BYTETOOLS

Split Audio by Silence

Automatically cut a long recording into separate tracks wherever it goes quiet. Ideal for tape rips, voice memos and interview sessions, with no upload.

Drop an audio file here or click to browseDetection and cutting both happen locally in your browser — nothing is uploaded.

What is the Split Audio by Silence?

The ByteTools Silence Splitter scans a recording for quiet gaps and turns each stretch of audio between them into its own file.

  • Detects quiet gaps automatically and cuts a separate track between each one
  • Adjustable threshold, minimum gap, minimum track length and edge padding
  • Lossless stream copy keeps the original audio bytes untouched
  • Every track is listed with its time range, length and an inline player
  • Splits up to 40 tracks from a single file
  • All detection and cutting happen locally - nothing is uploaded

How to use the Split Audio by Silence

  1. 1

    Drop a long audio file onto the box, or click to browse.

  2. 2

    Set the silence threshold in dB and how long a gap must last to count as a split.

  3. 3

    Set a minimum track length to discard fragments, and padding to protect the edges.

  4. 4

    Choose to keep the original format for a lossless copy, or re-encode to MP3.

  5. 5

    Click Split by Silence, then preview and download each track from the list.

About the Split Audio by Silence

The ByteTools Silence Splitter scans a recording for quiet gaps and turns each stretch of audio between them into its own file. Point it at a digitised cassette side, a run of voice memos recorded back to back, or an interview with clear pauses, and it hands you a numbered list of tracks ready to download.

You control what counts as silence. Set the level threshold, how long a gap must last before it becomes a split point, the shortest track worth keeping, and how much padding to leave either side so words are not clipped. Fragments below your minimum length are discarded, so a cough does not become track seven.

Detection and cutting both run in your browser through a WebAssembly build of ffmpeg, and nothing is uploaded. Keeping the original format copies the audio across untouched, which means no re-encoding, no quality loss, and a much faster job than re-rendering every track.

Frequently asked questions

How do I split one long recording into separate songs?

Drop the file in, leave the threshold near -35 dB, and set the minimum silence to about one second so only real gaps between songs count. Each stretch of audio becomes its own numbered track you can preview and download individually.

What silence threshold should I use?

Start at -35 dB. If the tool finds no gaps, the quiet parts are not actually that quiet, so raise the threshold toward -25 dB. If it cuts in the middle of songs, the threshold is too high - lower it toward -45 dB or lengthen the minimum silence.

Why did it say no silence long enough was found?

The gaps in your file never drop below the threshold for as long as you asked. Vinyl and cassette rips have surface noise sitting above true silence, so raise the threshold a little, or shorten the minimum silence length, and run it again.

Can I download all the tracks at once?

Not in one click. Browsers cannot save a folder of files without bundling them into a zip, and this tool deliberately adds no extra libraries. Each track has its own download button, and they save straight to your normal downloads folder.

Does splitting reduce the audio quality?

Not if you keep the original format. That option copies the existing audio data straight into each new file without decoding it, so the tracks are bit for bit the same as the source. Only the re-encode to MP3 option changes the audio.

Related tools