128, 192 or 320 kbps? Choosing the right MP3 bitrate

What kbps actually means

Bitrate is how much data the file spends on each second of sound. "kbps" means kilobits per second, where one kilobit is 1,000 bits and eight bits make a byte. So a 192 kbps MP3 uses 192,000 bits, or 24,000 bytes, for every second of audio, whether that second is a whisper or a full orchestra.

For comparison, uncompressed CD audio runs at 1,411.2 kbps. A 192 kbps MP3 is roughly one-seventh of that, and 128 kbps is about one-eleventh. The encoder's job is to decide what to leave out so that you don't notice the difference.

How MP3 saves space

An MP3 encoder splits the sound into short frames and many narrow frequency bands. It then runs a psychoacoustic model, a set of rules about what human hearing can and can't pick up, and spends its limited bits where they matter most.

• Masking. A loud sound hides quieter sounds at nearby frequencies, and briefly just before and after it. The encoder stores those hidden details roughly or not at all.
• High frequencies. At lower bitrates, the encoder filters out the very top of the frequency range, where hearing is least sensitive, to save bits for the midrange.
• Stereo tricks. Left and right channels are usually very similar, so the encoder can store what they share once and only the differences separately.

When there aren't enough bits to go around, you start to hear it. Cymbals and applause turn "swishy," and sharp attacks like castanets or a plucked guitar can pick up a faint pre-echo. Higher bitrates give the encoder more room, so those artifacts fade away. Past a certain point, extra bits buy you almost nothing you can hear.

File sizes at 128, 192 and 320 kbps

Because the bitrate is fixed, file size is simple arithmetic: bitrate × seconds ÷ 8. Here's what that looks like, with 1 MB counted as 1,000,000 bytes (your computer may show slightly smaller numbers if it counts in 1,048,576-byte units).

The files Konvert creates don't include tags or cover art, so real sizes land very close to these figures. To put them in perspective, a thousand four-minute songs at 192 kbps take about 5.8 GB. Videos can be up to 1 hour long, so the largest MP3 you can make here is about 144 MB at 320 kbps.

Constant vs. variable bitrate

Constant bitrate (CBR) spends the same number of bits on every second. It's predictable: you know the file size in advance, and every player seeks through it accurately.

Variable bitrate (VBR) lets the encoder spend more on complex passages and less on simple ones, aiming for a quality level rather than a size. It's usually more efficient. The popular LAME presets V2 and V0, for example, tend to average around 190 kbps and 245 kbps, depending on the music. The trade-off is less predictable sizes, and a few older players show the wrong duration or seek inaccurately.

Konvert uses constant bitrate. We checked the pipeline: the setting you choose is passed to FFmpeg's LAME encoder as a fixed bitrate (FFmpeg's -b:a option), not as a VBR quality level. So a 192 kbps file from Konvert really is 192 kbps from start to finish.

Why 320 kbps can't beat the source

An MP3 is always made from something. Here, that something is the audio track of an online video, which is already compressed. YouTube, for example, typically streams audio as Opus or AAC at bitrates that are well under 320 kbps. Anything that encoder discarded is gone before Konvert ever sees the file.

Encoding that stream at 320 kbps makes a very faithful copy of the compressed audio, including its limitations. It doesn't restore detail. It's like scanning a printed photo at a higher resolution: you get a bigger file of the same print. What 320 kbps does do is minimize the small, extra loss that any new lossy encode adds, which is why it's a fine choice when storage doesn't matter. If you need zero added loss, for example to edit and re-export, pick WAV or FLAC instead. The audio formats guide explains that trade-off in detail.

Hear it for yourself: a blind test

The only opinion that matters is your ears'. A short blind test tells you whether 320 kbps is worth the space for your music and your headphones.

1. Make two versions. In the converter, convert the same video twice, once at 128 or 192 kbps and once at 320 kbps. Pick something with cymbals, applause, acoustic instruments or long reverb tails, since those are the hardest to encode.
2. Use your best listening setup. Wired headphones in a quiet room work best. Most Bluetooth headphones re-encode audio with their own lossy codec, which can hide the difference you're testing for.
3. Pick a short passage. Choose 10 to 20 seconds where you think a difference would show.
4. Let someone else press play. Ask a friend to play one of the two files at random without telling you which. You say "low" or "high." Repeat ten times and keep score.
5. Read the result. Guessing gets you about five right. Scoring 9 or 10 out of 10 by pure luck has roughly a 1-in-93 chance, so a score like that means you really can hear it. Anything less suggests the smaller file is just as good for you.

Many listeners are surprised by the result, and that's the point. Test once, then choose with confidence.

Which bitrate to choose

When to skip MP3 altogether

MP3's biggest strength is that it plays everywhere. If you don't need that, two other formats are often better:

• OPUS for the smallest files and the least loss. When a video's audio is already Opus, as it typically is on YouTube, Konvert saves that stream without re-encoding it. That makes it the closest thing to the original, and usually smaller than a 192 kbps MP3.
• M4A for iPhone, iPad and Mac. AAC generally sounds better than MP3 at the same bitrate, and if the source is already AAC it's copied as-is.

Pulling audio for a video edit or a podcast episode? See extracting audio for editing, podcasts and transcripts. And if a conversion doesn't finish, the troubleshooting guide explains why and what to try next.

Questions about bitrate