← Blog 9 min read

MP3 vs FLAC vs WAV: Which Audio Format Should You Actually Use?

Audio formats are a mess of acronyms that mostly seem to do the same thing. MP3 is the one everybody knows, FLAC is the one audiophiles insist on, WAV is what recording software produces, and then there are AAC, M4A, OGG, Opus, AIFF and WMA cluttering up the list. Choosing badly is not usually catastrophic, but it can quietly cost you audio quality in a way that cannot be undone.

There is really only one distinction that matters. Get it right and the rest is detail.

Lossy and lossless: the only division that counts

Lossless formats keep every sample of the original recording. Decompress a lossless file and you get back the exact bits you started with - not something close, but identical. WAV, FLAC and AIFF are lossless.

Lossy formats make the file dramatically smaller by permanently discarding audio the encoder judges you will not notice - very quiet sounds masked by loud ones, frequencies at the edges of hearing. Done well it is genuinely hard to hear the difference. But it is a one-way door: the discarded detail is not recoverable. MP3, AAC, M4A, OGG, Opus and WMA are all lossy.

The size difference explains why lossy formats took over. Three minutes of CD-quality stereo audio is roughly 30 MB as WAV, around 15 MB as FLAC, and about 4 MB as a good MP3.

Generation loss: the mistake worth avoiding

This is the part that catches people out, and it is the single most useful thing on this page.

When you convert one lossy format to another - MP3 to AAC, OGG to MP3 - the file has to be decoded back to raw audio and then re-encoded. The second encoder has no idea what the first one already discarded, so it analyses the already-damaged audio and discards a fresh round of detail on top. This is generation loss, and it accumulates.

Once is usually fine and most people will not hear it. Do it four or five times - a file that has been through a messenger app, a video editor, a podcast platform and a conversion tool - and it starts to sound hollow and smeared, particularly on cymbals and sibilance.

The rule that follows is simple: always convert from the most original file you have. If you have the WAV, convert the WAV, not the MP3 you made from it last year. And if you are going to produce several different formats, make them all from the lossless source rather than from each other.

The related trap is conversion in the hopeful direction. Turning an MP3 into a FLAC does not restore anything. FLAC will faithfully preserve every flaw the MP3 encoder introduced, in a file three times the size. The one legitimate reason to do it is to freeze a file before repeated editing, so no further generation loss occurs.

The formats, and what each is actually for

WAV - uncompressed PCM, the standard working format for recording and editing. Enormous, universally supported, and no compression to decode makes it the least CPU-intensive to work with. Use it while producing, not for storing a library.

FLAC - lossless compression, typically around half the size of the equivalent WAV. This is the correct format for archiving anything you care about: it is open, unencumbered, widely supported, and bit-for-bit reversible. If you are ripping a CD collection, rip to FLAC.

AIFF - Apple's uncompressed format, functionally WAV's twin. The genuine difference is byte order: WAV stores samples little-endian, AIFF big-endian. It matters to software authors and to nobody else. Convert to it when a specific Apple audio tool asks for it.

MP3 - the lossy format with no compatibility problems anywhere. Technically it has been beaten by every codec designed since, but it plays on car stereos, cheap hardware, and anything with a headphone socket. Choose it when you do not control what will play the file.

AAC and M4A - AAC is the codec; M4A is the MP4-family container it usually travels in. Better quality than MP3 at the same bitrate, and the default across the Apple ecosystem, YouTube and most streaming. Choose AAC over MP3 unless you specifically need MP3's universality.

OGG (Vorbis) - open, patent-free, comparable to AAC. Common in games and open-source software, less supported by consumer hardware.

Opus - the current state of the art, and by a clear margin at low bitrates. Excellent for speech, excellent for music, low latency, fully open. It is the codec behind most modern voice chat. The catch is hardware: many older players have never heard of it. Choose Opus for anything that will be played by a browser or a modern app.

WMA - Microsoft's format from the Windows Media Player era. Effectively obsolete. If you have a library in it, the useful move is converting out - see our guide to rescuing a WMA library.

So which should you pick?

  • Archiving music you care about - FLAC. Lossless, half the size of WAV, and it will still be readable in twenty years.
  • Recording or editing - WAV while you work, then export to whatever you need.
  • A phone or portable player, where space matters - AAC/M4A, or MP3 if the device is old.
  • Maximum compatibility, no thinking required - MP3.
  • Podcasts and spoken word - MP3 for the widest player support; Opus if you control the playback.
  • Web and browser playback - Opus, with an MP3 fallback if you need to support very old browsers.
  • Anything going into Apple software that demands uncompressed - AIFF.

Bit depth and sample rate, briefly

Two numbers describe the raw audio underneath any format. Sample rate is how many times per second the waveform was measured - 44.1 kHz for CD, 48 kHz for most video work. Bit depth is how precisely each measurement was stored - 16-bit for CD, 24-bit in professional recording, where the extra headroom genuinely helps while mixing.

Neither is worth increasing after the fact. Converting a 44.1 kHz recording to 96 kHz adds file size and no information whatsoever.

One specific thing to know about our converters: they preserve the source sample rate and channel count exactly, and FLAC output preserves bit depth up to 24-bit. However, WAV and AIFF output is written as 16-bit. For CD-sourced material and virtually all consumer audio that changes nothing, since the source is already 16-bit. But if you are working with 24-bit studio recordings and need that depth preserved, convert to FLAC rather than WAV - it is lossless, smaller, and keeps all 24 bits.

What our converters do by default

So you know exactly what you are getting rather than guessing:

  • MP3 - LAME quality level 2, a variable-bitrate setting that lands around 190 kbps on typical music. Transparent for nearly all listeners.
  • FLAC - compression level 5, the standard balance of speed and size. Lossless regardless of the level; higher levels only make encoding slower.
  • AAC and M4A - 192 kbps.
  • OGG Vorbis - quality level 5, roughly 160 kbps.
  • Opus - 128 kbps, which for this codec is genuinely high quality.
  • WAV and AIFF - 16-bit PCM, little-endian and big-endian respectively.

One deliberate behaviour worth flagging: all tags and cover art are removed during conversion. That is a privacy decision rather than an oversight, and we explain the reasoning - and how to get your music library re-tagged afterwards - in why your audio tags disappear when you convert.

Converting

Common starting points, all free and without an account:

Every combination of MP3, WAV, FLAC, AAC, M4A, OGG, Opus, AIFF and WMA is on the audio tools page. Conversion happens entirely in server memory - your recording is never written to a disk and is discarded the moment your download is sent, which matters more than people expect for interviews, voice memos and anything else you would not post publicly.

Audio converters:
FLAC to MP3 WAV to MP3 WAV to FLAC MP3 to Opus All audio tools