What is an MP3 file?

MP3 is a codec: the encoding itself rather than the box it travels in. It is used for handing a finished file over, phones and streaming.

The extension is .mp3, and the full name is MPEG Audio Layer III. Both matter less than what the file can hold, which is what the rest of this page is about.

Fraunhofer IIS published it in 1993. The specification is ISO/IEC 11172-3.

A format that has been readable for that long is a format worth trusting with something you want back in ten years.

It is published in full, so anyone can implement it from the document rather than by inspection, which is why it turns up in so many programs and why files written twenty years ago still open. A published specification is not the same thing as a royalty-free one: where a format wraps a codec, the patent licensing is a separate question the standard does not settle.

MP3 discards detail to stay small, and what it discards does not come back. That is the right trade for something being looked at and the wrong one for something being edited: each re-save takes another bite.

MP3 can hold a layout that lets it start playing before it has finished arriving.

That matters mostly when converting: whatever the target cannot hold is dropped, usually without a warning.

Up to 2. Two covers stereo and is what almost everything is; more than two is surround, and it is the part most likely to be folded down when converting to a format that does not define it.

Audacity, VLC and iTunes read it, and so do most programs of the same kind.

If a file will not open, the format is rarely the problem — it is more often that the program predates it. Converting to something older is the reliable way past that, and it is what the rest of this site is for.

Every current browser reads it.

That makes it a safe thing to put on a page or attach to a message without wondering what the other end has installed.

MP3 can hold ID3 tags.

Whether any of it survives a conversion depends entirely on the target, and the honest answer is usually "some of it".

MP3 is meant to be handed over rather than worked in. Editing one is possible and rarely pleasant; the sane approach is to change the source and export again.

The recurring complaints: it degrades a little every time it is re-saved.

None of these is a reason to avoid the format. They are the things worth knowing before you are surprised by one, which is a different claim and a more useful one.

AAC sounds better at the same bitrate. Opus sounds better again, especially low down. Both are newer, both are technically superior, and neither has displaced MP3 for the one job that matters most often: handing an audio file to someone whose equipment you know nothing about.

That is the whole case for it in 2026. Every car stereo, every phone, every browser, every cheap player, every upload form and every piece of software written since 1995 reads an MP3. Compatibility compounds in a way that quality does not, and a file that plays everywhere beats a better file that plays in most places.

MP3 was covered by a patent pool held by Fraunhofer and Technicolor, and encoders required a licence for most of its life. The last of those patents expired in 2017, and Fraunhofer announced the licensing programme had ended.

The practical effect was on software rather than on files. Free and open-source projects could ship MP3 encoders without legal review, distributions stopped separating them into restricted repositories, and the awkwardness of a format everyone used but not everyone could legally write simply evaporated. Nothing about existing files changed.

For speech — interviews, podcasts, voice memos, lectures — 128 kbps is comfortable and 96 is defensible. For music, 192 kbps is the sensible default and 320 is the point past which almost nobody can distinguish it from the source in a controlled test.

The useful number is what an hour costs: about 57 MB at 128 kbps, 86 at 192, 144 at 320. Below 128 the damage becomes audible on music first in cymbals, applause and reverb tails, which are exactly what the encoder finds hardest to model.

Constant bitrate spends the same bits on silence as on a full orchestra. Variable bitrate spends according to difficulty, producing a smaller file at the same quality, and is the better choice for anything being kept.

Its one recurring annoyance is that some players estimate duration from file size and bitrate rather than reading it properly, so a VBR file shows a wrong length and seeking lands in the wrong place. A correctly written VBR header fixes this and most encoders write one; when a file misbehaves this way, that header is usually missing rather than the file being damaged.

MP3 discards audio it judges you cannot hear. Encoding an MP3 again discards more, guided by a model that does not know what the first pass removed — so quality falls faster than the bitrate suggests, and a 320 kbps file made from a 128 kbps file is larger and worse than the original.

If several versions are needed at several bitrates, make each from the best source you hold rather than from each other. This is also why converting an MP3 to WAV restores the file size and none of the audio: the discarded material is gone and no format brings it back.

MP3 has no native metadata, so tags were bolted on. ID3v1 is 128 bytes at the end of the file with fixed-length fields — 30 characters for a title, a genre chosen from a numbered list — and it is why old players truncate long track names.

ID3v2 sits at the front, has no practical size limit, handles Unicode, and holds cover art. Most files carry both for compatibility. When a track shows a mangled title in one player and the correct one in another, the two are usually disagreeing and the player is reading the older tag.

A great many recordings are stereo files carrying one microphone duplicated across both channels — Zoom calls, phone memos, single-source interviews. Encoding that as stereo spends twice the bitrate saying the same thing twice.

Mono at 96 or 128 kbps sounds identical to stereo at double the rate for a single voice, and halves the file. Keep stereo for music, for anything with a real stereo field, and for recordings with two people on separate channels.

AAC when the destination is a phone, a computer or a streaming platform — better at the same size, and everything modern reads it. Opus for voice, where it is dramatically better at low bitrates. FLAC when the recording is an archive copy and nothing may be lost. WAV when an editor or a transcription tool demands uncompressed input.

MP3 when you do not know what will open it. That is a narrower case than it used to be and it remains extremely common: an attachment to a stranger, a file for a car, an upload to a form whose accepted list was written years ago.

Encoding runs inside the browser tab, so the audio is never transmitted. That matters more than it sounds for what people actually convert: interviews under embargo, medical and legal recordings, voice notes, unreleased music.

It also removes the waiting. There is no upload to sit through and no daily allowance to run out of; the free tier stops at 100 MB a file and a hundred files to a batch, so a folder of hour-long recordings goes through in rounds of a hundred rather than one drop.