WAV to MP3 Encoder

How the WAV To MP3 works

WAV to MP3: getting a good result

Converting audio to MP3 has been the standard reason people hand a private recording to a random website, and it has never needed to be. A complete MP3 encoder is a few hundred kilobytes of code, and browsers have been fast enough to run one faster than real time for years. This tool loads a pinned copy of that encoder from this domain on your first export, feeds it the samples your browser already decoded, and assembles the frames it returns into a file your browser then downloads. There is no upload step to skip, because there was never a server in the path.

What the bitrate actually buys you. MP3 is lossy: the encoder throws away detail it models as inaudible, and the bitrate is the budget it has to work within. At 320 kbps the budget is generous enough that most people cannot tell the result from the source on most material. 192 kbps is the sensible default for music you will listen to rather than master from, and it is roughly a seventh the size of CD-quality WAV. 128 kbps is the old web standard and still perfectly usable for podcasts and voice notes. 96 kbps is for speech where file size matters more than fidelity — it will sound thin on cymbals and strings, because that is precisely the content that costs the most bits.

Why the source format barely matters. The page never parses the file itself. It hands the bytes to the browser's own audio decoder, which returns raw floating-point samples, and those samples are what gets encoded. Anything the browser can play, it can convert: WAV, MP3, M4A and AAC, FLAC and OGG on current desktop browsers. Decoding also means a lossy source is decoded back to samples before re-encoding, so converting an existing MP3 to a different bitrate is a genuine re-encode and loses a little more detail. Convert from the original whenever you still have it.

Sample rate is carried through, not changed. MP3 supports a fixed set of sample rates — 8, 11.025, 12, 16, 22.05, 24, 32, 44.1 and 48 kHz — and the output keeps whatever your browser decoded to, which on desktop is almost always 44.1 or 48 kHz. A file at some other rate is refused with a clear message rather than silently resampled, because resampling is a decision about quality that belongs to you and not to a converter.

Mono when you mean it. Keeping the source channel layout is right for music. Folding to mono is right for voice recordings, phone calls, interviews and anything destined for a single speaker: it halves the data the encoder has to describe, so a mono file at 128 kbps sounds better than a stereo one at the same bitrate. The fold averages the two channels rather than dropping one, which preserves everything panned to either side.

What encoding costs you in time. The encoder runs on the main thread and processes audio several times faster than real time on a modern machine, so a three-minute track takes a few seconds and the page is briefly busy while it works. Long files — a whole rehearsal, a lecture — will take proportionally longer and hold both the decoded samples and the finished MP3 in memory at once, so very large sources are better split before converting.

Other sources have their own pages explaining what each transcode keeps and loses: FLAC to MP3, M4A to MP3, AAC to MP3, OGG to MP3, Opus to MP3 and WebM to MP3.

Related tools