chain bypassed. In Ableton Live that is File > Export Audio/Video with Rendered Track set to All Individual Tracks. In Logic Pro it is File > Export > All Tracks as Audio Files. In FL Studio it is File > Export > Wave file with Split mixer tracks enabled. Everything else in this guide is detail on those three menu items and the settings around them.

A stem is a single rendered audio file containing one part of your arrangement: the drums, the bass, the lead vocal, the reverb return. Stems matter because they are the unit of exchange in every collaboration you will have. A mix engineer, a mastering engineer, a remixer, a sync licensor and an analysis tool all want the same thing, and they all reject deliveries for the same handful of reasons.

The four rules that cause almost every rejected stem delivery

Before any DAW-specific instructions, these are the errors that actually get stem sets sent back. Get these right and the menu path barely matters.

  1. Every stem starts at zero and has the same length. This is the big one. If your vocal does not enter until bar 17, the vocal stem must still start at bar 1 with 16 bars of silence in front of it. Stems that begin where the audio begins cannot be lined up again without guesswork, because nothing in the file records where it was supposed to sit. Export the full arrangement range for every stem, including any reverb or delay tail at the end.
  2. Bypass the master-bus chain. If you have a limiter, compressor or EQ across the 2-bus, take it off before rendering stems. Otherwise every stem is individually processed by a chain that was designed to act on the sum, and the stems will not add back up to your mix. Keep processing that belongs to the individual track: the vocal compressor stays, the mix-bus limiter goes.
  3. Turn normalisation off. Normalisation rescales each file independently, which destroys the balance between stems. Every DAW has this setting somewhere in the export dialog and it must be off.
  4. Match the session sample rate and use 24-bit. Do not resample on export and do not drop to 16-bit. 24-bit costs nothing but disk space and preserves the low-level detail that later processing will lift.

One thing you do not need to worry about is leaving a fixed amount of headroom on each stem. As long as no individual stem clips, the level it peaks at is not important, for the same reason the -6 dB headroom rule is folklore: whatever receives these files will change the gain first, and digital gain is free.

How to export stems in Ableton Live

  1. Set the export range. Select the whole arrangement, from the start to past the last reverb tail. Ableton exports the current selection or loop brace, so a short selection is the most common cause of truncated stems.
  2. Open File > Export Audio/Video.
  3. Set Rendered Track to All Individual Tracks. This renders every track in the set as its own file. Note that return tracks are rendered as separate files too, so your reverb and delay returns arrive as their own stems, which is correct and is what a mix engineer expects.
  4. Check the render settings. Sample Rate matches the set, Encode PCM on, File Type WAV, Bit Depth 24. Set Normalize to Off, and leave Render as Loop off.
  5. Render, then check the folder. You should have one file per track and return, all the same length.

The Ableton-specific trap is track-level effects on frozen or grouped tracks. Rendering All Individual Tracks gives you the individual tracks inside a group, not the group bus, so any processing sitting on the group is not included. If your group bus is doing real work, either render group tracks instead or flatten that processing onto the individual tracks first.

How to export stems in Logic Pro

  1. Select all regions, or set the cycle range to the full arrangement. As in Ableton, the export follows the range, so make sure it covers the tails.
  2. Open File > Export > All Tracks as Audio Files. If you only want a subset, select those tracks and use Selected Tracks as Audio Files instead.
  3. Set the format. WAV, 24-bit, and the session sample rate. Set Normalize to Off.
  4. Decide on the two checkboxes that matter. Bypass Effect Plug-ins should be unchecked, so stems sound the way you hear them. Include Volume/Pan Automation should be checked if you want your automation printed in, which is normally what you want for a mastering or analysis delivery.
  5. Check the length of the exported files. Logic names files after the track, so a quick sort by size will show you any stem that came out short.

Logic's own terminology is worth knowing: if your session is built on Track Stacks, exporting all tracks gives you the sub-tracks rather than the stack. For a stem set organised by section, export the stacks, or flatten each stack to a single track first.

How to export stems in FL Studio

  1. Open File > Export > Wave file. Choose the destination folder first; FL writes the split files into a subfolder there.
  2. Enable Split mixer tracks. It is in the Miscellaneous section of the render dialog. This is what turns one render into one file per mixer track.
  3. Set the format. WAV, 24-bit (or 32-bit float if you prefer), session sample rate, and make sure normalisation is off.
  4. Set the range to the full song, including tails. Use the Leave remainder option so delay and reverb tails are not cut off at the last bar.

The FL-specific trap is routing. Split mixer tracks splits by mixer insert, not by channel rack pattern, so anything routed to the same insert is summed into one stem. If you want the kick and the snare separate, they need separate mixer inserts before you render. Master-bus effects are not applied to the split files, which is the behaviour you want here.

Naming and packaging your stems

  • Name by content, not by track number. Kick.wav, Bass.wav, LeadVocal.wav. A folder of Audio 1 through Audio 24 makes whoever receives it do work you could have done in ten seconds.
  • Put the tempo and key in the folder name. SongTitle_128bpm_Amin is instantly useful to a remixer and costs nothing.
  • Zip the folder. A single archive cannot arrive partially, which is the failure mode of dragging thirty loose files into a browser or an email client.
  • Include a rough mix. One stereo bounce of the full mix, named clearly, so the recipient can hear the intended result and check that the stems sum to it.

Frequently asked questions

What is the difference between stems and multitracks?

Multitracks are every individual channel in the session, often dozens of them, including each drum microphone and each vocal take. Stems are grouped renders: one drum stem, one bass stem, one vocal stem. In practice the words are used loosely and people asking for stems often want multitracks, so it is worth confirming which one is expected before you spend an hour rendering.

Should stems be mono or stereo?

Export stems with the channel format the source uses. A mono bass track can be delivered mono, and a stereo synth must be stereo. If in doubt, deliver everything as stereo files; the extra size is trivial and a mono source rendered to a stereo file is unchanged.

Do I need to export stems for mastering?

Usually not. Conventional mastering works from a single stereo mix. Stems are for stem mastering, for handing the mix to someone else to finish, or for analysis that needs to see the parts separately. If you are simply getting a finished mix mastered, one stereo file is what is wanted.

What sample rate and bit depth should stems be?

Whatever the session already is, at 24-bit. Do not convert on export. Sample rate conversion is a real process with real artefacts, and it should happen once, deliberately, at the end of the chain rather than silently in a render dialog.

Why do my stems not add up to my mix?

Almost always master-bus processing. If a limiter or bus compressor was active during the render, each stem was processed as though it were the whole mix, and the sum will be louder and more distorted than the original. The other common cause is normalisation being left on.

What to do with the stems once you have them

Stems are the format that makes a mix diagnosable. A stereo bounce tells you the low mids are crowded; the stems tell you which part is crowding them. That distinction is why analysing your own exported stems is a different thing from source separation, which reconstructs approximate parts from a finished song. Your exported stems are the real signals, so anything measured from them is exact rather than inferred.

Mozonic takes stem sessions directly: upload WAV, AIFF or MP3 files loose, or a ZIP of the whole folder. Stems are sorted automatically into roles such as Vocals, Drums, Bass, Guitar, Keys, Synths, Strings, FX and Percussion, with a confidence label so you can correct anything it reads wrong. ZIP is recommended once a session goes beyond about 75 files or 1 GB. Stem sessions and stem DSP are Pro features.

If you only want to check the mix as a whole first, that does not need stems at all: upload a stereo bounce and you get true peak, integrated loudness, dynamic range, stereo width and a plain-language list of what is standing between the track and release-ready, measured against the platform targets each streaming service actually uses. Five analyses a month are free and no card is required.