Audio Converter

Choosing a format

MP3 is the universal option. Every device made in the last twenty-five years plays it, and if compatibility is your concern the decision ends here. It is not the most efficient codec available, but it is the only one nothing refuses.

M4A (AAC) is meaningfully better than MP3 at the same bitrate and is the default across Apple's ecosystem and on YouTube. Choose it when you know the destination handles it.

Opus is the most efficient of the group by a wide margin, particularly below 96 kbps — it is what most voice and video calling systems use. Support outside browsers and modern software is limited, so it suits web delivery rather than general sharing.

OGG Vorbis is the older open-format option, still used in games and by some open-source software. WAV is uncompressed and belongs in editing workflows, not distribution.

Bitrate, and what is actually audible

For music in MP3, 192 kbps is the point at which most listeners on most equipment stop being able to identify compression. 256 and 320 provide headroom for good headphones and for further editing, at proportionally larger files.

128 kbps is acceptable for casual listening and noticeably thin on good equipment, particularly in cymbals and reverb tails. Below that, music degrades quickly.

Speech is completely different. A podcast or voice note is intelligible at 64 kbps and perfectly pleasant at 96, because the human voice occupies a narrow frequency range. Encoding an interview at 320 kbps triples the file for no audible benefit.

AAC and Opus both deliver equivalent quality at lower bitrates than MP3, so if you choose one of those you can drop a step from these numbers.

Sample rate and channels

44.1 kHz is the CD standard and the default for music. 48 kHz is the standard for video and film audio. Converting between them is harmless; converting downwards to 22.05 kHz halves the data but removes everything above about 11 kHz, which is fine for speech and audibly dull for music.

Channels are the bigger lever for spoken content. A single microphone recording stored as stereo carries two identical channels — converting to mono halves the file size with no loss whatsoever, since there was never any stereo information to begin with.

For music, keep stereo. The spatial information is a substantial part of the listening experience and collapsing it is very audible.

Extracting audio from video

Drop a video file in and the video stream is discarded, leaving only the audio. This is the usual way to get a lecture recording, an interview or a piece of music out of a video container.

The conversion is a re-encode rather than a copy, so a little quality is lost. That is unavoidable when the source and target codecs differ, and inaudible at sensible bitrates.

One practical note on lossy-to-lossy conversion generally: each generation compounds the previous one's artefacts. Converting an MP3 to M4A and back to MP3 sounds worse than either step suggests. Convert once, from the best source you have.