How to reduce MP3 size

What decides the size

An MP3's size is its bitrate times its length. A 320 kbps file is about two and a half times the size of one at 128 kbps, however long it is. A smaller file means a lower bitrate, one channel instead of two, less audio, or a more efficient format. Making an MP3 smaller always means encoding it again, so some quality is lost; Audacity's manual warns that editing an MP3 and exporting it as MP3 loses quality twice, so keep a copy of the original.

In the browser, without uploading

The audio compressor takes MP3, WAV, M4A, FLAC, OGG and other audio files, one or several at once. The Compress MP3 page opens the same tool at 96 kbps. The options are:

• Compression: Light (160 kbps), Balanced (128 kbps), Small (96 kbps) or Smallest (64 kbps).
• Mono, which gives the whole bitrate to one channel. It suits speech.
• Fit under: 8 MB, 16 MB or 25 MB. The tool works out the bitrate from the length so that the file fits, using a lower one than your setting when it has to.
• Save as: MP3, or OGG with Opus, which sounds better than MP3 at low bitrates but is not played by a few older players.

Several files download as a zip, and each keeps its name with -compressed added. Everything happens inside your browser tab, so the audio never leaves your device.

What it saves

We tested a generated 60 second stereo clip of a tone with noise, encoded as a 320 kbps MP3 of 2.29 MB. The results, as shown by the tool:

• Light, 160 kbps: 1.14 MB, 50% smaller.
• Balanced, 128 kbps: 938 KB, 60% smaller.
• Small, 96 kbps: 704 KB, 70% smaller; as OGG Opus, 712 KB.
• Smallest, 64 kbps with Mono: 469 KB, 80% smaller.

A 128 kbps MP3 of the same clip (938 KB) came out at 704 KB at Small, 25% smaller, and at Light the tool gave back the original file unchanged, with a note that it was already smaller than that setting would make it, rather than a bigger copy. A test tone is not music, so these figures show the sizes to expect and not how the result sounds: for speech, 64 to 96 kbps is usually fine, and for music stay at 128 kbps or more.

Fitting under a size limit

Pick a Fit under size and leave the rest. The bitrate is the highest of 160, 128, 112, 96, 80, 64, 48 or 32 kbps, at or below your setting, that keeps the file under the limit, with 3% kept back for headers. An hour of audio needs about 54 kbps to fit in 25 MB, so it is written at 48 kbps, which is fine for speech, and works best with Mono ticked. If a very long file still comes out over the limit, the result says so and the answer is to split it, for example with the MP3 trimmer.

In Audacity

Audacity's manual says to open File, Export Audio and choose MP3. It offers variable, average and constant bit rate modes, and a default of about 170 to 210 kbps variable, which it says is not generally recommended for podcasts. Pick a lower setting or a lower fixed bit rate for a smaller file. To convert a WAV recording, see how to convert WAV to MP3.

Questions