How to Prep Your Stems for a Mixing Engineer
You’ve spent weeks on this song. The writing is done, the performances are tight, and you’re ready to hand it off to a mixing engineer. Then you drag a folder into a WeTransfer window and feel a small wave of doubt: Did I export this right?
The way you prep your stems can save your engineer hours of detective work, or create them. It also affects your budget, your timeline, and how good the final mix sounds. Here’s how to get it right.
First, what do we mean by “stems”?
People use the word loosely, so let’s get specific.
Multitracks are every individual track in your session: kick, snare, each guitar take, each vocal layer. This is usually what a mixing engineer wants most.
Stems are grouped submixes, like “Drums,” “Bass,” “Guitars,” “Vocals,” and “Synths.”
If you’re sending a mix engineer your song, ask which they prefer. Most will say multitracks, and some will accept stems if the session is huge or full of virtual instruments. This guide covers both, since the same principles apply.
Rule 1: Export at the right quality (and don’t downgrade)
Export at the same sample rate your session is running at, and use 24-bit WAV or AIFF. Here’s why that matters:
16-bit audio has a theoretical dynamic range of about 96 dB.
24-bit audio stretches that to about 144 dB.
That extra range gives your engineer far more room to turn things up, compress, and process without raising the noise floor. 16-bit is fine for the final distribution file, but it’s not what you want to hand off for mixing.
Don’t convert sample rates on export. If you recorded at 48 kHz, send 48 kHz. Converting up (say, 44.1 to 96 kHz) adds nothing but file size. Converting down throws away information.
And never send MP3s or other lossy files as stems. They can’t be un-compressed, and artifacts get worse the more processing is applied.
Rule 2: Make every file start at the same place
This is the single biggest time-saver, and the one people forget most often.
Every exported file should start at the exact same point, usually bar 1 or 0:00, and run to the same end point, even if that track is silent for the first two minutes. When all files are the same length and aligned, your engineer can drag them all into a session and everything lines up instantly. When they’re trimmed or exported from different starting points, someone has to manually figure out where things go.
Most DAWs have an “export all tracks” or “bounce in place” option that handles this. Set your loop or export range once, and run it.
Rule 3: Leave room (and don’t normalize)
Check your levels, but don’t chase loudness.
Aim for peaks around -6 dB on each track, and make sure nothing clips.
Turn off normalization on export. It rescales each file to a different gain, so the relative levels you balanced will be changed.
Remove any limiter or heavy compression from the master bus. Your engineer wants to apply their own, and a limiter that’s already squashing the mix can’t be undone.
The goal is clean, healthy signal with headroom, not a loud file.
To put numbers on it: 32-bit float files have a theoretical dynamic range far beyond anything you’d ever need (around 1,500 dB), which is why they can’t clip internally. If your DAW can export 32-bit float, it’s a safe option, but 24-bit is the widely accepted standard and every engineer can open it.
Rule 4: Print the effects that define the sound, and leave the rest
This one needs some judgment.
Print (bake in) effects that are part of the sound itself:
Guitar amp simulations and distortion
Creative synth processing or modulation
Vocal effects that are the creative signature, like a heavy vocoder or a bitcrusher
Sidechain pumping that’s essential to a track’s feel
Leave off (or send separately) effects that are about mixing:
EQ and compression meant to “fix” or “polish” a track
Reverb and delay, unless they’re a creative must-have
Master bus processing
If you’re unsure, export two versions: one with effects and one dry. It costs you a few more files and gives your engineer options. For time-based effects like reverb and delay, many people print them on their own separate stem so the engineer can adjust the amount without losing the original sound.
Rule 5: Name things like a human will read them
Your engineer might open 40 to 80 files. Make it easy.
Bad: Audio 12_bounce_FINAL_v3.wav
Good: 07_Lead_Vocal_Verse.wav, 12_Bass_DI.wav, 15_Kick_In.wav
A few tips:
Use numbers at the start so files sort in a logical order.
Include the instrument and what it is (e.g., “Gtr_Rhythm_L” and “Gtr_Rhythm_R” for a double-tracked pair).
Mark stereo vs. mono clearly. Don’t send a mono guitar as a stereo file with identical channels. It doubles the file size for no benefit.
Avoid special characters and spaces that can break on certain systems. Underscores are your friend.
Rule 6: Include a reference mix and notes
This might be the most underrated step. Send:
A rough mix of your own, bounced from your session. It tells your engineer where you were heading: the balance, the vibe, the energy.
A short note covering the tempo (BPM), key, and any reference songs you love the sound of.
Specific requests or concerns, like “the second chorus feels flat” or “I want the vocal upfront and intimate.”
Include a note on anything unusual too, such as a track that’s intentionally distorted, a sample that sounds weird on purpose, or a part that’s meant to be buried. If you don’t explain it, a good engineer may “fix” it.
Rule 7: Check your files before you send them
Before you upload, do a 5-minute sanity check:
Drag everything into a fresh session. Does it all line up from bar 1?
Play it back with all the faders at unity. Does it sound roughly like your rough mix?
Look for clipping, clicks, and pops. Zoom in on edits and fades.
Check for silent or empty files. An accidentally muted track export is more common than you’d think.
If you find issues now, you save a round of back-and-forth later.
A quick reality check on file size
Large sessions add up fast. Here’s the math for stereo vs. mono at 24-bit/48 kHz:
Mono: about 8.6 MB per minute
Stereo: about 17.3 MB per minute
So a 4-minute song with 40 mono tracks comes to roughly 1.4 GB. A session with 80 tracks, a mix of stereo and mono, can easily reach 3 GB or more. Use a service that handles large transfers (Dropbox, Google Drive, WeTransfer Pro, or your engineer’s preferred upload link), and zip the folder to keep it tidy.
Your quick checklist
24-bit WAV/AIFF at the session’s native sample rate
All files start at the same point and have the same length
No normalization, no master bus limiter, peaks around -6 dB
Creative effects printed, mix effects left off (or dry versions included)
Clear, numbered file names
Rough mix and a note with tempo, key, and references
Tested in a fresh session before sending
The bottom line
Prepping your stems well isn’t about being perfect. It’s about removing friction so your engineer can spend their time making your song sound incredible instead of sorting out your files. A little care on the front end means a better mix, a faster turnaround, and a happier collaboration.
Now go export something great.
Ready to Hear Your Song at Its Best?
You’ve done the hard part: you wrote it, performed it, and prepped your stems like a pro. Now let’s make it sound the way it did in your head.
We offer professional audio mixing and mastering for artists who care about every detail. Whether you’re releasing your first single or your tenth album, we’ll bring clarity, punch, and polish to your music while keeping what makes it yours.
Here’s what you get:
A mix built around your vision, guided by your reference tracks and notes
Mastering that’s release-ready for Spotify, Apple Music, and every major platform
Clear communication and a revision round so you’re happy with the result
Fast, friendly turnaround
Not sure if we’re a fit? Send a 30-second clip of your song and we’ll mix it for free so you can hear the difference before you commit.
👉 Book your project today or email us at info@spewmusic.com with your song details, and let’s get started.
Your music deserves to be heard. Let’s make it unforgettable.
Comments