Convert to FLAC
Lossless, and about half the size of a WAV.
or paste one from your clipboard
- Length
- ·
- Sample rate
- ·
- Channels
- ·
- Size
- ·
- Length
- ·
- Sample rate
- ·
- Channels
- ·
- Size
- ·
FLAC size depends on the material. Sparse or quiet recordings compress much further than dense ones, so treat the estimate as approximate.
How to use it
- Drop a WAV, AIFF or other lossless file onto the page. FLAC is worth doing from a lossless source; from an MP3 it only wastes space.
- There is no quality setting to choose. FLAC keeps every sample exactly, so the only thing that varies is how hard the encoder works to shrink it.
- Click Download. The first FLAC of your session loads the encoder, about 31 MB, reported as its own progress step.
- Expect roughly half the size of the equivalent WAV. Sparse or quiet material compresses further; dense loud masters compress less.
How lossless compression is possible at all
Lossy codecs shrink audio by deciding what you will not miss. FLAC shrinks it by noticing that audio is highly predictable from one sample to the next. The encoder takes a block of samples, fits a mathematical predictor to them. Either a fixed low-order polynomial or a linear predictor with coefficients calculated for that specific block, and then stores only the difference between what the predictor guessed and what actually happened.
Those differences cluster tightly around zero, which is precisely the case Rice coding handles well: small numbers get short codes, large ones get long codes. Nothing is discarded at any point; the predictor and the residuals together reconstruct the original exactly. That is also why the compression ratio varies so much with material. A solo cello line is easy to predict and compresses hard. A loud, limited, cymbal-heavy master looks much closer to noise, and noise is by definition unpredictable.
Origins and why it won the lossless argument
Josh Coalson started FLAC in 2000 and released version 1.0 in July 2001, with the reference implementation under a BSD licence and the format specification free of patent claims from the outset. Xiph.Org took the project under its umbrella in 2003, alongside Vorbis and later Opus. In 2024 the format was finally written up as a formal IETF specification, RFC 9639, which matters to the archives and institutions that need a standards-track document before they will commit collections to a format.
Competing lossless codecs existed and some compressed marginally better, but they were proprietary, slow to decode, or both. FLAC decodes fast enough for cheap hardware, seeks properly thanks to its seek table, and costs nothing to implement. That combination is why it ended up in Android, in network streamers, in Bandcamp downloads and in preservation workflows, while its rivals did not.
What FLAC is not for
It is not a repair tool. Everything it does is preservation of what it was given, so a FLAC made from a 192 kbps MP3 is a large, faithful copy of a 192 kbps MP3. It is not a distribution format for casual sharing either. Half of a WAV is still five or six megabytes a minute, which is a poor fit for messaging apps and email.
And it is not the lossless format Apple's ecosystem expects. Apple Music and the iOS Music app are built around ALAC, which is technically similar and equally lossless but a different file entirely. If your library lives in Apple software, FLAC works fine as an archive on disk and badly as the thing you actually play. Keep the FLAC, and make a lossy copy for the device.