Convert MP3 to WAV
Decode an MP3 into the uncompressed form editors and samplers expect.
How it works
This conversion is a decode, not a repair. The MP3 encoder discarded parts of the signal it judged inaudible, and that decision is permanent: no decoder can reconstruct data that was never stored. What comes out is the waveform the MP3 describes, written down in full, which is a far larger file containing exactly the same sound.
Expect roughly ten times the size. A 5 MB MP3 at 128 kbps becomes about 50 MB of 16-bit stereo WAV, because uncompressed audio costs about 10.1 MB per minute at CD settings regardless of what it came from.
The reasons this is still the right move
Audio editors and digital audio workstations want random access to samples. Seeking inside an MP3 means decoding forward from a frame boundary, so most tools decode the whole thing to PCM on import anyway. Handing them a WAV makes that step yours rather than theirs, and avoids the small differences between one program's decoder and another's.
Hardware samplers, older game engines, telephone and IVR systems, Windows event sounds and a good many scientific analysis packages read WAV and nothing else. So do a surprising number of upload forms, particularly for transcription and voice cloning services, which specify WAV because it removes every question about what the decoder did.
Bit depth, and why 24-bit is not an upgrade
Bit depth sets how finely each sample is measured. 16-bit gives 65,536 levels and about 96 dB of dynamic range, which is the CD standard and matches what an MP3 decoder produces for ordinary material. 24-bit gives 16.7 million levels and roughly 144 dB.
Choosing 24-bit here does not recover detail, because there is none to recover. What it does is leave headroom for what happens next: if the WAV is going into a chain of gain changes, filters and effects, the extra levels keep rounding errors from accumulating into audible noise across a dozen operations. As a delivery file, 16-bit is correct and half the size. 32-bit float is for editors that work internally in float and would rather not convert on import.
Sample rate is copied from the source unless you change it. Raising it is pointless, since it invents no detail and only makes the file larger. Lowering it below the source rate discards the top of the frequency range for real.
Questions
Will this improve the quality of my MP3?
No, and nothing can. The information the encoder removed is gone from the file. A WAV made from a 96 kbps MP3 is a large, uncompressed copy of a 96 kbps MP3, artefacts included. Anyone offering to restore or enhance a lossy file is either guessing at the missing content or selling you something.
Is my file uploaded?
No. Decoding uses the browser's own audio decoder, running in a Web Worker on your machine, and the WAV is assembled there too. Watch the Network tab while you convert: after the page loads there are no further requests. This particular conversion needs no extra module at all, because writing uncompressed audio requires no encoder.
How big will the WAV be?
Multiply the length in minutes by 10.1 MB for 16-bit stereo at 44.1 kHz, or by about 15.1 MB for 24-bit. Mono halves both. A three-minute song lands near 30 MB and a one-hour recording near 600 MB. Files above 4 GB are written in the RF64 form, which lifts the ceiling the original WAVE specification imposed.
Why does the WAV sound identical rather than better?
Because it is identical. The decoder reproduces the same samples your music player produces when it plays the MP3, and writes them to disk instead of to the speakers. The value of the conversion is the format, not the sound: what changes is which programs will accept the file and what they can do with it once open.