Best Export Settings for Mixing & Mastering (24-bit, Headroom) | MegaMix AI
Export from the native session rate
Avoid unnecessary sample-rate conversion between stages. Convert only when a delivery target requires it. If you produced at 48 kHz, bounce 48 kHz until a distributor or video project needs 44.1 or 48 specifically. Consistent rates reduce surprises and processing artifacts.
Upsampling a 44.1 kHz session to 96 kHz for “quality” before mixing does not restore high-frequency information you never recorded. It makes bigger files and another conversion later.
Video collabs often want 48 kHz. Music stores accept 44.1 or 48. Match the project, then convert once at the end if required.
Bit depth and headroom guidance
Use 24-bit for mix and stem exports in most workflows. It offers practical dynamic precision without 32-bit float file bloat for sharing. Leave headroom before mastering so dynamics processing has room to work cleanly, typically a few dB of peak room, not a normalized 0 dBFS mix print.
16-bit is for specific deliveries like CD after dither. Do not mix in 16-bit because an old forum said so.
32-bit float bounces can be useful internally. Many collaborators still want 24-bit WAV. Ask, then export that.
File format best practices
Use WAV for production and mastering stages; use lossy formats for preview sharing only. Name files clearly with version labels and date to avoid confusion. AIFF is fine if your Mac world prefers it; do not send a .m4a as a master.
Do not master from an MP3, then bounce a WAV of that. You are mastering codec smear. Go back to the mix WAV.
Zip stems with a readme of sample rate and start bar. Future you is a collaborator too.
Final QC before delivery
Check fades, starts, and silence tails. Small export mistakes can break a release timeline: a 200 ms of extra click, a missing last reverb tail, a bounce that started at bar 2. Audit metadata and loudness consistency before sending to distribution.
Play the file in a different app than your DAW. DAWs can mask export bugs with playback settings, oversampling differences, or plugin delay compensation that did not print.
Confirm the file is actually the version you approved, not mix_v3 sitting in the same folder as master_v1. Open the WAV, play 0:00, and match the date in the name to your notes.
Export workflow that prevents redo week
Same locator every time. Same bit depth. Same naming pattern. Boring is the point.
Offline bounces can differ from realtime if plugins misbehave. If something sounds extra crunchy, try realtime once as a check.
- Set export range from true start through last tail.
- WAV, 24-bit, session sample rate, no extra limiter on the bounce unless it is a preview.
- Bounce mix and stems with identical starts.
- Listen in Finder/Explorer or a phone after transfer.
- Dither only when reducing bit depth for a specific delivery.
- Archive mix print and master with dates in the names.
Export settings checklist
If you skip the listen-outside-the-DAW step, you will learn this lesson on release day. Another app and a phone transfer catch export bugs the session view hides.
MP3 is for texts, not for DistroKid. Upload the 24-bit WAV you already QC’d, then encode previews from that master.
- WAV (or AIFF), not MP3, as the master source.
- 24-bit mix/stem prints; dithered 16-bit only when required.
- Native sample rate until a real target needs conversion.
- Headroom on mix prints; no stem normalization to 0 dBFS.
- Identical start times on stems.
- Filename includes version and date.
- Playback check outside the DAW.
A 320 kbps MP3 can fool you in a car. It will not fool a mastering stage or a second encode to AAC.
Export for MegaMix AI and then for stores
Bounce 24-bit WAV stems or a rough mix at session rate with headroom, upload at /app, then download the mixed or mastered WAV MegaMix AI gives you. Use that WAV for the distributor, not a voice-memo rerecord of the speakers. No DAW is required after the bounce if you are done editing.
The 7-day free trial is a good time to standardize this path on one song. If the AI export sounds different from in-app monitoring, you still QC the downloaded file on a phone.
Keep SFX as WAV too. MegaSFX AI at /megasfx downloads belong in the session bounce, not as a last-minute MP3 layer on the master. For demos, workflow, and more mix AI FAQs, see the Mix AI hub.
FAQ
Should I dither during mix export?
Usually reserve dithering for final bit-depth reduction at delivery, such as 24-bit to 16-bit CD. Bouncing 24-bit to 24-bit does not need dither; Dithering every stem “just in case” is superstition.
Is 48 kHz always better than 44.1 kHz?
Not always. Consistency with your production pipeline matters more; 48 kHz is common for video; 44.1 kHz is classic for CD-centric music; Neither magically improves a closet vocal.
Can I master from MP3?
Not recommended. Use lossless mix exports for mastering quality; Mastering an MP3 prints codec artifacts into the “high quality” WAV you then upload.
Why do my exports sound different?
Check playback normalization, plugin oversampling, offline versus realtime render, and whether a limiter was on the export bus. Also check you are not comparing a loud master to a quiet mix print.
Should I export at 32-bit float?
It can be useful inside a DAW to avoid internal clips. For sharing and most AI uploads, 24-bit WAV is the practical standard; Ask the receiving tool if it prefers float; do not assume bigger is better.
Do I need interleaved stereo WAV?
Yes for almost every distributor and collaborator. Split-mono stems are a special studio request; Default to standard stereo interleaved WAV unless someone specifies otherwise. More on the Mix AI hub. See Start free trial or plans.