Every few weeks somebody sends me a hard drive’s worth of session files for an Atmos mix, and we lose two days untangling it before I can mix a single note. So here’s the whole answer up front. To prepare your Dolby Atmos stems, I need four things: your stems at 48 kHz / 24-bit with all your processing printed, your stereo FX busses included, your mastered stereo mix as a reference, and labels a stranger could understand. That’s it. The rest of this article is why each one matters, because the why is where people usually go wrong.
Send your stems wet, not dry
This is the one that surprises people most. Artists assume an Atmos mixing engineer wants raw, untouched tracks so we can “do our thing.” We don’t. I want your stems POST-processing: every EQ move, every compressor, everything that made your stereo mix sound like your record. Your mix decisions are the record. My job in Atmos isn’t to remix your song, it’s to give the song you already made a bigger room to live in. When Dolby Atmos stems arrive dry, I end up guessing at a sound you spent months dialing in, and nobody wins that game.
Your FX busses are the secret weapon
Reverbs, delays, parallel compression: print those busses separately and include them. Here’s why this matters more in Atmos than anywhere else. In a stereo mix, your reverb tail fights the guitars for the same narrow window. In an Atmos room, that tail can live behind the listener, wrapped around them, anywhere. The FX busses are the most spatial material in your whole session, and they’re exactly the thing most people forget to send. A separate reverb bus gives me something I can place. A reverb baked into the vocal track gives me nothing to work with.
The mastered stereo mix is non-negotiable
I reference-check every Atmos mix against the released stereo version, constantly, all session long. Your fans know the record. If the chorus vocal sits differently in the Atmos version for no reason, they’ll feel it even if they can’t name it. The stereo master is my north star for balance, tone, and intent. I wrote more about how the two formats relate in stereo vs Atmos, but the short version: the Atmos mix should feel like the same record with the walls taken away, and I can’t hit that target if I can’t see it.
Labeling: the five minutes that saves five hours
Label every stem in a format like this: DRUMS, BASS – Electric, BASS – Direct Input, VOCALS – Lead, VOCALS – Harmonies, GTR – Rhythm Left. Not “Audio_04_final_v2_FINAL.” I’ve genuinely received that file. Twice. Every file should also start from the same point (bar one, beat one), so everything lines up the moment I drop it in the session. Clear labels mean I spend the session mixing your music instead of playing detective, and you’re paying for mixing, not detective work.
The boring specs that aren’t optional
48 kHz, 24-bit, WAV. Atmos delivery pipelines run at 48 kHz natively, so sessions recorded at 44.1 get resampled somewhere along the chain, and I’d rather that happen deliberately than accidentally. If your session lives at 44.1 or 96, tell me before you export and we’ll sort out the cleanest path. But don’t just hit bounce and hope.
Getting the files to me
Zip the whole thing and name the folder something a stranger could read: artist name, song title, stems. Send it however you like, Dropbox, Google Drive, WeTransfer, a drive in the mail if the project is big enough. A full album of printed stems with FX busses can run past a hundred gigabytes, so start that upload the night before rather than the morning your deadline lands.
Give each song its own folder. If you’re sending four songs and every one of them contains a file called VOCALS – Lead, something is going to overwrite something.
And send a short note with it. Which take is the keeper, which bus is doing something unusual, whether that buried bridge vocal is buried on purpose. Thirty seconds of context from you beats an hour of me guessing.
What you get back
When the mix is approved, your deliverables include the Atmos master as an ADM BWF file (the format Apple Music, Amazon, Tidal, and Avid Play all accept), plus 7.1, 5.1, and binaural headphone versions. Everything is delivered at -18 LUFS, the loudness standard Dolby and Universal Music Group both recommend, so your release passes QC the first time instead of bouncing back from the distributor. If you’re wondering what all this costs, I published real numbers here.
The questions I get after people read this
What if my mix bus chain is part of the sound?
Then print it into the stems, or send both versions. If your bus compressor is doing something you love, run the stems through it so that character comes along. If you’re not sure, send them both ways and tell me which is which. Extra files cost you nothing. Guessing costs a day.
Do you need my session file or tempo map?
Usually not. Printed stems that all start at bar one are enough for most songs. But if there are tempo changes, or you want me to hit a specific edit, send the tempo map. It takes you a minute and it saves me reverse engineering it.
What if I’m missing one stem?
Tell me before you send, not after. A missing stem caught on day one is a re-export. The same stem caught on day two is a re-export plus everything I built on top of the wrong balance.
Can you work from just my stereo master?
Not for a real Atmos mix, no. Upmixing a stereo file is a different service and it sounds like exactly what it is. The whole point of the format is placing things that were never separate in stereo, and I can’t place what I can’t reach.
How many stems is too many?
There’s no count that scares me. Atmos gives us 128 audio elements to work with, so a 60 or 80 stem session is normal. What matters more is that they’re grouped sensibly. Forty separately printed percussion hits aren’t more useful to me than one well-built percussion bus.
My session is stuck at 44.1 kHz. Now what?
Tell me before you export. If re-exporting at 48 isn’t realistic, I’ll handle the conversion once, deliberately, in one place. What I want to avoid is the same file getting resampled three times by accident on its way through the chain.
The checklist, one more time
Stems at 48 kHz / 24-bit with processing printed. Stereo FX busses included. Mastered stereo reference. Labels that make sense, all files starting from bar one. Do those four things and your Dolby Atmos stems will drop into my session ready to mix, your first playback happens a day sooner, and your budget goes into the mix instead of the cleanup. And once the mix is signed off, Dolby Atmos mastering and QC takes it the rest of the way to release. Got a session that’s almost ready but you’re not sure about one of these? Ask me before you export. That question costs nothing and regularly saves a day.
Todd Hooge is the founder of Sonic Forest Studios and mixes in Dolby Atmos full-time. His immersive work has over 448 million verified streams.