Choosing the Right Audio Format
By WuTools editorial team · Updated
Five formats cover almost every audio use case in 2026. Pick the wrong one and you either waste bandwidth (WAV everywhere) or quietly lose quality forever (re-encoding MP3 over MP3). This guide compares WAV, FLAC, MP3, Opus, AAC on file size, quality, compatibility, and use case — so you can confidently pick a format before you hit "export". When you're ready to actually convert, our audio toolbox handles all five.
Lossless vs lossy: the only distinction that really matters
Lossless formats (WAV, FLAC, ALAC) reproduce the original samples exactly. Decode a FLAC file and you get the same bits as the original PCM — every bit, no rounding. They are larger but bit-perfect.
Lossy formats (MP3, AAC, Opus, Vorbis) discard information that human hearing won't notice — masked frequencies, near-silence, super-high partials — to shrink the file. The 'loss' is real and irreversible: each generation of lossy re-encoding accumulates artefacts. For master files, edits, and archival, use lossless. For final delivery to listeners, use lossy.
WAV — the uncompressed master
WAV (Microsoft RIFF) stores raw PCM samples with a small header. No compression, no metadata to speak of, near-universal support. A typical 16-bit / 44.1 kHz stereo WAV is 10 MB per minute; 24-bit / 96 kHz is 33 MB per minute.
Use WAV for: working masters during editing, multi-track session files, anything you'll re-render. Avoid WAV for delivery: a 1-hour podcast in WAV is 600 MB; in Opus it's 30 MB. Same audio, 20× smaller.
FLAC — the archival lossless
FLAC (Free Lossless Audio Codec) compresses PCM by 30–60% with no quality loss. It's bit-identical to the original on decode, and supports rich metadata (artist, album, ReplayGain, embedded artwork). Standard for music ripping (CD → FLAC), high-resolution downloads (24-bit / 96 kHz from Bandcamp, Qobuz), and long-term archives.
FLAC is well supported on Android, Linux, Windows, modern macOS (Big Sur+), VLC, foobar2000, and most car head units. Apple's preferred lossless is ALAC, which is technically equivalent and almost always interchangeable. If iTunes / Apple Music compatibility matters, choose ALAC; otherwise FLAC.
MP3 — the universal lossy fallback
MP3 (MPEG-1 Audio Layer III) is the format that won the format war in 1999 and has been universally supported ever since. Every player, every browser, every embedded device. Encoder of choice is LAME; modern V0 (variable bit-rate, around 220 kbps) is transparent for most listeners.
MP3 has aged: at the same bit-rate Opus and AAC sound noticeably better, especially below 96 kbps. But its compatibility is unmatched — if you don't know what device will play the file, MP3 at 192–256 kbps CBR is the safe bet. Old podcast directories, voicemail systems, and many audiobook players still expect MP3.
AAC — the platform-default modern lossy
AAC (Advanced Audio Coding) was designed to replace MP3. It's the default lossy format on Apple platforms (iTunes, Apple Music, AAC-LC at 256 kbps), YouTube (uses AAC-LC alongside Opus), and most streaming services. At 128 kbps AAC is roughly equal to 192 kbps MP3.
Variants matter: AAC-LC for music, HE-AAC for low-bitrate streaming (32–64 kbps), xHE-AAC for the lowest end. The most common file extensions are .m4a (audio-only AAC in MP4 container) and .aac (raw stream).
Opus — the modern open-source king
Opus is the IETF standard codec (RFC 6716) developed by Xiph.org and Mozilla. It is the best lossy audio codec we have at 2026: better than AAC at every bit-rate from 6 kbps voice to 510 kbps stereo music, and the only codec that handles voice and music seamlessly with a single encode (it adapts internally between SILK voice mode and CELT music mode).
Use Opus for: podcasts (32–64 kbps mono is transparent for speech), web audio, video conferencing (Discord, Zoom, WhatsApp use it), modern web video. Limitations: not supported by older Apple devices before iOS 14, and a small share of car head units. For internet-only delivery to modern phones and computers, Opus is the right answer.
Sample rate: 44.1 vs 48 vs 96 kHz
44.1 kHz is the CD standard. Pick it for music intended for CD, music streaming, or anything aiming at parity with the existing music ecosystem. Theoretically supports up to 22.05 kHz audio content — covers full human hearing.
48 kHz is the video standard. Every camera, every video editor, every TV broadcast format runs at 48 kHz. If your audio will end up in a video, record at 48 kHz to avoid resampling.
96 kHz and 192 kHz are high-resolution rates used in studio mastering. The extra bandwidth (up to 48 kHz or 96 kHz) is inaudible to humans, but provides safety margin for processing — pitch-shifting, time-stretching, and some plugins benefit from oversampling. For final delivery to listeners, 96 kHz is usually overkill: downsample to 44.1 or 48 kHz.
Bit depth: 16 vs 24
16-bit (CD standard) gives 96 dB of dynamic range, more than enough for any final delivery. The noise floor of any real listening environment is louder than the quantization noise of 16-bit.
24-bit gives 144 dB and is the standard for recording and editing. The extra headroom matters during processing — applying gain to a 24-bit file leaves more room before you lose the noise floor — but it's invisible after final mastering. Record and edit in 24-bit; master to 16-bit for delivery.
Pick-the-format cheat sheet
Voice memo, transcription source: Opus 24–48 kbps mono. Tiny file, perfectly intelligible.
Podcast distribution: Opus 64 kbps stereo (or 48 kbps mono) for modern platforms; MP3 128 kbps stereo as a fallback for old podcast apps.
Music streaming, casual listening: AAC 192–256 kbps for Apple compatibility; Opus 128 kbps for everything else.
Music archival, hi-fi listening: FLAC at the source rate (CD: 16/44.1; hi-res: 24/96).
Working master during editing: WAV 24-bit / 48 kHz (or 96 kHz if you'll do heavy time/pitch processing).
Phone-call quality: Opus 12–16 kbps narrowband; the codec is designed to be transparent for speech at extremely low bit-rates.
Loudness: a separate problem
Once you've chosen a format, you still have to set the loudness. The current broadcast standard is EBU R128 / ITU-R BS.1770: −23 LUFS integrated for broadcast TV in Europe, −16 LUFS for podcasts (Apple Podcasts target), −14 LUFS for music streaming (Spotify, YouTube). Use our Loudness Meter to measure and the Volume Booster or Audio Compressor to adjust before export.
Related WuTools
- Audio Equalizer — Tone-shaping before export
- Vocal Remover — Karaoke / acapella stem extraction
- Volume Booster — Bring quiet recordings up to streaming loudness
- Audio Fade — Add intro / outro fades before final export
- Audio Joiner — Merge clips losslessly into one file
- Audio Splitter — Cut long takes into per-track files
- Audio Compressor (Dynamics) — Even out the dynamics for podcast / voice
- Loudness Meter — Measure LUFS to hit broadcast / streaming targets
Frequently asked questions
Is FLAC really the same quality as WAV?
Yes — bit-identical. FLAC's compression is lossless, so the decoded samples match the source PCM exactly. The only differences are file size (FLAC ~ 50–70% of WAV) and CPU cost (FLAC needs a tiny amount of decode work).
Should I record at 48 kHz or 96 kHz?
48 kHz for everything that's going into a video. 96 kHz only if you'll do heavy time-stretching, pitch-shifting, or multiple effects passes that benefit from extra headroom. For pure music recording that goes straight to delivery, 44.1 kHz is fine.
Will my audience notice if I export podcast as MP3 vs Opus?
At 64 kbps Opus and 128 kbps MP3 are roughly comparable for speech — both transparent on most listening setups. Opus uses half the bandwidth, which matters for mobile listeners on metered data. The risk with Opus is older podcast apps that don't decode it; MP3 is the safer fallback.
Can I convert MP3 to FLAC to recover quality?
No. Transcoding MP3 → FLAC just stores the lossy MP3 data losslessly — the discarded frequencies are gone. You get a bigger file with no audible improvement.
Why is my exported MP3 quieter than the original?
Encoders apply −0.5 to −1 dB headroom to avoid inter-sample clipping after the lossy decode. If exact loudness matters, normalize to a target LUFS (use our Loudness Meter) and let the player handle volume — don't fight the codec.
What's the difference between .m4a and .aac?
.m4a is AAC audio packaged in an MP4 container (with metadata, chapters, embedded art). .aac is raw AAC stream with no container. Most apps prefer .m4a; some streaming hardware wants raw .aac.
Is 320 kbps MP3 better than 256 kbps AAC?
No, AAC at 256 kbps slightly beats MP3 at 320 kbps in blind listening tests. AAC is a more modern codec — expect ~30% better compression at equal quality.
Why does my voice memo sound bad after upload?
Many platforms re-encode on upload. iPhone Voice Memos are AAC at 64 kbps; if you re-export from a video editor as 96 kbps Opus, you've stacked two lossy encodes. Convert from the original .m4a once, not via intermediate formats.
