Audio

Audio Formats for Music Production and Distribution

Published Mar 19, 2026 9 min read By ChangeThisFile Team

Quick Answer

Record and mix in WAV or AIFF at 48 kHz / 24-bit. Deliver stems as WAV. Master to WAV for the CD master, FLAC for hi-res distribution, and let distributors handle lossy encoding for streaming. Keep session files and bounces in lossless formats throughout the production chain — lossy encoding happens once, at the very end.

Music production involves dozens of format decisions — from the initial recording through mixing, mastering, and distribution to every platform. Get them wrong and you either waste storage on unnecessarily large files or introduce quality degradation that compounds at every stage.

The governing principle is simple: stay lossless until the final delivery step. Every recording, every bounce, every stem, every mix revision should be WAV or AIFF. Lossy compression (MP3, AAC, OGG) happens once — at the very end, when the mastering engineer or distributor converts your lossless master for streaming and download platforms. One generation of lossy encoding is inaudible at modern bitrates. Two or more is not.

This guide covers the complete format pipeline from tracking through distribution, with specific recommendations for each stage.

Session Formats: Recording and Tracking

Every track recorded in your DAW should be WAV (or AIFF on Apple platforms) at 48 kHz / 24-bit minimum. This is non-negotiable for professional work.

Parameter	Recommended	Acceptable	Avoid
Format	WAV	AIFF	MP3, AAC, OGG
Sample Rate	48 kHz	44.1 kHz, 96 kHz	Below 44.1 kHz
Bit Depth	24-bit	32-bit float	16-bit

Why 48 kHz over 44.1 kHz: If any track might be used in video content (music videos, sync licensing, YouTube), 48 kHz avoids sample rate conversion at the video editing stage. If you're producing exclusively for CD release, 44.1 kHz is fine.

Why 24-bit: Recording headroom. At 24-bit, you can record at -18 dBFS average (the widely recommended "sweet spot" that emulates analog levels) and still have 126 dB of clean dynamic range. At 16-bit, the same recording level leaves you with only 78 dB before the noise floor.

Why not 96 kHz: At 96 kHz, file sizes double, disk I/O doubles, and plugin CPU usage increases (some plugins oversample, others process twice as many samples per second). The audio quality benefit for final playback is nil. Use 96 kHz only if your workflow involves heavy pitch-shifting or time-stretching where aliasing might be an issue.

Stem Delivery and Collaboration

When sending individual tracks or stems to a mixing engineer, collaborator, or remixer, the format matters:

Format: WAV, always. Not FLAC (some DAWs don't import FLAC directly), not MP3 (quality loss), not your DAW's native format (Logic/Ableton/Pro Tools session files aren't interchangeable).
Sample rate and bit depth: Match your session settings. If you recorded at 48/24, export stems at 48/24. Never upsample (44.1 → 48) or increase bit depth (16 → 24) for stems — this adds data without adding quality.
Start point: All stems should start at the same point (bar 1, beat 1, or session start). This ensures the mixing engineer can drop all files into a new session and they align perfectly. Export from the start of the session, not from where audio first appears.
Naming convention: Clear, consistent names. "01_Kick.wav", "02_Snare.wav", "03_Bass_DI.wav", "04_GTR_Left.wav". Numbered prefix ensures correct ordering in file browsers.

For transferring large stem packages (a 40-track session can easily be 5-10 GB as WAV), compress as ZIP or use cloud storage. ZIP doesn't alter the audio data — it's fully lossless file-level compression.

Mixing: Working Format and Bounces

During mixing, your DAW processes audio internally in 32-bit or 64-bit float. The source files are 24-bit WAV, but every plugin, fader movement, and bus summing happens at float precision. This provides unlimited headroom — intermediate stages can exceed 0 dBFS without clipping.

When you export (bounce) a mix:

Mix revisions: WAV, 48 kHz, 24-bit (or 32-bit float). Keep every revision. Storage is cheap. "Mix_v3_vocal_up_1dB.wav" is worth the 50 MB when the client asks you to go back to "the version from last Tuesday."
Print stems (grouped submixes for mastering): WAV, matching session settings. Common stems: drums, bass, guitars, keys, vocals, effects. The mastering engineer may want these for parallel processing or if the mix needs targeted adjustments.
Pre-master mix: WAV, 48/24 or 48/32f. No limiting, no final processing. Peaks at -3 to -6 dBFS to give the mastering engineer headroom. This is the file you send to mastering.

Never bounce to MP3 or AAC during production. Not for reference mixes, not for client previews, not for "quick checks." Lossy bouncing introduces artifacts that may mask or mimic mixing issues. If you need to send a quick reference, bounce to WAV and let the recipient's playback handle any necessary transcoding.

The Mastering Chain: Format at Every Stage

Mastering is where format decisions have the most downstream impact, because the mastering output is what reaches listeners.

Mastering Input

Accept: WAV or AIFF, 44.1-96 kHz, 24-bit or 32-bit float. Reject: anything lossy. The mastering engineer needs the best possible source material. If a client sends an MP3, the mastering session starts with compromised audio that can't be recovered.

Processing

All processing happens at 32-bit or 64-bit float internally. EQ, compression, limiting, stereo enhancement, and loudness optimization — all at full precision. The mastering engineer's session doesn't touch the file format until the final export.

Output Formats

The mastering session produces multiple deliverables:

CD master: WAV, 44.1 kHz, 16-bit (with dither — TPDF or noise-shaped). This is the Red Book standard. ISRC codes embedded if applicable.
Hi-res master: WAV or FLAC, 48-96 kHz, 24-bit. For hi-res stores (Qobuz, HDtracks) and hi-res streaming (Apple Music Lossless, Tidal HiFi Plus). No dither needed when staying at 24-bit.
Digital distribution master: WAV, 44.1 kHz, 16-bit or 24-bit. This goes to your distributor (DistroKid, TuneCore, CD Baby), who handles encoding to MP3, AAC, Vorbis for each streaming platform.
Archive master: WAV, session sample rate, 24-bit (or 32-bit float). The highest-quality version for your archives. Encode to any target from this.

Distribution: What Each Platform Receives

When you upload through a distributor, they handle the encoding:

Platform	You Upload	They Encode To
Spotify	WAV or FLAC, 44.1+ kHz, 16/24-bit	OGG Vorbis 96-320 kbps
Apple Music	WAV or FLAC, up to 192 kHz/24-bit	AAC 256 kbps + ALAC lossless
Amazon Music	WAV or FLAC	Multiple tiers including FLAC lossless
YouTube Music	WAV or FLAC (via video upload)	AAC 128-256 kbps
Tidal	WAV or FLAC, up to 96 kHz/24-bit	FLAC (HiFi), AAC (standard)
Bandcamp	WAV or FLAC, 16-24 bit	MP3 V0, FLAC, AAC, OGG, WAV (buyer chooses)

The takeaway: always upload the highest quality lossless master you have. Let the platforms handle encoding to their target formats. Never upload MP3 to a distributor — they'll re-encode it, compounding artifacts.

Sample Libraries and Plugin Compatibility

If you work with sample libraries (Kontakt, Splice, Loopmasters), format compatibility matters:

WAV: Universal. Every sampler, every DAW, every plugin reads WAV. Deliver samples as WAV.
AIFF: Common in Apple-centric workflows. Most samplers read AIFF natively. Logic and Apple Loops use AIFF.
FLAC: Growing support but not universal. Kontakt doesn't natively read FLAC as of recent versions. Convert to WAV before loading into samplers.
MP3/OGG: Some game engines (Unity, Unreal) import these directly for compressed audio assets. For music production, never use lossy samples — artifacts in source material become amplified through processing.

When creating your own sample libraries: 48 kHz / 24-bit WAV is the standard. Include tempo and key information in filenames or metadata. Trim start and end points cleanly. Normalize to -1 dBFS peak if providing one-shots; leave dynamic range intact for loops.

Session Backup: Formats and Strategy

Your production files are the raw material from which everything else derives. Losing a session with 40 tracked instruments is losing months of work. Backup strategy by format:

Session files (.als, .ptx, .logicx, .rpp): Platform-specific, not interchangeable. Back up to at least two locations.
Audio files (WAV/AIFF): These are the actual recordings. Back up separately from session files — they're the irreplaceable part.
Bounce archive: Every mix revision, every mastered version. WAV for active projects, FLAC for long-term archival (50% smaller with zero quality loss).
Plugin presets and recall sheets: Document your processing chain. If a plugin update breaks compatibility, you'll need to recreate settings.

Storage math: A typical album session (40 tracks × 4 minutes × 48 kHz × 24-bit) is roughly 10-15 GB. With mix revisions and bounces: 20-30 GB. A 4 TB external drive holds 130-200 album projects. Invest in two drives — one working, one backup.

Converting Production Audio Files

Common production conversions: preparing stems from mixed formats, converting for distribution, creating reference copies.

The format pipeline for music production is WAV in, WAV through, and WAV/FLAC out — with lossy encoding happening exactly once at the final distribution step. Every deviation from this principle risks introducing artifacts that compound through the production chain. Storage is cheap; re-recording a session because you discovered artifacts in your only copy is not.

Need to prepare audio for distribution or archival? Convert WAV to FLAC, FLAC to WAV, or AIFF to WAV — free at ChangeThisFile.

Key Takeaways

Record at 48 kHz / 24-bit WAV. Stay lossless through mixing and mastering. Encode lossy once at the end.
Deliver stems as WAV at session settings, all starting from the same point, with clear naming.
Mastering produces multiple deliverables: 44.1/16 for CD, 48-96/24 for hi-res, and a lossless master for your distributor.
Upload the highest quality lossless file to your distributor. They handle encoding for each streaming platform.
Never bounce to MP3 during production — not for mix revisions, not for client previews, not for anything.
Back up session files and audio separately. FLAC is ideal for long-term archival — 50% smaller than WAV, zero quality loss.

Frequently Asked Questions

Should I record at 44.1 kHz or 48 kHz for music production?

48 kHz is the safer default. It matches the video/broadcast standard, so if any of your music ends up in a video (music video, sync license, YouTube content), no sample rate conversion is needed. If you're exclusively producing for CD release and will never touch video, 44.1 kHz is fine. The audio quality difference between 44.1 and 48 kHz is inaudible.

What format should I send stems to a mixing engineer?

WAV at your session's sample rate and bit depth. All stems starting from the same point (session start). Clear naming convention with numbered prefixes. Never send MP3 stems, FLAC stems (compatibility issues with some DAWs), or DAW session files (they're not cross-platform). ZIP the stems for transfer if needed — ZIP compression is lossless.

Do I need 96 kHz for professional music production?

No. 48 kHz / 24-bit is the professional standard. 96 kHz doubles file sizes, increases CPU load for plugins, and provides no audible quality improvement for the final output. The only justified use of 96 kHz is when doing heavy DSP processing (extreme pitch-shifting, time-stretching) where aliasing artifacts at 48 kHz might fold into the audible range.

Should I upload WAV or FLAC to my music distributor?

Either works — both are lossless. WAV is accepted by every distributor without exception. FLAC is accepted by most (DistroKid, TuneCore, CD Baby all accept FLAC). FLAC is half the size of WAV, making uploads faster. Check your specific distributor's accepted formats, but in 2026, FLAC support is nearly universal.

What is dithering and do I need it for my master?

Dithering adds controlled random noise when reducing bit depth (24-bit to 16-bit). Without it, truncating bits creates audible distortion on quiet passages. You need dithering when creating a 16-bit CD master from a 24-bit session. You don't need it when staying at 24-bit (for hi-res distribution) or when uploading to streaming services that accept 24-bit files. Dither once, at the final bit-depth reduction step.

Can I use FLAC in my DAW session instead of WAV?

Some DAWs support FLAC natively (Reaper, Audacity). Others don't (Pro Tools, Logic as of recent versions). Even in DAWs that support FLAC, using it adds decode overhead for every track on every playback cycle. WAV is the universal standard for production — zero decode overhead, guaranteed compatibility. Use FLAC for archival and distribution, WAV for active production.

What happens if I upload MP3 to a music distributor?

The distributor will re-encode your already-lossy MP3 into AAC (for Apple), Vorbis (for Spotify), and other formats. Each lossy re-encoding compounds artifacts — frequencies are discarded again, compression noise accumulates. The result sounds measurably and sometimes audibly worse than encoding from a lossless source. Always upload WAV or FLAC.

How should I back up my music production files?

Follow the 3-2-1 rule: 3 copies, 2 different media types, 1 offsite. Session files plus raw audio WAVs on your working drive, mirrored to an external drive, and a third copy on cloud storage or a drive at a different physical location. For long-term archival, convert audio to FLAC (50% smaller) but keep session files in their native format. A typical album project is 20-30 GB — manageable on modern drives.

Ready to convert your files?

Use ChangeThisFile to convert between 600+ formats — free, fast, and private.

Start Converting