How to trim and clean a voice recording in your browser

Start with a copy of your recording

Open the audio cutter and choose a voice recording at least twenty seconds long. This example keeps the region from 12 to 20 seconds. For a shorter file, choose timestamps within its duration instead.

The tools read and process audio locally. Keep the original file, and save your completed download before reloading or closing the tab. Each input is limited to 100 MB; the current edit is limited to ten minutes and 128 MB of decoded mono or stereo audio.

Keep an eight-second excerpt

Enter 12 in Start and 20 in End, then choose Preview selection. Adjust the boundaries to avoid cutting through a word. Click Keep selection: the current edit should now show eight seconds.

Selecting a region alone does not change the exported file. Keep selection makes it the current edit. Undo restores the previous recording while the session remains open.

Reduce noise gently

Choose Enhance your audio in the studio workflow to carry the current clip into the enhancer. Select Clear voice as a starting point, then lower noise reduction if the source is already clean. Click Apply settings.

Compare Original recording with Current edit at the same device volume. Pay attention to soft consonants and breaths. If speech sounds metallic or watery, Undo and try less noise reduction. These filters cannot reliably remove competing speech, strong echo, or clipping.

Choose the output for your next step

Use WAV when another editing pass is planned. Use MP3 or M4A for convenient sharing, or Opus when the receiving app supports it. Converting compressed audio to WAV or FLAC does not restore information lost in the source.

Click Prepare download and save the completed file. File extensions and MIME types match the selected output, but playback support still depends on the receiving device.

References

• FFmpeg audio filters
• ffmpeg.wasm usage