Best Audio Format for Audacity: Import, Edit, and Export

Audacity is the most-installed free audio editor on the planet — over 100 million downloads since 1999, and the active 3.x line is more capable than most people realize. Format handling is the foundation of any Audacity workflow: pick the right import format and you save CPU and avoid quality loss; pick the right export format and your file works on the destination platform without surprises. This guide walks through Audacity's full format pipeline, from project files to delivery.

Audacity's Internal Format: 32-bit Float PCM

Whatever you import into Audacity — MP3, M4A, FLAC, OGG, WMA, WAV, AIFF, Opus — gets decoded to 32-bit floating-point PCM at the project's sample rate the moment it lands on the timeline. This is by design. 32-bit float gives Audacity essentially unlimited headroom (peaks up to roughly +770 dB without clipping) and bit-perfect arithmetic during edits. Every effect, fade, normalization, and gain change happens in 32-bit float.

The practical implication: format quality differences between source files do not affect editing operations. Editing a 128 kbps MP3 in Audacity is no slower or noisier than editing a WAV, because both have already been decoded to identical-precision PCM. The differences appear at export time when you re-encode.

The Project File: AUP3

Audacity 3.0 (released 2021) introduced AUP3, a single SQLite database file that contains the project's audio data, edits, labels, tracks, and settings. Earlier versions used .aup XML files plus an _data folder full of fragmented audio chunks — a notorious source of "missing file" errors when users moved or renamed projects. AUP3 fixes that by bundling everything.

• Saved internally as 32-bit float PCM at the project sample rate
• Single file, easy to back up or move
• Lossless round-trip: close and reopen with no quality change
• File size: roughly the same as uncompressed WAV of the same audio length, plus a small SQLite overhead

For long-term archival, export the project to a separate WAV or FLAC file. AUP3 is intended as the live working format, not a delivery format.

Best Format for Importing

Audacity's broad import support comes through two paths: native decoders for WAV, AIFF, FLAC, OGG Vorbis, and MP3 (via the bundled LAME library); and FFmpeg-based decoders for M4A, AAC, OGG Opus, WMA, AC-3, AMR, and most video container audio tracks. The FFmpeg integration is now bundled with Audacity 3.4 and later — older versions required a separate download from the Audacity site.

Recommendations:

• WAV or AIFF for editing source. No decode overhead, no compression artifacts, sample-accurate seeking.
• FLAC for lossless source you do not want to inflate to WAV size.
• MP3, M4A, OGG are fine to import for editing, but if you plan to export back to a lossy format, decode them to WAV first to avoid one extra generation of lossy damage.

If a file fails to import — usually a sign that FFmpeg is missing or out of date — convert it to WAV using a browser-based tool like M4A to WAV or OGG to WAV before importing.

Bit Depth Handling

Audacity's internal 32-bit float means bit depth conversions happen at export time, not during editing. The bit-depth dropdown in File > Export controls the output:

• 16-bit: CD-quality, standard for distribution (MP3 export is implicitly limited to 16-bit per encoder spec)
• 24-bit: standard for tracking, mixing, and mastering deliverables
• 32-bit float: for project-to-project handoff to other DAWs that support it (Pro Tools, Reaper, Logic)

Dithering options appear when downsampling from 32-bit float to 16-bit. Triangular dither is the safe default; shaped dither (TPDF noise-shaped) sounds slightly cleaner on quiet content.

Export Formats by Use Case

MP3 Export: LAME Bundled

Audacity stopped requiring a separate LAME download in version 2.3.2 (2019). LAME is now bundled and MP3 export works immediately on a fresh install. The export dialog exposes:

• Bitrate mode: Constant, Variable, Average
• Quality preset: for VBR, q0 (best, ~245 kbps avg) through q9 (worst)
• CBR bitrate: 8, 16, 24, 32, 40, 48, 56, 64, 80, 96, 112, 128, 160, 192, 224, 256, 320 kbps
• Channel mode: Joint stereo (default), Stereo, Mono
• Force export to mono: required for ACX audiobook compliance

FLAC, OGG Vorbis, and Opus Export

All three are exported via Audacity's bundled libraries. FLAC defaults to compression level 5 (the official sweet spot); level 8 gives 1–3% better compression for ~5x the encoding time. OGG Vorbis quality 5 (~160 kbps) is the recommended setting for music. Opus uses VBR by default with bitrate selectable from 6 to 510 kbps — 96 kbps is transparent for music, 32 kbps for speech.

Recommended Workflow

1. Record or import source as WAV (or FLAC for lossless source)
2. Save the AUP3 project to keep edits non-destructive
3. Apply effects, edits, normalization
4. Export the master to WAV or FLAC for archive
5. From that master, export delivery formats (MP3 for distribution, M4A for Apple, etc.)

Keeping a clean WAV/FLAC master means future deliveries to new platforms encode from a high-quality source rather than compounding lossy generations.

Sample Rate Conversion in Audacity

Audacity 3.x uses libsoxr internally for high-quality sample rate conversion. The default Best Quality preset is essentially transparent for typical audio. Access via Tracks > Resample, or change the project rate at the bottom-left of the window for new tracks.

For collaborative workflows where stems arrive at mixed sample rates: conform on import (Audacity prompts when a clip's rate differs from the project rate), do all editing at the project rate, then export at the destination's required rate. Repeated SRC operations accumulate small artifacts; one well-placed conversion is cleaner than three.

For the broader picture on lossless masters, see PCM audio explained, what is PCM, and fix audio clicks and pops for cleanup workflows in Audacity.