Audio Converter
Any format in, any format out. Nothing leaves your device.
or paste one from your clipboard
- Length
- ·
- Sample rate
- ·
- Channels
- ·
- Size
- ·
- Length
- ·
- Sample rate
- ·
- Channels
- ·
- Size
- ·
Change the format in the bar below and this updates. WAV and MP3 need no download; every other format loads a converter once.
How to use it
- Drop your audio file onto the page, or click to browse. The file is read straight off your disk. It is never sent to a server.
- Check the waveform and play it back, so you know the file decoded correctly before you spend time encoding it.
- Pick the output format. MP3 and WAV encode immediately; the other formats load a converter once, and the page tells you before it starts.
- For lossy formats, choose a bitrate. The estimated output size updates as you change it.
- Click Download. The finished file is written in the tab and saved straight to your downloads folder.
What a conversion actually does
There is no such thing as changing an audio file's format in place. Every conversion is two separate operations: a decoder turns the compressed bitstream back into a plain list of numbers, one amplitude value per channel, tens of thousands of times per second, and then an encoder packs those numbers back down using an entirely different set of rules. The middle stage is the same raw signal regardless of what you started with, which is why a file that has been through several formats carries the scars of all of them.
That middle stage is also why the file you get back has no metadata. Tags, cover art, chapter marks and gapless-playback hints live in the container, not in the audio, and the container is discarded when the audio is decoded. Nothing here reads them and nothing writes them. If a podcast episode arrives with chapters and you convert it, the chapters are gone.
Lossless, lossy, and which direction you are travelling
WAV, AIFF and FLAC are lossless: decode one and you get back precisely the numbers that went in. MP3, AAC, Vorbis, Opus and WMA are lossy, and they work by modelling what human hearing cannot detect, quiet sounds sitting next to loud ones at nearby frequencies, mostly, and simply not storing it. The discarded information is gone permanently.
This makes direction matter enormously. Lossless to lossy is a normal, sensible operation: you choose how much to throw away. Lossy to lossless is almost always a mistake, because it faithfully preserves damage while multiplying the file size. Lossy to a different lossy format is the worst case of all, because the second encoder wastes bits describing the first encoder's artifacts as though they were music. If you must transcode between lossy formats, give the second encoder more headroom than the first had: a 128 kbps MP3 re-encoded at 192 kbps AAC survives better than one re-encoded at 128.
Why some formats are instant and one download is not
WAV needs no encoder at all. It is a 44-byte header followed by the raw samples, so it is written the instant you click. MP3 uses a compact JavaScript port of the LAME encoder, around 110 KB, which loads faster than most images on a typical web page.
AAC, Vorbis, Opus, FLAC, AIFF and WMA have no equivalent. Getting them into a browser means shipping a WebAssembly build of ffmpeg, which is roughly 31 MB. It is served from this site's own domain, downloaded once, and kept in memory for the rest of your session, so the second and subsequent conversions are as fast as the first was slow. Choosing a format that needs it is always an explicit choice on your part, and the page says so before the download begins rather than after.