How to Prepare Stems for a Mixing Engineer: 2026 Export Checklist
How to export and send stems for mixing: file format, naming, dry vs wet, vocals, references and notes. Step-by-step for Logic, Ableton and FL Studio.
Horia Stan is a music producer and sound engineer at The One Records in Bucharest, Romania, who mixes and masters songs for artists worldwide, remotely and in writing.
To prepare stems for a mixing engineer, export every track as its own lossless WAV file. Use the same sample rate and bit depth for all of them, and start every file at the same point. Name every file clearly and send a rough mix plus written notes with them. Most of the quality of a remote mix is decided before the engineer opens a single file.
Quick answer
- Format: 24-bit WAV, at the same sample rate as your session. Every file starts at bar 1 or 0:00, so they line up in any DAW.
- Content: One file per track, with no master bus processing. Send vocals dry, and add a wet print only if the effect is part of the sound.
- Levels: Nothing clipping and no normalizing. Keep the balance of your rough mix if the engineer asks for it.
- Extras: A rough mix bounce, the BPM and key, one or two references, and a short written brief.
Why stem prep decides the mix
A mixing engineer can only work with what you send. A stereo bounce gives them very little to work with. Badly named, misaligned or clipped stems cost the first hours of the job, before any mixing happens.
There is also money involved. Most engineers price per song and include a fixed number of revision rounds. If the first round goes on fixing a missing backing vocal or an offset kick, that round is gone. If you are still comparing prices, read how much mixing and mastering costs in 2026 first.
The full stem checklist
| Item | What to send | Why |
|---|---|---|
| File format | WAV (or AIFF), 24-bit | Lossless, opens in every DAW |
| Sample rate | Same as your session, same for every file | Mixed rates cause pitch and timing errors |
| Start point | Every file starts at bar 1 or 0:00 | Drag and drop alignment in any DAW |
| Length | Every file runs to the end of the song | No missing tails |
| Naming | Number + instrument + part, e.g. "07 Lead Vox Verse" | No guessing what "Audio 14" is |
| Levels | No clipping, normalize off | Clipped stems cannot be fixed later |
| Master bus | Bypassed | The engineer builds the final bus |
| Rough mix | One stereo bounce of your current mix | Shows your intended balance and vibe |
| Info | BPM, key, tempo changes | Needed for delays, tuning and edits |
| Brief | Written notes plus 1-2 references | Defines what "done" means |
Sonarworks and Soundfly give the same core advice. Consolidate every track into one continuous file, keep the format consistent, clean up noise, and ask the engineer what they want before you export.
Dry or wet stems?
This question causes more back and forth than anything else on the list.
Dry stems have no effects. The engineer gets full control over EQ, compression, reverb and delay.
Wet stems keep your processing printed in. That matters when the effect is part of the sound. A distorted 808, a synth that only makes sense through its delay, or a vocal chop built from a reverb tail are all examples.
My rule of thumb:
| Source | Send dry | Send wet | Send both |
|---|---|---|---|
| Lead vocal | Yes | Only as reference | If the chain is part of the sound |
| Backing vocals, doubles, ad-libs | Yes | No | Rarely |
| Synths and pads | No | Yes | If the effect is creative, not corrective |
| 808 and bass | No | Yes, if saturation is the tone | Often useful |
| Drums | Yes | No | If you used parallel processing as the sound |
| Sound design, FX, risers | No | Yes | No |
When in doubt, send both and label them clearly. "07 Lead Vox DRY" and "07 Lead Vox FX PRINT" take two minutes to export. Rebuilding an effect you forgot to describe takes much longer.
Vocals need extra care
Vocals sell the song. They are also where stem prep most often goes wrong.
- Export every vocal track separately. Lead, doubles, harmonies, ad-libs, whispers. Do not merge them into one "vocals" file.
- Comp first. Send the final comped take, not 12 playlist takes. Delete unused takes, as Sonarworks recommends. If you want the engineer to choose between takes, say so in writing.
- Say whether tuning is done. If the vocal is already tuned, send a tuned version and, ideally, an untuned one. If tuning and editing are part of the job, say so before the price is agreed, because it takes time. My vocal tuning cost breakdown explains why.
- Mark the creative effects. A telephone filter on one line or a reversed reverb into the chorus: write down the timestamp.
If you are after the tight, heavily edited modern pop vocal, with pitch locked, doubles aligned and harmonies stacked, the stems matter even more. That sound is built from clean, separate, well-labeled vocal files. I explain how it works in how the polished Romanian pop vocal sound is made.
How to export stems in Logic Pro, Ableton Live and FL Studio
Logic Pro
Use File > Export > All Tracks as Audio Files. Apple's guide covers the dialog. File names come from the track names, so rename the tracks before you export.
Two common mistakes come up in third-party guides. An active track or region selection can limit what gets exported. A metronome left on can end up in the files. Clear the selection and turn the click off first.
Logic's default exports do not include effects on the stereo output. That is what you want for stems. Bounce the rough mix separately with File > Bounce > Project or Section.
Ableton Live
In the Export Audio/Video dialog, set Rendered Track to "All Individual Tracks". According to the Ableton manual, this renders the post-fader signal of each track, including Group Tracks. Return and Main tracks get their own files, and every file has the same length. That makes alignment trivial.
There is an option called "Include Return and Main Effects". Leave it off for dry stems. Turn it on only if you deliberately want wet prints. The same dialog has a Normalize switch. Keep it off, so your relative levels survive the export. Set Bit Depth to 24.
FL Studio
Route every instrument to its own mixer insert first. Then go to File > Export > Wave file and enable "Split mixer tracks" in the render window. FL writes one file per mixer track.
Do not skip the routing step. Several instruments sharing one insert will come out as one stem. Users on the Image-Line forum have also reported edge cases with channels routed into other channels. Play every exported file once before you upload.
Any DAW: the 60-second check
- Import all stems into a new, empty project at the same sample rate.
- Put every fader at 0 dB and press play.
- Compare against your rough mix. Balance aside, it should be the same song: no missing parts, no offset, no clicks at region edges.
If that test passes, your stems are ready.
Levels, headroom and the rough mix
You do not need to hit a specific number. You need two things.
- Nothing clipping. If a track hits 0 dBFS in your session, the overload is printed into the file. Pull the fader down, or lower the gain before the plugins, and export again.
- No normalize. Normalizing every stem to the same peak wipes out the balance you built.
Some engineers want stems exported at your rough mix fader levels. Others prefer everything at unity. Ask. Soundfly makes the same point: the answer depends on the engineer, so confirm it before exporting.
The rough mix is not optional. It is the clearest brief you can give. It shows which elements you consider the hook, how loud you hear the vocal and where the drop hits. These are things words describe badly.
The written brief
Keep it short and specific. This is the template I would want to receive:
| Field | Example |
|---|---|
| Song title and artist | "Title" - Artist Name |
| BPM and key | 92 BPM, F minor, no tempo changes |
| Release target | Spotify and Apple Music single |
| References | 1-2 released songs, with what to take from each |
| Must keep | The lo-fi drum loop sound, the vocal chop in the intro |
| Must fix | Chorus feels small, vocal gets lost in bridge |
| Vocal status | Comped, not tuned, tuning wanted |
| Stem notes | Synths printed wet, vocals dry, FX folder for risers |
| Deadline | Release date, not "ASAP" |
References need a reason. "Like this song" is vague. "The vocal level and dryness of this one, the low end of that one" is useful. My reference track method explains how to choose them.
A clear brief also speeds up revisions. Each round of notes can be checked against the brief, not against a shifting memory of what the song "should feel like". If you are still choosing who gets these files, these 12 questions cover what to ask before you pay.
Folder structure and delivery
Zip one folder per song:
Song Title - Stems(all WAV files, numbered)Song Title - Rough Mix.wavSong Title - Notes.txt(or the brief pasted into your email)References(links are fine, files are optional)
Upload to any cloud drive that keeps the original files untouched, and send one link. Do not send stems one by one over a messaging app, because many apps compress audio. If you also want mastering after the mix, the delivery rules for that step are different. I cover them in how I prepare mixes for mastering.
When doing it yourself is fine, and when to hand it off
Exporting stems is a job anyone can do. Follow the checklist, run the 60-second check, write the brief, and any competent engineer can start immediately.
Mixing the song yourself makes sense when it is a demo, a sketch for your audience or a release you treat as practice. It also makes sense when you have good monitoring and time to learn. If you are unsure which stage your song actually needs, mixing vs mastering breaks it down.
Hand it off when the song is going on streaming platforms and has to stand next to the references you picked. The same goes when the vocal needs serious editing, or when you have listened to your own mix so long that you cannot hear it anymore.
I mix and master remotely for artists worldwide, fully async and in writing. You send the stems and the brief, I send back the mix, and the notes go back and forth by email until it is right. Prices start at โฌ500 per song. Average turnaround is 2-3 days. Two revision rounds are included. Vocals in any language, from heavily edited modern pop to a natural live feel. If your stems are ready, send the brief here. For the remote process step by step, read online mixing and mastering for international artists.
Frequently Asked Questions
What format should stems be in for mixing?
Send 24-bit WAV files at the same sample rate as your session, with one file per track. Every file should start at the same point, bar 1 or 0:00, and run to the end of the song. AIFF works too, but WAV is the safest choice across operating systems and DAWs. Never send MP3 or other lossy files as mixing stems.
Should I send dry or wet stems to a mixing engineer?
Send vocals and drums dry. Send wet prints for anything where the effect is part of the sound, such as distorted bass, synths built around a delay, or sound design. When you are unsure, export both versions, label them DRY and FX PRINT, and let the engineer choose.
Should I leave effects on the master bus when exporting stems?
No. Bypass the master bus when you export stems, because the engineer builds the final bus processing. Bounce one separate rough mix with your master chain on, so they can hear the result you were aiming for.
How should I name stems for a mixing engineer?
Use a number, the instrument and the part, for example "01 Kick", "07 Lead Vox Verse" or "12 BV Harmony High". Numbering keeps the files in order in any DAW. Clear names save the engineer from guessing what "Audio 14" is.
Do I need to send a rough mix with my stems?
Yes. The rough mix shows the engineer your intended balance, which elements are the hook and how loud you hear the vocal. These are things that are hard to describe in words. Send it as one stereo WAV bounce, together with BPM, key and one or two reference tracks.
Should stems be normalized or exported at a specific level?
Do not normalize stems, because it erases the balance between tracks. Make sure nothing clips in your session. Ask the engineer whether they want stems at your rough mix fader levels or all at unity, because preferences differ.
How do I export all tracks as stems in Logic Pro?
Use File > Export > All Tracks as Audio Files. Rename your tracks first, because the file names come from them. Clear any track or region selection, turn the metronome off, and choose WAV at 24-bit. Bounce the rough mix separately with File > Bounce > Project or Section.
Continue reading
AI Mixing vs a Mixing Engineer in 2026: RoEx, Neutron or a Human?
AI mixing vs hiring a mixing engineer in 2026: what RoEx Automix, Neutron 5 and BandLab AutoMix really do, what they cost, and when a human is worth it.
Stem Mastering vs Stereo Mastering: Which One Do You Need?
Stem mastering vs stereo mastering: what each one is, real prices, how many stems to send, and when stems are worth it or just hide a mix problem.