Home › Guides › Audio Formats Explained: MP3, WAV, FLAC and AAC

Audio Formats Explained: MP3, WAV, FLAC and AAC

Audio 8 min read

A three-minute song can be 3 MB or 30 MB depending on nothing but the format you saved it in. Understanding why takes about five minutes and saves a lot of guessing later.

The one division that matters

Every audio format falls on one side of a single line: lossy or lossless.

Lossless formats keep every sample of the original recording. WAV stores it raw and uncompressed; FLAC compresses it the way a ZIP file does, typically to around half the size, with nothing discarded. Decompress a FLAC and you get back a file bit-for-bit identical to the WAV it came from.

Lossy formats permanently discard audio on the theory that you will not miss it. They exploit how hearing works — a loud sound masks a quiet one at a similar frequency, and nobody hears above roughly 20 kHz — so they delete what is unlikely to be perceived. MP3 and AAC both work this way, and both achieve enormous size reductions doing it.

The four formats

 MP3AAC / M4AWAVFLAC
TypeLossyLossyLossless, uncompressedLossless, compressed
3 min of stereo audio~3 MB~3 MB~30 MB~15 MB
Quality at same sizeGoodBetter than MP3PerfectPerfect
CompatibilityPlays on literally everythingExcellentUniversalGood, not universal
Best forSharing, maximum compatibilityStreaming, Apple devicesEditing, productionArchiving

MP3

Thirty years old and still the safest choice when you do not know what will play the file. Technically it has been beaten — AAC sounds better at the same bitrate — but its compatibility is unmatched: every phone, car stereo, browser, and cheap MP3 player made this century handles it. When you are emailing a file to somebody and cannot ask what they use, send MP3.

AAC (usually seen as .m4a)

The successor to MP3, and genuinely better at low bitrates — at 128 kbps the difference is audible, and at 96 kbps it is obvious. It is the default across Apple devices and most streaming services. Support is excellent nearly everywhere now, with only old hardware occasionally objecting.

WAV

Raw, uncompressed audio: no compression, no loss, no processing. That makes it the correct format to work in — recording, editing, mixing — because every save is lossless and nothing degrades. It also makes it enormous. Roughly 10 MB per minute of stereo CD-quality audio, which is why WAV is a working format rather than a distribution one.

FLAC

Lossless like WAV, but compressed to around half the size. If you want a perfect archive copy without wasting disk space, FLAC is the right answer. The trade-off is compatibility: excellent on computers and modern phones, still absent from some car stereos and older devices.

Bitrate, and what the numbers mean

Bitrate is how much data is spent per second of audio, measured in kilobits per second. For lossy formats it is the main quality dial: more data means less thrown away.

BitrateSounds likeUse for
64 kbpsNoticeably thin; fine for speechVoice memos, audiobooks
128 kbpsAcceptable; some artefacts on complex musicCasual listening, podcasts
192 kbpsGood; most people hear no problemGeneral music
256–320 kbpsVery hard to distinguish from losslessMusic you care about
The honest position on 320 kbps: in properly conducted blind tests, most listeners cannot reliably tell a good 256 kbps encode from the lossless original on ordinary equipment. Going higher costs storage and buys very little. If you genuinely want perfection, the answer is FLAC, not a bigger MP3.

Generation loss: the thing to actually avoid

Because lossy encoding discards data every time it runs, re-encoding compounds the damage. Convert an MP3 to another MP3 and the second encoder throws away more — working partly from artefacts the first one introduced. Do it a few times and the result is audibly hollow and swirly.

Never treat a lossy file as a master. Keep originals in WAV or FLAC, and export to MP3 or AAC only as the final step. Converting MP3 → FLAC does not restore anything; it just stores the damaged audio losslessly in a bigger file.

Sample rate and channels

Two other numbers appear in audio settings. Sample rate (44.1 kHz, 48 kHz) is how many times per second the waveform was measured; 44.1 kHz is CD standard and captures everything humans can hear, so higher rates matter for production headroom rather than listening. Channels is mono or stereo — and for a single voice recorded on one microphone, mono is not a compromise, it is correct, and it halves the file size for nothing.

Choosing, in practice

  • Sending audio to someone — MP3 at 192 kbps. It plays everywhere and sounds fine.
  • Publishing a podcast — MP3 at 96–128 kbps mono for speech. Bigger files help nobody and cost your listeners data.
  • Archiving your own recordings — FLAC. Perfect quality at half the size of WAV.
  • Editing or mixing — WAV throughout, exporting to lossy only at the end.
  • Filling a phone or old MP3 player — AAC if it is supported, MP3 if you are unsure.
Try it: Switch between formats with Convert Audio, change the bitrate directly using Change Audio Bitrate, shrink a file with Compress Audio, or clean up a recording with Noise Reduction.

Common questions

Does converting MP3 to WAV improve the quality?

No. It produces a much larger file containing exactly the same audio, artefacts included. Nothing that MP3 discarded comes back.

Which is better, MP3 or WAV?

They are for different jobs. WAV is perfect quality and huge, so you edit in it. MP3 is smaller and slightly degraded, so you share in it. Neither is universally better.

Why is my WAV file so large?

Because it is uncompressed — roughly 10 MB per minute in stereo. That is normal. Convert to FLAC to halve it losslessly, or to MP3 to shrink it by around ninety per cent.