You finished the mix at 11:47pm. The client signed off an hour ago. Now you’re staring at the export window in your DAW and asking yourself the same questions every engineer asks at this moment of final mix delivery: WAV or AIFF? 24-bit or 16-bit? 44.1 or 48? One file or three? Do they want stems? Did anyone mention an instrumental? What sample rate did they ask for again?
This is the part of the job nobody teaches. There are 400 YouTube tutorials on EQing a kick drum. There are maybe four blog posts on what to actually send the client when the mix is done. So most engineers wing it. They export one stereo WAV, attach it to an email, and hope it’s what the client wanted.
It usually isn’t.
Final mix delivery is where most engineers leak professionalism. Not in the mix itself, which is great. Not in the communication, which has been fine. In the last 15 minutes of the project, where a sloppy file naming convention or a missing instrumental version makes the whole engagement feel less polished than it actually was. This post is the full playbook for what to send, what format to send it in, and how to deliver it so the project lands cleanly instead of dribbling out across four follow-up emails.
What “Final Mix Delivery” Actually Means
Most clients don’t know what they want from final mix delivery because they’ve never had to specify it. They hired you to mix the song. They assume the “mix” will arrive in whatever form makes sense. So the burden is on you to define what “makes sense” before you bounce a single file.
A complete final mix delivery is usually some combination of:
- The main stereo master (the song as it will appear on streaming services)
- An instrumental version (no lead or backing vocals)
- A TV mix (vocals only on the lead, everything else intact, used for live TV and karaoke licensing)
- An acapella (vocals only, no instruments)
- Stems or stem groups (drums, bass, vocals, instruments as separate files, useful for sync licensing and remixes)
- A reference MP3 (a lower-resolution version for the client to send around without leaking the master)
You don’t include all six on every project. Most independent releases need the first two. Sync placements need all five. Major-label projects often want stems on top of everything. The right answer depends on where the song is going, and your job is to ask before you bounce.
If you’re delivering anything other than a single stereo master, write it into your scope and contract before the project starts. Stem delivery in particular takes meaningful extra time (an hour to two hours on top of a mix, depending on the session) and should be priced as a line item, not absorbed.
The File Format Decision Tree
The format question is simpler than most engineers make it. There are really only three deliverable formats that matter for final mix delivery: streaming WAV, distribution WAV, and reference MP3.
Streaming WAV (24-bit, sample rate matches the session). This is the file that goes to a mastering engineer or to a service like DistroKid for streaming distribution. 24-bit because it preserves headroom for the mastering process. Sample rate matches the session (44.1kHz if you mixed at 44.1, 48kHz if you mixed at 48). Leave at least 3dB of headroom on the master bus. No master bus limiter unless the client specifically asked you to skip mastering and just deliver a finished file.
Distribution WAV (16-bit, 44.1kHz). This is the file that goes on CDs, gets uploaded to legacy services, or gets handed to a producer who wants a final-final version. Dither it down from your 24-bit master. This is also the file you’d send to a sync supervisor unless they specify otherwise.
Reference MP3 (320kbps, 44.1kHz). This is the file the client sends to their manager, posts in their group chat, plays in their car, listens to on the train. Always include it. The client will create their own version anyway by re-encoding your WAV in iTunes if you don’t, and the result will be worse than the one you’d have made.
That’s the entire format question for 95% of mixes. The exceptions are vinyl pre-masters (don’t dither, leave more headroom, talk to the cutting engineer), Atmos and surround deliveries (different file structure entirely, only relevant if you’ve been hired for an immersive mix), and broadcast deliveries (loudness-normalized to specific LUFS targets, region-dependent). If you’re delivering one of these and don’t know what you’re doing, ask the client what spec they need and get it in writing before you bounce.
File Naming: The Convention That Saves Everyone
Most final mix delivery problems are file naming problems. A client opens your delivery folder and sees:
song.wav
song_v2.wav
song_inst.wav
song_INST_V2.wav
song_final.wavThey have no idea which one is the actual final. They guess. They guess wrong. They upload the wrong file to DistroKid. They call you panicking three weeks later when their release sounds different from what you delivered.
A clean file naming convention prevents all of this. The format that works across genres and clients:
[ArtistName] - [SongTitle] - [VersionType] - [Date] - [SampleRate-BitDepth].wavSo a final main mix for an artist called Lyra delivered on May 25, 2026 at 24-bit/48kHz becomes:
Lyra - Easy Money - MAIN - 2026-05-25 - 48k-24b.wav
Lyra - Easy Money - INSTRUMENTAL - 2026-05-25 - 48k-24b.wav
Lyra - Easy Money - TV - 2026-05-25 - 48k-24b.wav
Lyra - Easy Money - ACAPELLA - 2026-05-25 - 48k-24b.wav
Lyra - Easy Money - MAIN - 2026-05-25 - REFERENCE.mp3That naming convention is verbose, and that’s the point. The client can read the filename from across the room and know exactly what they’re looking at. There’s no version ambiguity. There’s no question about format. If they email you in six months asking for the file, you can find it in five seconds because it sorts alphabetically by artist and chronologically by date.
Avoid version numbers in the filename of the final mix delivery. Use them during revision rounds (v1, v2, v3) but not in the final. Once a file is final, it doesn’t need a version number because it’s the only version that matters. Adding “_FINAL” or “_v7_FINAL” to the filename creates the same kind of versioning chaos the file name chaos post covers in detail.
What Goes in the Delivery Folder Structure
A single zip file with a clean folder structure beats five separate emails with attachments. Always.
The folder structure that works:
[ArtistName] - [SongTitle] - Final Delivery [Date]/
├── 01_Streaming_WAV/
│ ├── [ArtistName] - [SongTitle] - MAIN - 24b48k.wav
│ ├── [ArtistName] - [SongTitle] - INSTRUMENTAL - 24b48k.wav
│ ├── [ArtistName] - [SongTitle] - TV - 24b48k.wav
│ └── [ArtistName] - [SongTitle] - ACAPELLA - 24b48k.wav
├── 02_Distribution_WAV/
│ ├── [ArtistName] - [SongTitle] - MAIN - 16b44k.wav
│ └── [ArtistName] - [SongTitle] - INSTRUMENTAL - 16b44k.wav
├── 03_MP3_References/
│ ├── [ArtistName] - [SongTitle] - MAIN - 320kbps.mp3
│ └── [ArtistName] - [SongTitle] - INSTRUMENTAL - 320kbps.mp3
└── README.txtThe numeric prefixes (01, 02, 03) force the folders to sort in a deliberate order regardless of how the client’s operating system would normally sort them. The first folder the client opens is the most important one. The README file inside the root folder explains what’s there and what each file is for.
If stems are part of the final mix delivery, they go in their own folder labeled 04_Stems with each stem named clearly: Drums.wav, Bass.wav, LeadVocal.wav, BackingVocals.wav, Keys.wav, Guitars.wav. Stems should be the same length as the main mix (full song from downbeat to fade), bounced from the start of the session, and time-aligned so they play back in sync when imported into another DAW.
The README File Nobody Writes
The single most useful thing you can include in a final mix delivery is a short text file that explains everything. Most engineers don’t write one. The 30 seconds it takes to write a README saves 30 minutes of email back-and-forth later.
A useful README contains:
FINAL MIX DELIVERY
[Artist Name] - [Song Title]
Delivered: [Date]
Engineer: [Your Name]
FILES INCLUDED:
- MAIN: Final stereo mix, 24-bit/48kHz WAV. Send this to your mastering engineer
or upload directly to streaming if not mastering separately.
- INSTRUMENTAL: Full mix minus lead and backing vocals. Use for live shows,
karaoke, or sync placements.
- TV: Lead vocals only on top of instrumental. Used for live TV appearances
and some licensing contexts.
- ACAPELLA: Vocals only, no instruments. Useful for remixes and sync.
DISTRIBUTION VERSIONS:
- 16-bit/44.1kHz WAVs for legacy distribution or CD pressing.
REFERENCE MP3s:
- 320kbps MP3s of the main mix and instrumental. Use these to share around
privately. Do not use these for distribution or mastering.
MASTERING NOTES:
- Main mix has approximately -6dB of headroom on the master bus.
- No master bus limiter applied.
- LUFS-I measured at approximately -14 (pre-master).
NEXT STEPS:
- If sending to a mastering engineer, share the MAIN 24-bit WAV only.
- If self-distributing without mastering, use the 16-bit MAIN file.
- Any questions, reply to this email and I'll get back to you within 24 hours.
This project is now closed. Any future revisions will be billed at my
hourly recall rate of $X per hour.That last line is the close. It states that the project is complete. It quotes the recall rate so there’s no ambiguity later. It’s friendly, it’s clear, and it ends the engagement on a professional note rather than leaving the door open for indefinite tweaks.
How to Deliver the Files Without Looking Sloppy
The way you deliver the final files affects how the project feels to the client as much as the files themselves. Three patterns work, and one widely-used pattern actively damages your perceived value.
The pattern to avoid: WeTransfer. WeTransfer is fine for sending stems to a client at the start of a project. It’s sloppy as a final delivery method. The link expires in seven days. The interface is built for one-time transfers. The whole experience feels like emailing a contractor a zip of files, because that’s what it is. For a $1,000 mixing project, a $20,000 album, or a sync-bound single, WeTransfer is the equivalent of handing a luxury watch to a customer in a plastic grocery bag.
Pattern 1: Dropbox or Google Drive with a permanent link. Set up a folder specifically for that project. Share a link that doesn’t expire. The client can come back to it months later when they need the files again. Pin the link in the delivery email and reference it in your record-keeping so you can find it again if asked.
Pattern 2: A delivery portal built for music. Several tools exist specifically for sending finished audio files professionally. They look more polished than a Drive folder, often include playback so the client can listen in-browser before downloading, and signal that you take final mix delivery seriously as part of the craft. The downside is they’re another subscription. Worth it if you’re delivering more than five projects a month.
Pattern 3: A clean upload-and-download workflow tied to your intake. The same place clients sent their stems at the start of the project can be the place they download the final files. This is what session.trackbloom.com is built for. The session that started as an intake link comes full circle as a delivery point, which keeps the entire project organized in one location instead of scattered across Dropbox, WeTransfer, and email. The client experience is cohesive from upload to final delivery, which is what professional engagements are supposed to feel like.
Whichever pattern you choose, pick one and use it consistently. Switching delivery methods project to project signals that you don’t have a system, and the absence of a system is what makes engineers look like hobbyists even when their mixing work is professional.
The Delivery Email Template
The email that accompanies a final mix delivery should be short, structured, and unambiguous. Here’s the template that works for most projects:
Subject: Final delivery: [Artist] – [Song Title]
Hi [name],
Final mix delivery for [Song Title] is ready. You can download everything here: [LINK]
What’s in the folder:
- Main stereo mix (24-bit/48kHz WAV) for mastering or direct streaming upload
- Instrumental, TV, and acapella versions (same format)
- 16-bit/44.1kHz distribution copies
- 320kbps MP3 references for sharing
- A README file with a full breakdown
A few notes:
- The main mix has approximately 6dB of headroom on the master bus. No limiter applied. Ready for mastering.
- The link will stay active indefinitely, so you can come back to it whenever you need.
Please confirm in writing that this delivery is approved. Once confirmed, I’ll archive the session and consider the project complete. Any changes after sign-off would come in at my hourly recall rate, which is $X per hour with a one-hour minimum.
It’s been great working on this. Looking forward to hearing it out in the world.
[Your name]
That email does a lot of work in a small space. It tells the client what they’re getting and how to use it. It surfaces the mastering-readiness of the main file. It asks for explicit written sign-off. It quotes the recall rate before any post-delivery tweak conversation can begin. And it closes the relationship warmly without leaving the door propped open for endless follow-up.
The Stems Question: What to Send and How
Stems are a separate skill, and most engineers under-charge for them. A clean stem export is genuinely an extra hour or two of work on top of a mix, sometimes more. If you’re including stems in a final mix delivery, the rules that matter:
Bounce stems from the start of the session to the end, not just from the first downbeat. They need to be time-aligned so they play in sync when reimported. Include silent space at the head if there’s a count-in or pre-roll.
Decide on groupings before you bounce. The standard stems are drums, bass, lead vocal, backing vocals, keys, guitars, and FX. Pop and electronic mixes might add synths and percussion as their own groups. Hip-hop mixes might separate ad-libs from main vocals. Ask the client what they need the stems for: sync licensing usually wants traditional groupings, remixers usually want more granular separation, dance floor edits usually want a TV mix plus drums isolated.
Include effects on the stems unless the client specifically asks for dry stems. A drum stem with the room reverb and parallel compression printed on it sounds like the mix. A dry drum stem sounds like a multitrack. Mostly the former is what people want, but get confirmation in writing.
Name the stem files with the song name in front and the stem type at the end: Lyra - Easy Money - STEM - Drums.wav. This keeps everything sorting cleanly alongside the main mix files in the same folder.
What to Hold Back Until Payment Clears
If there’s an outstanding invoice, the final mix delivery is your last point of professional pressure. Most engineers hand over everything the moment they think the mix is done, then chase payment for the next 30 days. This is backwards.
The professional pattern: deliver reference MP3s before final payment. Hold the WAVs and stems until the invoice clears.
The delivery email in this case looks slightly different:
Hi [name],
Reference MP3s of the final mix are attached. These are watermarked and not intended for distribution, but they’re good enough for you to confirm sign-off.
Final WAV files, stems, and distribution copies will be delivered as soon as the outstanding invoice clears. Invoice attached for $X. Once payment is confirmed, I’ll send the full delivery folder within 24 hours.
Thanks, and looking forward to wrapping this up.
This is not aggressive. This is standard professional practice in every industry that delivers digital files. Wedding photographers don’t release high-res images before final payment. Designers don’t release source files before final payment. Mix engineers shouldn’t either.
If a client pushes back on this, the answer is in your contract. The “stems-and-files-on-payment” clause covered in the mix engineer contract post exists specifically for this moment. Reference it. Stand by it. The clients who object to paying before receiving high-quality files are exactly the clients who would have disappeared once they had the files.
Common Final Mix Delivery Mistakes
A short list of mistakes that show up repeatedly, ranked by how much they damage perceived professionalism:
Inconsistent sample rates. Mixing the song at 48kHz, then exporting the master at 44.1 without sample-rate-converting properly. Or worse, exporting at the original session rate without telling the client. Always confirm the target sample rate before bouncing the final mix delivery and write it into the filename.
Limiter on the main mix file. Unless the client explicitly asked you to skip mastering, do not put a limiter on the main mix bus. A mastering engineer needs the dynamic range. A client uploading directly to streaming usually still wants the headroom for their distributor’s processing.
Mono compatibility not checked. Some platforms still play in mono. Check the master in mono before bouncing and confirm nothing important disappears. Phasing issues that vanish in stereo will haunt you in mono.
Missing dither. If you’re going from 24-bit to 16-bit, dither the file. If you don’t know what dither is, learn before your next final mix delivery. It’s not optional.
Sending the wrong file as “final.” This is the most common mistake and the one that costs you the most. Triple-check the filename before you send the link. Open the file. Listen to the first 10 seconds. Confirm it’s actually the version you meant. The 30 seconds this takes saves you from the worst client conversation in mixing: “this isn’t the version we approved.”
What Great Final Mix Delivery Looks Like
The clients you want to keep don’t remember most of what happened during the mix. They remember the start (when they sent stems and felt nervous about whether it would go well) and the end (when they got the files and felt either professional or amateur).
Great final mix delivery feels like getting a finished product from a real business. Files named clearly. Folders organized logically. A README that explains what’s there. An email that walks them through what to do next. Explicit sign-off. A recall rate quoted in advance. A delivery link that doesn’t expire. The whole package signals that they hired a professional, which is what they wanted to believe when they paid the deposit.
Bad delivery feels like a contractor disappearing. One zip file in an email. No labels on the versions. No instructions. Crossed fingers that the client will figure out which file to upload. A WeTransfer link expiring in seven days. The client’s last interaction with you is confusion, and confusion is what they’ll remember when their friend asks who mixed their record.
The mixing work is the same in both cases. The difference is the last 30 minutes of the project. Spend them well.