Convert WAV to MP3

Drop WAV files here

or click to choose — 16-bit, 24-bit and 32-bit float, one or many

Each WAV is checked first: its loudest point, whether the two channels are really the same, and whether one side is silent. The result says what was found and what was done about it.

Convert WAV recordings to MP3 for email, podcasts, upload forms and phones. Before encoding, each file is checked: peaks above full scale in 32-bit float WAVs are lowered so the MP3 doesn't clip, stereo files whose two channels are identical become mono at half the size, and a recording with sound on only one side is turned into mono from that side. Everything runs in your browser and nothing is uploaded.

Why WAV files are so big, and what MP3 does about it

A WAV file stores every sample as a plain number, with no compression at all. CD-quality stereo (16-bit, 44.1 kHz) takes about 10.6 MB per minute, and the 24-bit and 32-bit float files that recorders and music software export are half as big again or double. MP3 keeps what the ear notices and drops what it doesn't, so the same minute fits in 1 to 2 MB. That's the difference between a recording you can email and one you can't.

FormatSize per minuteOur 20-second test file
WAV, 16-bit stereo10.6 MB3.53 MB
WAV, 24-bit stereo15.9 MB5.29 MB
WAV, 32-bit float stereo21.2 MB7.06 MB
MP3, 256 kbps stereo1.92 MB641 KB
MP3, 128 kbps stereo or mono0.96 MB321 KB

MP3 is lossy, so keep the WAV if you'll edit the recording again. For listening and sharing, 256 kbps stereo is very hard to tell apart from the original, and that's the default here.

32-bit float WAVs can go above full scale

Field recorders that record in 32-bit float, and music software exporting a mix, can produce peaks louder than "full scale", the ceiling a normal audio file can hold. A float WAV keeps them, so nothing sounds wrong in your editor. The MP3 encoder works with ordinary samples, though, and anything above the ceiling gets cut off flat: a harsh crackle on every loud peak.

We measured it with a test file peaking 4 dB above full scale. Converted as it was, the MP3 came back with 301,850 samples pinned flat at the ceiling. With the default setting, the converter finds the loudest point first and lowers the whole file just enough to put it at −1 dB, 5 dB in this case. The resulting MP3 had no clipped samples at all. Each result says how far the peaks went and how much the volume was lowered. A WAV that never goes above full scale is left at its own level.

"Stereo" recordings that are really mono

Lots of WAV files have two channels without any stereo in them. Phone recorders, video calls and interview kits often write the same signal to both sides, and an MP3 of that spends half its bytes on a copy. The converter compares the two channels sample by sample. When they're identical, the MP3 is mono: our 20-second test came out at 321 KB instead of 641 KB, with the same sound.

The other common case is a lavalier or USB microphone plugged into one input, so the voice is on the left and the right is silent. Played as stereo, it sounds as if it's coming from one earbud. The usual fix, mixing the two sides down to mono, averages the voice with silence: on our test file the mono version came out 6.5 dB quieter, at half the signal level. Here, a silent side is detected and the mono MP3 is taken from the channel that has the sound, at its original level. You can still choose Always stereo if you need both channels as they are.

Choosing the bitrate

MP3 bitrateSize per minuteGood for
64 kbps mono0.48 MBVoice notes, lectures, meeting recordings
128 kbps mono0.96 MBInterviews and podcasts where quality matters
128 kbps stereo0.96 MBMusic for background listening, upload limits
256 kbps stereo1.92 MBMusic, mixes, demos you send to others (default)
320 kbps stereo2.4 MBThe highest MP3 allows; little audible gain over 256

If a file has to fit under a limit, divide the limit by the length: a 4-minute song under 10 MB can go up to 320 kbps, while a 60-minute interview under 25 MB needs 48 kbps mono or lower.

Tags and timing

If the WAV carries a title, artist, album or date (the "INFO" fields many recorders and editors write), they're copied into the MP3's ID3 tag, and the result lists which ones came across. MP3 encoders add about 25 ms of silence at the start. This converter writes the header that tells players to skip it, so the MP3 starts on the same sample as the WAV. In our tests the start matched to the sample in stereo, mono and single-channel files.

Your recordings stay on your device

The WAV is read and encoded inside your browser, using a WebAssembly build of the LAME MP3 encoder. Nothing is uploaded, there's no account, and there's no limit on file size beyond what your device's memory can hold. Long recordings take a little longer, but they never leave your computer.

Frequently asked questions

Does converting WAV to MP3 lose quality?

Yes, a little: MP3 is lossy and WAV isn't. At 256 kbps stereo, the default, the difference is very hard to hear on music, and the file is about a fifth of the size of a 16-bit WAV. Keep the WAV if you'll edit the recording again, and share the MP3.

How much smaller will the MP3 be?

A 16-bit stereo WAV takes about 10.6 MB per minute. At 256 kbps the MP3 takes 1.92 MB per minute, and at 128 kbps 0.96 MB. Our 20-second test file went from 3.53 MB to 641 KB at 256 kbps. 24-bit and 32-bit float WAVs shrink even more, because the MP3 size depends only on the bitrate.

Why does my 32-bit float WAV crackle after converting?

Its peaks probably go above full scale. A float WAV can store that, but an MP3 encoder cuts those peaks off flat, which sounds like crackling on loud parts. In our test, a file peaking 4 dB over full scale came out with 301,850 clipped samples. This converter measures the peak and lowers the volume just enough to avoid it, and says by how much.

My recording only plays in one ear. Can the MP3 fix that?

Yes. If one channel is silent, as happens with a microphone plugged into one input, the converter makes a mono MP3 from the channel that has the sound, so it plays in both ears. A plain stereo-to-mono mix would average the voice with silence and make it 6.5 dB quieter, as it did in our test.

What bitrate should I choose for a podcast or voice recording?

For speech, 64 kbps mono is clear and takes under half a megabyte per minute. 128 kbps mono is a safe choice for a podcast with music beds. If the WAV's two channels are identical, the default setting already makes a mono MP3.

Can it convert several WAV files at once?

Yes. Drop as many as you like; each is checked and converted in turn, and you can download them one by one or all together as a ZIP.

Are my WAV files uploaded?

No. The WAV is read and encoded inside your browser with a WebAssembly build of the LAME encoder. Your recordings never leave your device, and there's no account or file limit.