How to use Audio Compressor
- 1Drop in an audio file
- 2Pick a preset — Voice, Small, Standard, High or Maximum
- 3Or set a target size in megabytes and let it work backwards
- 4Compare the finished size against the original
- 5Download the compressed file
What Audio Compressor does
Most audio compressors offer a bitrate slider and stop there. That misses where the savings actually are, because a file's size is the product of three things: how many bits per second the encoder spends, how many channels it encodes, and how many samples per second each channel holds. Bitrate is only one of the three.
For speech that omission is expensive. A voice recording in stereo at 44.1 kHz is storing two nearly identical channels and a frequency range that contains almost nothing — human speech lives below about 8 kHz. Folding to mono halves the data immediately, and dropping the sample rate to 22 kHz halves what remains, before the bitrate is touched at all.
When to use it
- Voice notes and interviews
- The Voice preset changes all three settings at once and routinely produces a file a tenth the size of the original with no meaningful loss for speech.
- Podcast episodes
- An hour of stereo audio at full quality is a large download. Compressed for speech it becomes something that streams comfortably on a phone connection.
- Attachments that will not send
- Mail servers commonly refuse anything over 20–25 MB. Target-size mode lets you name a figure rather than guessing at settings and re-checking.
- Archiving recordings you rarely play
- Lectures, meetings and voice memos kept for reference rather than listening pleasure do not need full fidelity.
Audio Compressor settings explained
- Voice
- 64 kbps, mono, 22 kHz. All three levers pulled at once, which is why it is far smaller than 64 kbps alone would be. Ideal for speech, podcasts and voice notes; wrong for music.
- Small
- 96 kbps, keeping the source channels and sample rate. A noticeable but acceptable quality cost on music, and a substantial size reduction.
- Standard
- 128 kbps. The usual choice, and hard to distinguish from the original on most material through ordinary headphones or speakers.
- High
- 192 kbps. Transparent for nearly everyone on nearly everything, while still being a fraction of a lossless file.
- Maximum
- 320 kbps, the highest a lossy format offers. Worth choosing if the file will be re-encoded later; otherwise the saving over High is modest and the quality gain is not audible.
- Target size
- Name a size in megabytes and the bitrate is derived from it and the recording's duration. Useful against a hard attachment limit, where the constraint is the number rather than the quality.
Why Voice is so much smaller
| Preset | Bitrate | Channels | Sample rate |
|---|---|---|---|
| Voice | 64 kbps | Mono | 22 kHz |
| Small | 96 kbps | Unchanged | Unchanged |
| Standard | 128 kbps | Unchanged | Unchanged |
| High | 192 kbps | Unchanged | Unchanged |
| Maximum | 320 kbps | Unchanged | Unchanged |
Limits and what to expect
- Compressing an already-lossy file adds a second generation of loss whatever preset you choose.
- Voice mode discards frequencies above about 11 kHz, which ruins music but is inaudible on speech.
- 150 MB per file, and long recordings decode slowly because samples are held uncompressed in memory.
- Target-size mode estimates from duration and bitrate, so the result lands near your figure rather than exactly on it.
- Exporting to a format other than MP3 or WAV downloads a ~32 MB engine once, then caches it.
Troubleshooting
- Music sounds muffled after using Voice
- Voice drops the sample rate to 22 kHz, which removes everything above about 11 kHz — cymbals, air, brightness. That range carries nothing in speech and a great deal in music. Use Standard or High for anything musical.
- The file barely got smaller
- It was already compressed at or below your chosen preset. Re-encoding at a similar bitrate cannot shrink it further and only adds loss. Check the original's bitrate before compressing again.
- The target size was not hit exactly
- The bitrate is calculated from your figure and the duration, and the encoder varies slightly around it. Aim a little below a hard limit — 4.5 MB against a 5 MB cap — rather than exactly at it.
- A long recording will not load
- Decoded audio is uncompressed in memory: an hour of stereo is several hundred megabytes regardless of the compressed size. Split long recordings before compressing.
Your files and your privacy
The file is decoded and re-encoded inside your browser tab on your own device. Nothing is uploaded. That matters more here than for most audio tools, because the recordings people compress to send are frequently interviews, meetings and voice notes — conversations that were not meant to pass through a third party.
What to use next
FAQ
How do I make an audio file smaller without losing quality?
For speech, the Voice preset achieves a large reduction with no meaningful loss, because it removes stereo and high frequencies that speech does not use. For music there is no free saving — every reduction is a genuine trade, though at 192 kbps most people cannot hear it.
Why is the Voice preset so much smaller than 64 kbps alone?
Because it changes three things at once: bitrate, channels and sample rate. Folding stereo to mono halves the data, dropping to 22 kHz halves what remains, and only then does the bitrate apply. A bitrate slider on its own leaves the other two savings untouched.
What bitrate should I use for a podcast?
The Voice preset if it is speech only — 64 kbps mono at 22 kHz is entirely adequate and produces a file that downloads quickly. If the episode has music beds or the production quality matters, Standard at 128 kbps keeping stereo is the safer choice.
Can I compress to a specific file size?
Yes — enter a target in megabytes and the bitrate is calculated from it and the recording's length. The result lands close to your figure rather than exactly on it, so aim slightly under a hard limit.
Does compressing twice make it worse?
Yes. Each lossy encode works from an already-degraded signal and discards more, so damage accumulates. Always compress from the highest-quality source you have rather than from a file you compressed earlier.
Is my audio uploaded?
No. Decoding and encoding both run in your browser on your own device, with nothing transmitted to any server.