Exporting & Sharing
When your mix is ready, export it to a standard audio file you can share, stream, or master further. Open the export dialog from ☰ ▸ Export….
Export is offline — AnalogDAW renders the mix faster (or slower) than real time through the full mixer, inserts, automation, and master chain, so the result is bit-for-bit what you hear, not a live capture. Playback and editing pause while it renders; you can Cancel at any time. Multicore rendering does not change the file contents: repeated exports of the same session produce the same audio.
What to export — the four modes
Pick a mode at the top of the export dialog:
| Mode | Renders… |
|---|---|
| Mix | The whole session as one stereo file, start to finish. |
| Loop Region | Just the loop region, played through once. Set the loop first (see The Interface). |
| Markers | The section between two markers you choose. Needs at least two markers. |
| Stems | Every track and mix bus as its own stereo file, in Tracks and Buses folders — for handing a mix to a mastering engineer or another DAW. |
File formats
| Format | Best for |
|---|---|
| WAV | Uncompressed, universal — the standard for delivery and archiving. |
| AIFF | Uncompressed, Apple's equivalent of WAV. |
| FLAC | Lossless but compressed (smaller files). Capped at 24-bit. |
| AAC | High-quality compressed (256 kbps) — small files for sharing/streaming previews. |
WAV/AIFF/FLAC export at your session's bit depth (Session Settings); FLAC is capped at 24-bit. AAC is the only lossy option here, and it is handled specially so it still honours your master ceiling — see Lossy exports and your ceiling.
Dither
AnalogDAW mixes internally in 32-bit floating point. Writing that to a 16-bit file has to round every sample to a coarser grid, and rounding on its own produces distortion that follows the music instead of a steady noise floor — you hear it as a gritty edge on reverb tails, fade-outs, and quiet passages, exactly where there is least signal to hide it.
To prevent that, AnalogDAW automatically adds TPDF dither whenever it writes a 16-bit file — exports, bounces, and flattened take folders alike. It is a tiny amount of noise (about one least-significant bit), added before rounding, and it trades that signal-following distortion for a constant, far less audible hiss. There is nothing to switch on.
- 24-bit and 32-bit exports are not dithered, deliberately. At 24 bits the rounding error already sits far below the noise floor of any microphone, converter, or room your material has passed through, so dither would only add noise to mask something inaudible. 32-bit float is not rounded at all.
- If you are sending a mix to a mastering engineer, export 24-bit or 32-bit. Dither belongs at the very last step, on the final delivery file. Exporting 16-bit stems that will be processed further means the dither gets processed too, and the true final render dithers again on top of it.
- Choose 16-bit for final delivery — CD, or a file going straight out as-is.
Loudness normalization
Turn on Normalize and pick a target to render your export at a consistent loudness (LUFS). The targets are the same delivery targets the Mastering Assistant offers: Streaming Balanced (-14 LUFS), Streaming Conservative (-16 LUFS), Streaming Loud (-12 LUFS), Video Streaming (-13 LUFS), and CD / Club (-9 LUFS).
With Normalize on, the file you export is the normalized one — it is written at the target loudness, under the name and in the place you chose. There is no second copy to tidy up. This is ideal for delivering previews that play back at the "right" level on streaming services. Leave it off when delivering to a mastering engineer, who will set final loudness themselves.
AnalogDAW finishes exports as a transaction. If a mix or stem batch is cancelled or fails while it is being rendered, normalized, or copied into place, existing files are restored and an incomplete batch is not left behind. If the operating system also prevents a previous file from being restored, AnalogDAW keeps its recovery copy and shows the folder path in the error instead of deleting it.
Normalize only applies gain. There is no limiting or compression involved, and the gain is capped so the file never exceeds your master Maximizer's ceiling (or -1 dBTP if the Maximizer is switched off). If your master fader or trim is above unity, it counts toward that ceiling — the fader sits after the Maximizer, so it is part of what reaches the file. Turning a mix down to -14 or -16 LUFS therefore always works, but a louder target can fall short: a mix that already peaks near full scale has no headroom left to gain up, so asking for -12 or -9 LUFS may land well below it. To genuinely reach a loud target, master the mix to it — the Mastering Assistant has a maximizer, this does not.
Lossy exports and your ceiling
AAC is a lossy format. It does not store your waveform — it stores a model of it, and the decoder rebuilds something close but not identical. One consequence surprises people: a lossy file's peaks are not the peaks you gave it. Encoding a loud master routinely pushes the decoded file 1-2 dB above the ceiling it was limited to, which is why delivery specs ask for headroom on lossy formats. No limiter can prevent this, because the overshoot happens after the limiter, inside the encoder.
So AnalogDAW checks. An AAC export is rendered to full-quality audio first, then encoded, then decoded again and measured. If the delivered file breaches your Maximizer's ceiling — or -1 dBTP, if the Maximizer is switched off — it is re-encoded from a slightly quieter render until it doesn't. You get the ceiling you asked for in the file you actually ship. On the rare material where a handful of passes still can't reach it, the export completes and tells you the level it did land on, so the decision is yours rather than silent.
Three things follow from that:
- An AAC export takes longer than a WAV of the same mix — it is encoded and measured more than once.
- Your AAC will be a little quieter than your WAV. That difference is the headroom needed to land the delivered file on your ceiling; how much depends on the material, and a mix with headroom to spare needs none at all. It is not a loss of quality, and streaming services normalize loudness anyway.
- A range that cuts off mid-note gets a little extra trim. True peak is measured the way delivery specs define it, which includes the step where the file stops — a hard cut at a loud moment reconstructs above its own samples, and nothing upstream of the cut can prevent that. It costs well under a decibel, and ending your export range in silence (or on a fade) avoids it entirely.
- With Maximizer look-ahead off, expect a larger difference. At zero latency the Maximizer holds the samples to your ceiling but not the waveform between them, so a render can already sit 1-2 dB over the ceiling in true-peak terms before the encoder adds anything. The AAC export gives that back, and the WAV does not. Turn look-ahead on when you are exporting a master and the two land much closer together.
WAV, AIFF and FLAC are lossless — they reproduce the render exactly, so they land on your ceiling precisely. At 16- or 24-bit all three carry identical audio (the files differ in container and packing, the samples do not); at a 32-bit float session FLAC still writes 24-bit, so it is the one of the three that is not a bit-for-bit copy of the render. If you want to verify a master against its ceiling, measure a lossless export.
File name and location
- For Mix, Loop Region, and Markers, give the export a file name (a sensible default is suggested). On macOS you choose where to save; on iOS, exports are offered through the Share sheet so you can save to Files or send them on.
- Stems are written as a set of files into named folders.
Every mode offers the session's own name, undecorated — no -mix or
-section suffix. On iPhone and iPad, where there is no save panel to warn you,
a bounce never replaces an earlier one: export a mix and then a marker section,
and the second one gets the same name with a counter — a session called Session
gives you Session 2. A stem export does the same with its folder. Rename them in
Files if you want something else. On a Mac the save panel already asks before
replacing anything, so the name you type is the name you get.
When an export finishes, you'll see an "EXPORT SAVED" confirmation in the transport bar.
Exporting and freezing a track use the same offline renderer, so they cannot run at once. If a track is still freezing — or still unfreezing — AnalogDAW says so and asks you to try the export again once it finishes. (In the other direction, a freeze started during an export is refused the same way.) Frozen tracks export exactly as you hear them — a mix or stem bounce uses each frozen track's printed audio, so freezing never changes your export.
The diagnostic bundle (support)
Unrelated to mixing, but in the same menu: ☰ ▸ Export Diagnostic Bundle…
(also ⌘⌥D on macOS) gathers technical details about your session and system
into a small bundle you can send if you ever need to report a problem. It
omits all audio, full file paths, and plug-in state — but it does include
file names, track names, plug-in names, and session metadata to help diagnose
the issue.
Next: Keyboard Shortcuts →