About this tool
Audio compression here reduces file size through a lower lossy encoding bitrate for uploads, messaging, and web delivery. It differs from dynamic-range compression, which changes the gap between loud and quiet passages; this tool primarily changes encoded data volume.
Acceptable settings depend on content. Solo speech usually tolerates lower bitrate than complex music, while background noise, reverberation, and high-frequency instruments consume more coding capacity. Judge by listening instead of pursuing the smallest number alone.
How to use it
Upload and note source size
Check source quality and duration and avoid another generation from heavily compressed input when possible.
Choose a target bitrate
Begin with the highest value that can meet the size need and step down for comparison.
Listen and compare
Check sibilance, metallic artifacts, stereo image, and quiet passages before downloading.
Supported range and limits
- How it compresses
- Re-encodes MP3 at a lower bitrate; duration, sample rate and channel count are unchanged
- Typical result
- 320 kbps down to 128 kbps saves roughly 60%; whether speech suits a lower step depends on the source and an actual listening check
- Irreversible
- A lossy re-encode. High frequencies already discarded do not return if you raise the bitrate later
- Poor candidates
- Files that are already low-bitrate; compressing again degrades audibly for little saving
- vs Change Bitrate
- This one frames the choice around shrinking the file; Change Bitrate is for hitting an exact number
- Where it runs
- FFmpeg compiled to WebAssembly in your browser; the file is never uploaded
When you would use it
Meeting an attachment limit
Reduce a recording to fit an email, form, or messaging platform.
Optimizing spoken delivery
Create a smaller podcast preview, course narration, or voice-note version.
Getting speech under an attachment limit
A lower bitrate can reduce a meeting recording substantially, but the result still needs an actual listen before delivery.
What to know before you start
- Lower bitrate permanently discards sound information, so keep a high-quality original.
- Size also depends on duration and channels; twice the duration needs roughly twice the data at one bitrate.
- Dynamic-range compression, loudness normalization, and codec compression are separate processes.
Related concepts
- codec compression
- An encoding process using perception to reduce the amount of data representing sound.
- dynamic range
- The level difference between the quietest and loudest audio passages.