Audio Converter

Audio comes in many containers and codecs, and the one you have is rarely the one the next program wants. A video editor asks for WAV, a phone wants M4A, a web page prefers Opus. This converter uses your browser’s own decoder to read nearly any file, then writes it again in the format you choose, with an optional new sample rate or a mono mix-down, all without leaving your computer.

Decoding and encoding

Every audio file is a stream of numbers, either the raw sample values (WAV) or a compressed description of them (MP3, AAC, Opus, OGG, FLAC). Converting means decoding to raw samples, optionally processing them, and encoding again. Decoding is free of surprises; encoding to a lossy format throws away information that the codec judges hard to hear. WAV and FLAC keep everything, so converting FLAC to WAV is lossless. Converting MP3 to WAV merely stores the decoded MP3, with its losses, in a bigger file.

A worked example

Press Sample to load a four-second stereo WAV at 44,100 Hz. Convert with the defaults (keep the sample rate, WAV 16-bit) and the page writes sample-song-converted.wav of 689 KB: 4 seconds × 44,100 samples × 2 channels × 2 bytes = 705,600 bytes of sound plus a 44-byte header, which is 689 KB. Tick Mix down to mono and the file halves to about 345 KB. Choose 22.05 kHz and it halves again, since there are half as many samples per second. Choose 32-bit float and it doubles to about 1.3 MB, as each sample takes four bytes.

Sample rate and channels

Lowering the sample rate reduces the highest frequency that can be stored to half the rate: 22.05 kHz keeps sound up to about 11 kHz, which is fine for speech and loses the sparkle of music. The resampler uses a long windowed-sinc filter to avoid aliasing, the false tones that a cheap method adds. Mixing to mono averages the channels, which also removes anything that was out of phase between them; for a spoken recording it halves the size at no real loss.

Choosing a compressed format

For Opus and AAC the Bitrate box appears. A speech recording is clear at 48 to 64 kbps; music is usually good from 128 kbps. Opus only accepts certain sample rates, so a file at an unusual rate is resampled to 48 kHz for it, and both codecs take at most two channels. If the browser lacks an encoder, the page says so and leaves the WAV options.

Edge cases

Browsers decode audio at the sample rate of their audio system for formats other than WAV, so an MP3 recorded at 44.1 kHz may arrive at 48 kHz; the readout shows the actual rate. A file that has an unusual codec, or that is damaged, will show an error. Video files also work, giving the sound track. Metadata such as tags and cover art is not copied to the result.

Limits

MP3 cannot be written because the encoders available for the web do not fit the licence of a free tool; AAC and Opus give better quality at the same size anyway. Gapless playback information and chapter marks are not preserved. The conversion happens on your device, so very long files are limited by memory.

Related tools

• Merge & Join Audio Files combines converted files into one.
• Extract Audio from Video is the quick way to get sound out of a video.
• Audio File Size Calculator predicts the size of a WAV before you make it.
• Audio Cutter trims the file before or after converting.