An MP3 at 128 kbps takes 0.96 MB per minute, 192 kbps takes 1.44 MB and 320 kbps takes 2.4 MB. 320 kbps is the highest bitrate the MP3 format allows, and it only helps if the source had that much detail to begin with: re-encoding a 128 kbps file at 320 kbps makes it 2.5 times bigger without making it sound better. For listening on phones and earbuds, 192 kbps is a sensible middle.
Size per minute, per song and per hour
An MP3's size is its bitrate multiplied by its length. Bitrate is measured in kilobits per second and a byte has 8 bits, so kilobytes = kbps × seconds ÷ 8. The table counts 1 MB as 1,000,000 bytes and leaves out tags and cover art, which add a little on top.
| Bitrate | 1 minute | 4-minute song | 1 hour |
|---|---|---|---|
| 128 kbps | 0.96 MB | 3.84 MB | 57.6 MB |
| 192 kbps | 1.44 MB | 5.76 MB | 86.4 MB |
| 256 kbps | 1.92 MB | 7.68 MB | 115.2 MB |
| 320 kbps | 2.40 MB | 9.60 MB | 144 MB |
| WAV, CD quality (1,411.2 kbps) | 10.58 MB | 42.34 MB | 635.04 MB |
The WAV row is for comparison: CD-quality audio is 44,100 samples per second × 16 bits × 2 channels = 1,411,200 bits per second, uncompressed. Windows counts 1 MB as 1,048,576 bytes, so it shows slightly smaller figures; a 9.6 MB song appears as about 9.16 MB.
Why the ceiling is 320 kbps
MP3 (MPEG-1 Audio Layer III) doesn't allow arbitrary bitrates. For 32, 44.1 and 48 kHz audio the permitted values, as listed in the LAME encoder's documentation, are 32, 40, 48, 56, 64, 80, 96, 112, 128, 160, 192, 224, 256 and 320 kbps. There is no standard 384 or 512 kbps MP3. If 320 kbps isn't enough for your purpose, the answer is a lossless format such as FLAC or WAV, not a bigger MP3.
Why 320 kbps can't rescue a 128 kbps file
MP3 is lossy. When an encoder makes a 128 kbps file, it decides which parts of the sound you're least likely to notice and discards them. That information is no longer in the file. Decoding the 128 kbps MP3 gives you audio that already lacks those parts, and encoding it again at 320 kbps carefully preserves that reduced version, using more bits to do it.
It can even get slightly worse. Each lossy encode makes its own decisions, so a second pass can add new artifacts on top of the old ones. The rule is to encode once, from the best source you have (a WAV, a FLAC or the original video), and choose the bitrate at that point.
The same goes for audio taken from video. The soundtrack of an MP4 is usually already compressed AAC; converting it with Video to MP3 at 320 kbps keeps it as close as possible to what's in the video, but can't go beyond it.
CBR vs VBR
CBR (constant bitrate) spends the same number of bits on every second, so the size is exactly predictable, which is why the table above works. VBR (variable bitrate) lets the encoder spend more bits on complex passages and fewer on simple ones; LAME bases each frame's bits on how much error the ear is estimated to tolerate, trading a predictable size for more even quality.
Which bitrate for what
- Speech (lectures, audiobooks, voice memos): 128 kbps is plenty.
- Music on a phone, earbuds or in the car: 192 kbps.
- Music you're keeping long-term or playing on good speakers: 320 kbps, or keep a lossless FLAC copy and make MP3s from it when needed.
- Ringtones and short clips: size barely matters; 30 seconds at 192 kbps is 720 kB.
- Editing: work in WAV and export to MP3 at the end, so you encode only once.
If space is tight, consider AAC (M4A) or Opus rather than pushing MP3 lower; both generally reach similar quality at lower bitrates than MP3, though MP3 still plays on more old hardware. You can change format and bitrate with the Audio converter, and trim clips with the Audio cutter.