Every audio file format exists to answer one question: how much quality are you willing to trade for file size? During recording, the answer is none — you capture everything the microphone hears. During mastering, you export to the destination's exact specifications. During distribution, you let the platform compress to whatever codec they use internally.
This guide covers every format in the music production chain, from DAW session files through recording, mixing, mastering, and distribution. With actual sample rates, bit depths, and bitrate targets for each stage.
DAW Session Formats
DAW (Digital Audio Workstation) session files store the entire project state: track layout, plugins, automation, routing, and references to audio files on disk. The session file itself is typically small; the audio files it references are where the space goes.
Major DAW File Formats
Ableton Live (.als): Gzipped XML that references audio files in the project folder. Ableton can "Collect All and Save" to bundle everything into one folder. The .als file itself is typically 1–50MB; audio files add up to GB. Ableton also uses .alc for clips/presets and .adg for device groups.
Logic Pro (.logicx): macOS bundle (folder). Contains project data, audio files (if managed), movie files, and plugin settings. Logic can embed all media or reference external files. A typical session with 30 tracks of recorded audio: 2–8GB as a managed bundle.
Pro Tools (.ptx): Industry standard for professional studios. The .ptx file is the session; audio lives in an "Audio Files" folder alongside it. Pro Tools is extremely particular about file locations — moving audio files breaks the session. Use "Save Copy In..." to create portable session bundles.
FL Studio (.flp): Binary project file. FL Studio can store audio samples directly in the .flp file or reference external files. This makes .flp files self-contained but potentially very large (500MB+ with many audio samples embedded).
Reaper (.rpp): Plain-text project file (human-readable). Audio files are referenced from the project folder. The .rpp file itself is tiny (100KB–1MB) regardless of project complexity. Reaper's format is the most inspectable of any DAW.
Cross-DAW Exchange
Stems (WAV): The universal exchange format. Export each track as a separate WAV file (same length, same start point). Any DAW can import WAV stems. You lose plugin settings and automation, but the audio is preserved perfectly.
AAF/OMF: Advanced Authoring Format carries edit decisions, fades, and automation between Pro Tools, Resolve, Premiere, and some others. Useful for video post-production handoff. Not supported by Ableton, Logic, or FL Studio.
MIDI: Standard MIDI files (.mid) transfer note data between any DAW. No audio, no plugin settings, no mixing data — just notes, velocities, timing, and basic controller data. Use for sharing compositions when the recipient will use their own sounds.
Recording Formats: Capturing Audio
Recording demands the highest quality format because everything downstream depends on this source material.
WAV: The Recording Standard
Spec: WAV (Waveform Audio) stores uncompressed PCM (Pulse Code Modulation) audio. The recording standard is 24-bit/48kHz, which gives 144dB dynamic range and captures frequencies up to 24kHz (well above the 20kHz human hearing limit).
Why 24-bit, not 16-bit: 16-bit gives 96dB dynamic range. In a quiet recording room (30dB noise floor), that leaves 66dB between the noise floor and clipping. 24-bit's 144dB range gives 114dB of usable headroom. This means you can record at conservative levels (-18dBFS) without losing resolution in quiet passages. The difference is real and audible in recordings with wide dynamic range (classical music, acoustic performances).
Why 48kHz, not 44.1kHz: 44.1kHz is the CD standard. 48kHz is the video/film standard. If your music will ever sync to video (which it might — licensing, music videos, sync placements), 48kHz avoids sample rate conversion later. For music-only production, 44.1kHz is fine, but 48kHz costs nothing extra and provides a small anti-aliasing benefit.
Higher rates (96kHz, 192kHz): Diminishing returns for most music production. The audible benefit above 48kHz is debatable. The costs are real: double the file size (96kHz), 4x the file size (192kHz), higher CPU load for plugins, and compatibility issues with some hardware. Use 96kHz if you plan extreme time-stretching or pitch-shifting. Skip 192kHz.
File sizes: 24-bit/48kHz stereo WAV = 17.3MB per minute. A 40-track, 4-minute recording session: ~2.8GB. A full album session (12 songs, 40 tracks, including takes): 25–50GB.
AIFF: Apple's Uncompressed Format
AIFF (Audio Interchange File Format) is technically identical to WAV in audio quality. Same PCM data, different container. WAV uses RIFF headers; AIFF uses IFF headers. The decoded audio is bit-for-bit identical.
When you'll see AIFF: Logic Pro defaults to AIFF for recording. Some sample libraries ship as AIFF. Historically Mac-centric workflows used AIFF while Windows used WAV. In 2026, this distinction is irrelevant — every DAW on every platform reads both. Use whichever your DAW defaults to.
Convert AIFF to WAV | Convert AIFF to MP3 | Convert AIFF to FLAC
Mastering Formats: Final Output
Mastering output formats depend on where the music is going. Each distribution channel has specific requirements.
Mastering Output Specs by Destination
| Destination | Format | Bit Depth | Sample Rate | Loudness Target |
|---|---|---|---|---|
| CD (Red Book) | WAV | 16-bit | 44.1kHz | No standard (typically -9 to -14 LUFS) |
| Spotify/Apple Music | WAV or FLAC | 16-bit or 24-bit | 44.1kHz | -14 LUFS (Spotify), -16 LUFS (Apple) |
| Tidal HiFi | FLAC | 24-bit | 44.1–96kHz | -14 LUFS |
| Vinyl cutting | WAV | 24-bit | 96kHz | Per cutting engineer |
| Film/TV sync | WAV | 24-bit | 48kHz | Per spec (-24 LKFS broadcast) |
| Digital distribution (DistroKid, etc.) | WAV | 16-bit or 24-bit | 44.1kHz | -14 LUFS recommended |
Dithering: When converting from 24-bit to 16-bit (for CD or 16-bit distribution), apply dither. Dithering adds microscopic noise that prevents the quantization distortion (harsh truncation artifacts in quiet passages) that occurs when reducing bit depth. Use MBIT+ (iZotope), POW-r 3, or triangular dither. Apply dither exactly once, as the absolute last step in the mastering chain.
Stem Delivery for Mastering
If you're sending tracks to a mastering engineer, they may ask for stems: grouped submixes that sum to the full mix.
Typical stem groups: Drums/percussion, bass, guitars/keys, vocals, effects/ambient. Each stem is a WAV file at the session's native sample rate and bit depth (24-bit/48kHz or higher). All stems must start at the same timecode and be the same length.
Headroom: Leave 3–6dB of headroom on the mix bus (peak around -6dBFS to -3dBFS). Don't clip. Don't master before sending to a mastering engineer — remove any limiter or maximizer on the mix bus. They need dynamic range to work with.
Distribution Formats: Getting Music to Listeners
Once mastered, the audio enters distribution channels that each apply their own processing.
FLAC: Lossless Distribution
FLAC (Free Lossless Audio Codec) compresses audio without losing any data — decoded FLAC is bit-for-bit identical to the original WAV. Compression ratio is typically 50–70% (a 50MB WAV becomes 15–25MB FLAC).
Where FLAC is used: Bandcamp (primary format for purchased downloads), Tidal HiFi, Amazon Music HD, Qobuz, archival downloads. FLAC supports up to 32-bit/655kHz, covers art embedding, and has full metadata support (Vorbis comments).
ALAC: Apple's lossless codec. Functionally identical to FLAC but in an M4A/MP4 container. Apple Music uses ALAC for lossless streaming. If you're distributing exclusively through Apple's ecosystem, ALAC is the native choice. For everything else, FLAC is more universal.
Convert WAV to FLAC | Convert FLAC to WAV | Convert FLAC to MP3
MP3, AAC, and OGG: Lossy Streaming
MP3 (MPEG-1 Audio Layer 3): The format that started digital music. At 320kbps CBR, transparent quality for most listeners on most playback systems. At 128kbps, audible artifacts in cymbals, reverb tails, and stereo imaging. MP3's psychoacoustic model is 30 years old — newer codecs (AAC, Opus) sound better at the same bitrate.
AAC (Advanced Audio Coding): Apple's preferred format, used by iTunes/Apple Music, YouTube (audio track), and many streaming services. AAC at 256kbps (Apple's standard) is perceptually transparent and roughly equivalent to MP3 at 320kbps. AAC handles transients and stereo imaging better than MP3.
OGG Vorbis: Open-source lossy codec used by Spotify (320kbps for Premium) and many games. Quality is comparable to AAC at equivalent bitrates. Good handling of very low bitrates (96–128kbps) where MP3 struggles.
Opus: The newest and best lossy codec. Outperforms everything at every bitrate. Designed for real-time communication but excellent for music. Opus at 128kbps sounds better than MP3 at 192kbps. Growing adoption but not yet universal in music distribution.
Convert WAV to MP3 | Convert FLAC to MP3 | Convert WAV to OGG | Convert MP3 to AAC
What Streaming Platforms Actually Do
Upload WAV or FLAC to your distributor (DistroKid, TuneCore, etc.). They transcode to each platform's required formats:
- Spotify: OGG Vorbis at 96/160/320kbps depending on user's plan and settings. Normalizes to -14 LUFS.
- Apple Music: AAC 256kbps, ALAC lossless, Dolby Atmos (if spatial mix provided). Normalizes to approximately -16 LUFS (Sound Check).
- YouTube Music: AAC up to 256kbps. The video platform uses Opus.
- Tidal: AAC 320kbps (HiFi), FLAC 16-bit/44.1kHz (HiFi Plus), MQA for "Masters" tier.
- Amazon Music HD: FLAC up to 24-bit/192kHz.
The takeaway: upload the highest quality master you have (24-bit WAV or FLAC). The platforms handle all downstream encoding. Never upload an MP3 — re-encoding a lossy format produces worse quality than encoding from lossless.
Sample Libraries and MIDI
Beyond recorded audio, musicians work with samples, loops, and MIDI data that have their own format requirements.
Sample Library Formats
WAV samples: Universal. Every sampler, every DAW. Kontakt, EXS24, SFZ, Decent Sampler — all can load WAV. For sample library creation, record at 24-bit/48kHz minimum. Include loop points in the WAV metadata for sustained instruments.
Kontakt (.nki/.nkm/.nkc): Native Instruments' proprietary format. The dominant format for commercial sample libraries. Kontakt libraries use .nki (instrument), .nkr (resource container), .nkc (cache). You need Kontakt to use these.
SFZ: Open-source sampler format. A text file that maps WAV files to key ranges, velocities, and round-robins. Supported by sforzando (free), ARIA, and others. Use for distributing sample libraries without requiring a specific sampler.
MIDI Files
Standard MIDI Files (.mid) store performance data: note on/off, velocity, pitch bend, control changes, program changes, tempo, and time signature. No audio. File sizes are tiny (10–100KB for a complex arrangement).
Type 0: All tracks merged into one. Simple but loses track separation.
Type 1: Multiple tracks preserved. The standard for DAW exchange. Each track retains its channel assignment and name.
MIDI is the only format that transfers musical performance data between any DAW, any platform, any era. A MIDI file from 1990 plays perfectly in 2026. No other music technology format has this longevity.
Audio format decisions follow a simple gradient: maximum quality for capture and editing, appropriate quality for the distribution channel, and the platform handles the rest. The biggest mistake musicians make is premature lossy conversion — bouncing to MP3 during production, or uploading MP3 to a distributor that will re-encode it again.
Keep everything lossless until the final distribution step. Let the streaming platforms do their own encoding from your lossless source. Your mastering engineer, your distributor, and your listeners' ears will thank you.