The short answer is that the 48 kHz music file grew almost exactly 2.5 times when its requested bitrate increased from 128 to 320 kbps. The 24 kHz mono speech file behaved differently: requests for 192, 256, and 320 kbps all produced the same roughly 160 kbps, 1.12 MB result. Selecting a larger number does not guarantee that every MP3 can use it.
Bitrate is the amount of data budgeted for each second of audio. Think of it as giving an encoder more space on each page. More space cannot recreate information missing from the source, but it can reduce how much a good source must discard during lossy compression.
This experiment measured bytes and effective data rates. It did not include a blinded listening panel, so it cannot establish a universal point at which people hear a difference. The measured facts still provide useful rules for choosing a setting.
Music scaled with the setting, while 24 kHz speech stopped at 160 kbps
At 128, 192, 256, and 320 kbps, the files measured 1.47, 2.21, 2.95, and 3.69 MB—almost perfectly proportional to bitrate.
The 24 kHz file uses MPEG-2 Layer III, whose standard bitrate table ends at 160 kbps.
A 320 kbps MP3 is still lossy and cannot restore information that an earlier lossy encode removed.
We encoded two different WAV sources at the same four requested settings
- 1
The music source was a 92.112-second, 48 kHz, stereo, 16-bit PCM WAV measuring 17,685,676 bytes.
- 2
The speech source was a 56.000-second, 24 kHz, mono, 16-bit PCM WAV measuring 2,688,044 bytes. We excluded an already lossy MP3 source because its earlier loss would confound the comparison.
- 3
Toolit's current browser processing path converted both sources with libmp3lame CBR (constant bitrate, which assigns a similar data amount to each second) at requested values of 128, 192, 256, and 320 kbps.
- 4
We recorded the exact byte count of all eight outputs and calculated whole-file average data rate as file bytes × 8 ÷ duration. Small headers and other overhead can make this value slightly higher than the nominal audio bitrate.
- 5
We did not conduct a listening-quality test. These are size and format measurements for two sources with the current encoder path, not a quality ranking for every genre or listening environment.
The 48 kHz music file grew predictably with bitrate
The music outputs closely matched the requested settings. Relative to the 1.475 MB result at 128 kbps, 192 kbps was about 1.5 times larger, 256 kbps about twice as large, and 320 kbps about 2.5 times larger. For CBR MP3 of equal duration, the chosen bitrate is the dominant file-size factor.
At 128 kbps, audio receives about 128,000 bits—or 16,000 bytes—per second. That is approximately 0.96 MB per minute. The same arithmetic gives about 1.44 MB per minute at 192, 1.92 MB at 256, and 2.40 MB at 320, plus small file overhead.
| Requested setting | Music output | Calculated average rate | Relative to 128 |
|---|---|---|---|
| 128 kbps | 1,474,605 bytes · 1.475 MB | 128.07 kbps | 1.00× |
| 192 kbps | 2,211,885 bytes · 2.212 MB | 192.10 kbps | 1.50× |
| 256 kbps | 2,949,165 bytes · 2.949 MB | 256.14 kbps | 2.00× |
| 320 kbps | 3,686,445 bytes · 3.686 MB | 320.17 kbps | 2.50× |
The 24 kHz speech file became 160 kbps at every higher request
The speech source produced a different pattern. The 128 kbps output measured 0.897 MB, but requests for 192, 256, and 320 kbps all produced the exact same 1,121,805-byte file and an average whole-file rate of about 160.26 kbps.
A 24 kHz MP3 uses the MPEG-2 Layer III range. LAME's official usage documentation lists 8–160 kbps for that range. MPEG-1 Layer III at 32, 44.1, or 48 kHz supports up to 320 kbps. The encoder therefore settled on 160 kbps within the format's valid range.
Do not assume that a 320 kbps selection always creates a 320 kbps file. Low-sample-rate speech and other sources with different format constraints should be checked after conversion.
| Requested setting | Speech output | Calculated average rate | Observed result |
|---|---|---|---|
| 128 kbps | 897,453 bytes · 0.897 MB | 128.21 kbps | Encoded at 128 |
| 192 kbps | 1,121,805 bytes · 1.122 MB | 160.26 kbps | 160 kbps ceiling |
| 256 kbps | 1,121,805 bytes · 1.122 MB | 160.26 kbps | 160 kbps ceiling |
| 320 kbps | 1,121,805 bytes · 1.122 MB | 160.26 kbps | 160 kbps ceiling |
A higher bitrate cannot restore detail missing from the source
A higher bitrate gives the encoder more room to preserve information. It does not recover sound removed by an earlier lossy encode. Converting an existing 128 kbps MP3 to 320 kbps can make the file larger while adding another lossy generation.
LAME's official guidance also says that lossy codecs such as MP3 should not be used for archiving. Keep WAV or FLAC when you may edit or convert again, and create MP3 copies for delivery and playback.
A practical way to choose
For voice recordings, lectures, and meetings, 128 kbps is a sensible place to start when compatibility matters. In a source like this 24 kHz speech file, choosing 256 or 320 provides no bitrate benefit because the valid MP3 range stops at 160.
For general music sharing, 192 kbps can be a starting balance between size and data budget. Consider 256 or 320 when the source is good, storage is less constrained, and you specifically want more room for a lossy delivery copy. Whether listeners hear a difference depends on the source, content, equipment, and listener.
- Speech, lectures, and meetings: start at 128 kbps and verify the actual output
- General music sharing: compare 192 kbps against your size limit
- More encoding headroom with fewer size constraints: consider 256 or 320 kbps
- Archival and future editing: retain WAV or FLAC
- Already-lossy MP3 input: a higher MP3 setting will not recover the source
Do not turn this size test into a universal listening claim
This test measured one music file and one speech file. Different instruments, high-frequency content, noise, stereo width, and encoder versions can change how a given bitrate sounds.
We did not run a blinded listening comparison. The data cannot prove that people always hear a difference between 192 and 320 kbps—or that they never do. It does establish the predictable CBR size relationship and the format ceiling observed with 24 kHz MP3.
Official documentation used for this analysis
Choose for the source and job, not the largest number
For the 92-second 48 kHz music source, increasing 128 to 320 kbps grew the output from 1.475 MB to 3.686 MB—almost exactly 2.5 times. For the 56-second 24 kHz speech source, requests for 192, 256, and 320 all became the same 160 kbps, 1.122 MB file.
Check whether the source is speech or music, its sample rate, the delivery limit, and the playback environment before choosing. Preserve important originals losslessly and treat MP3 as a purpose-sized delivery copy.