How to split an audio recording at silence

You have one long recording, but the music or takes are separated by pauses. Splitting at silence can turn it into smaller clips without hunting for every boundary by hand. The important part is checking each cut: quiet notes, breaths, and reverb tails can look like silence even when you want to keep them.

You can use the Audio Editor to split a recording by silence, then listen to the sections and make any needed adjustments before exporting.

Split the recording at silent gaps

1. Open the Audio Editor and open your recording. If you have already started a multitrack project, work on the track that contains the recording.
2. Listen through once and note which pauses should become boundaries. Decide whether to keep short gaps between phrases, breaths, or the decay after a note.
3. Use the split-by-silence feature on the recording. If the editor offers a silence threshold or minimum gap length, start with a setting that avoids treating every brief dip in volume as a new section.
4. Inspect the resulting clips from the start of the recording to the end. Play a few seconds around each cut, especially where a note fades out or a new take begins quietly.
5. Adjust any boundaries that cut off useful sound. Use cuts or fades where needed, then export the edited audio as WAV or MP3.

Keep the original recording unchanged until you are happy with the split. If the first pass creates too many pieces, make the silence detection less sensitive or require a longer quiet gap, if those controls are available, and run it again.

Choose which gaps should create a split

Use a short gap for clearly separated takes

If each section ends with a clean pause and the next starts clearly, a shorter minimum gap can find the boundaries quickly. Listen closely to the start of each clip afterward. A soft pickup note or the first consonant of a spoken slate can begin before the sound becomes loud.

Use a longer gap when the recording has brief pauses

For a performance with short rests, choose a longer minimum gap if the editor provides that option. This helps prevent a pause between musical phrases from becoming a separate clip. It is usually better to get fewer, larger sections than many tiny clips you have to combine again.

Keep the decay when the music rings out

Reverb is the lingering sound after a note stops. A recording may look quiet before that tail is fully gone, especially when the waveform is zoomed out. If a split removes the natural ending of a chord or note, move the boundary later or add a short fade so the clip does not end abruptly.

Keep room tone if it matters to the demo

Room tone is the low-level background sound in a recording, such as a quiet fan or the noise floor of a microphone. A silence detector may treat it as sound, so a gap with room tone might not register as silence. Decide whether you want that ambience between sections. You can trim the gaps by hand if the automatic split leaves them attached to the clips.

Fix common split problems

1. Undo an overly aggressive pass. If the recording has been split into many tiny pieces, undo the operation if the editor offers undo. Otherwise, reopen the original recording and try again with a less sensitive threshold or a longer required gap, when those options are available.
2. Check for clipped beginnings and endings. Play the first and last second of each section. If a quiet pickup or reverb tail is missing, adjust that boundary rather than lowering the whole track's volume.
3. Join sections that should stay together. A short rest inside one musical phrase may have triggered a split. Join the related clips or remove the extra boundary, then replay the phrase to make sure the timing still sounds natural.
4. Soften clicks at a cut. A cut through a waveform can create a click, even when the gap looks nearly silent. Move the boundary into a quieter point or apply a brief fade at the edge.
5. Export a fresh copy. If the exported file still contains the original long recording or has unexpected gaps, check that the edited version is the one being exported. Export again as WAV for an uncompressed copy, or MP3 for a smaller file to share.

Why silence detection can miss a pause

Silence in an audio editor does not always mean a row of samples with exactly zero sound. It usually means the audio stays below a level for long enough. A quiet room, microphone hiss, or a sustained note can keep the signal above that level. Conversely, a soft note or a fading reverb can drop below it and be mistaken for a pause.

If a threshold control is available, raising the level can make more quiet passages count as silence, but it can also split quiet music. Lowering the level can preserve those passages, but may miss pauses with background noise. A minimum-gap control helps ignore very short rests. There is no single correct value for every recording, so judge the result by listening to the cuts.

Your recording and edited audio stay on your device while you work in the browser.