tool 03 / 04 — audio
Cut sound on the waveform
cut locally · nothing uploaded
about this tool
The split.tools audio splitter turns one long recording into separate tracks: a digitized tape into songs, a podcast episode into chapters, a rehearsal recording into takes, a voice memo into clean segments.
Drop a file and the full waveform renders instantly — you cut by sight, clicking where the sound actually changes, instead of scrubbing by ear. A silence detector can rough in the markers for you by finding every stretch of dead air, and per-segment playback lets you audit each cut before exporting.
Tracks export as 16-bit WAV at the source sample rate — lossless from the decoded audio, so no extra compression generation is added to your files. Decoding, analysis, and export all run on your device through the Web Audio API; recordings never touch a server.
how it works
- 1 cm
Drop an audio file on the mat
MP3, WAV, M4A, and OGG all decode locally, and the full waveform draws in seconds — even for hours-long recordings.
- 2 cm
Place your cuts
Click the waveform to drop a marker at each boundary, or press 'Cut at silences' to auto-mark every gap of dead air. Preview any segment with its Play button.
- 3 cm
Export the tracks
Each region between markers becomes a numbered track. One track saves as a WAV file; several arrive as one ZIP, with the timestamp list doubling as chapters.
frequently asked questions
Which audio formats can I split?
Anything your browser can play: MP3, WAV, M4A/AAC, OGG, and FLAC in most browsers. The file is decoded locally with the Web Audio API.
Why do tracks export as WAV instead of MP3?
To avoid adding a lossy compression generation in the middle of your workflow. WAV slices are bit-faithful to the decoded audio; encode to MP3 once, at the final publishing step, and your audio only pays the compression cost a single time.
How does silence detection work?
The tool slides a 20-millisecond window across the recording measuring RMS loudness. Stretches that stay below a quiet threshold for long enough count as silence, and a cut marker is placed at the middle of each. It works best on speech and album rips with real gaps.
Can I split a podcast into chapters with this?
Yes — place a marker at each topic boundary and the segment table gives you both the chapter audio files and the start timestamps for your show notes. YouTube and most podcast apps turn those timestamps into tappable chapters.
Is my recording uploaded anywhere?
No. Decoding, waveform rendering, silence analysis, and WAV encoding all happen in your browser tab. Unreleased demos and private recordings stay on your machine.
Will splitting reduce my audio quality?
No. Cutting selects samples — it doesn't re-encode them. The exported WAV contains exactly the samples your source decoded to, at the original sample rate.