Guide · Audio basics

128, 192 or 320 kbps? Choose a bitrate you will not regret.

Bitrate is a budget, not a quality score. Here is what each number buys, what it costs in storage, and when the biggest number is the wrong choice.

Bitrate is the single number people use to judge an MP3, and the one most often misunderstood. It is not a quality score. It is a budget: how many bits per second the encoder may spend. This guide explains what that budget buys, when it is wasted, and how to choose a number you will not regret.

What a bitrate actually measures

A bitrate is a data rate, usually written in kilobits per second (kbps). A 192 kbps MP3 spends roughly 192,000 bits on every second of audio. Divide by eight and you get bytes: about 24 KB per second, or 1.44 MB per minute.

MP3 encoders can work in two modes. Constant bitrate (CBR) spends the same budget every second. Variable bitrate (VBR) spends more on complex passages and less on simple ones, usually producing better quality at the same average size. Braise exports CBR because the resulting file size is predictable and every player handles it identically.

What each bitrate costs in storage

BitratePer minutePer hourTypical use
96 kbps≈ 0.72 MB≈ 43 MBMono speech, audiobooks, voice notes
128 kbps≈ 0.96 MB≈ 58 MBPodcasts, casual listening, small downloads
192 kbps≈ 1.44 MB≈ 86 MBMusic on phones and cars — the default sweet spot
256 kbps≈ 1.92 MB≈ 115 MBCritical listening, good headphones, archiving a copy
320 kbps≈ 2.40 MB≈ 144 MBMaximum CBR; only worth it from a clean source

These are decimal megabytes. Your file manager may show mebibytes (MiB), which are about 5% smaller for the same file. A few kilobytes of ID3 tags and padding are added on top.

What you can actually hear

Perception is personal, and it depends on the recording, the headphones and the room. Still, some broad patterns hold:

  • 96 kbps: speech remains intelligible and clean; music starts to sound dull, with smeared cymbals and a weak stereo image.
  • 128 kbps: acceptable for casual music, but dense mixes can develop a metallic edge. This is where the “MP3 sound” reputation comes from.
  • 192 kbps: transparent for most listeners on most equipment. A sensible default for music you will actually listen to.
  • 256 kbps: a safety margin for critical listening. The difference from 192 is real but subtle on typical gear.
  • 320 kbps: the top of the CBR range. Useful for archiving a lossy copy or for editing headroom; not a cure for a poor source.

The source matters more than the number

Every lossy encode discards information. A second encode cannot recover it; it can only approximate the first result again. This is called generation loss, and it is why an MP3 made from a WAV can sound excellent while an MP3 made from another MP3 often sounds worse than the original.

Practical rule: convert once, from the best source you have, and keep that output. If you must re-encode, go from lossless (WAV or FLAC) rather than from a previous MP3.

Mono, stereo and sample rate

Mono halves the data needed for the same perceived quality, which is why speech benefits so much. A mono file at 96 kbps can sound better than a stereo file at 128 kbps for a single voice, because the encoder is not wasting bits on a second near-identical channel.

Sample rate is a separate axis. CD audio is 44.1 kHz; video often uses 48 kHz. MP3 supports both. Braise decodes everything to 44.1 kHz so the output is consistent. Higher-resolution sources (96 kHz, 192 kHz) are resampled; there is no audible benefit in an MP3.

How to choose, in one minute

Recommended starting points
Single voice, audiobook, voice memoMono, 96 kbps
Podcast with musicMono or stereo, 128 kbps
Music on a phone or in a carStereo, 192 kbps
Music on good headphonesStereo, 256 kbps
Archiving a lossy copy from a clean masterStereo, 320 kbps
Source is already a 96 kbps MP3Keep it, or export at 128 kbps at most

Three myths worth dropping

  • “320 kbps is CD quality.” No. A CD is uncompressed 1411 kbps PCM. 320 kbps is a high-quality lossy approximation.
  • “Converting to a higher bitrate improves the file.” No. It cannot add detail that was never captured.
  • “VBR is always better.” VBR is usually more efficient, but CBR is easier to predict, seek and stream. Both are valid; the encoder matters more than the mode.

If you want to test your own ears, encode the same track at 128, 192 and 320 kbps and compare on the equipment you actually use. Most people are surprised by how close 192 is to 320 — and by how obvious 128 can be on a dense mix.

FAQ

Bitrate questions

Is 320 kbps always better than 192 kbps?

No. It is always larger. Whether it is better depends on the source and the listener. A clean CD rip at 320 kbps preserves more detail than 192 kbps; a 96 kbps AAC source re-encoded at 320 kbps carries the same limitations in a bigger file.

Should I use VBR or CBR?

VBR spends more bits on complex passages and fewer on silence, so it usually gives better quality per byte. CBR is simpler and more predictable for streaming and editing. Braise exports CBR because it is transparent about the file size and plays everywhere.

What bitrate is “CD quality”?

There is no honest answer. A CD is 1411 kbps uncompressed PCM. A well-encoded 192–256 kbps MP3 can be perceptually close for most music, but it is not bit-identical. Use 256–320 kbps if you want a generous safety margin.

Does converting MP3 to a higher bitrate improve it?

Never. The missing detail is already gone. Re-encoding at a higher bitrate only makes a larger file with the same artefacts, plus a new generation of loss. Keep the original.

What about 96 kbps for podcasts?

Speech is forgiving. A mono 96 kbps MP3 is a common podcast delivery choice and sounds clean on phones and in cars. Stereo 128 kbps is a reasonable step up for music or mixed content.

Try it on a real file.

Export the same track at two bitrates and compare. That is the only benchmark that matters.