Audio Trimmer

Cut a section out of an audio file. Drop it in, drag the handles across the waveform or type exact times, preview the selection, and export it as WAV. The cut happens on raw samples rather than by re-encoding, so the cut itself loses nothing. A WAV is decoded at its own sample rate, read from the file header, so nothing is converted; a compressed file is decoded at your browser's audio rate, and the page tells you which rate was used rather than pretending the question does not arise. The waveform shows the minimum and maximum for every pixel rather than one sampled value, which is why short transients appear at all and why the shape does not change when you resize the window. Whatever your browser can play, this can open. Nothing is uploaded.

Export options

DepthSize per minute (stereo 44.1k)Use for
16-bit~10 MBAlmost everything
24-bit~15 MBHanding to an audio editor
32-bit float~20 MBFurther processing without clipping

When to use it

  1. Cut a ringtone out of a song.
  2. Take a quote from a podcast episode.
  3. Trim silence from the start of a voice note.
  4. Cut a sample for a video edit.
  5. Shorten an audio file to fit an upload limit.

How to Use

  1. Drop in your audio file.
  2. Drag the handles, or type exact start and end times.
  3. Add a short fade if the cut clicks, then export.

Features

  1. Waveform with min/max per pixel
  2. Drag handles or type exact times
  3. Preview just the selection
  4. Fade in and out
  5. Sample-accurate, no re-encoding
  6. 16, 24 or 32-bit float WAV
  7. Opens anything your browser can play
  8. Runs entirely in your browser

FAQs

Does trimming lose quality?

The cut does not — the selection is copied sample for sample from the decoded audio, with no re-compression, and exporting as WAV keeps it that way. The honest caveat is decoding: browsers resample to their own audio rate, so a WAV has its rate read from the header and the decoder set to match, while a compressed file is decoded at your browser's rate. The tool shows which rate was used.

Why is the export WAV rather than MP3?

Two reasons. Encoding MP3 in a browser needs a bundled encoder of several hundred kilobytes, and re-encoding audio that was already lossy loses quality again. A WAV handed to whatever you are using next avoids both. Convert afterwards if you specifically need MP3.

Which formats can I open?

Whatever your browser can play, since it does the decoding — MP3, AAC and M4A, WAV, FLAC and Ogg on most browsers. If a file will not open, it is a format your browser has no decoder for rather than a limit here.

Why is the file so much larger?

WAV is uncompressed. A minute of CD-quality stereo is about 10MB whatever it was compressed to before. Drop to 16-bit if you have it higher, or convert to a compressed format afterwards.

How precise is the cut?

To the sample. Drag for a rough selection and then type exact times to refine it — at 48kHz a sample is about 20 microseconds.

What do the fades do?

They ramp the volume up at the start and down at the end of the selection, which removes the click you get when a cut lands mid-waveform. Even 10 milliseconds is usually enough.

Is my audio uploaded?

No. It is decoded, trimmed and encoded in your browser.

Technical Details

The file is decoded with the browser's own `decodeAudioData`, which means whatever your browser can play, this can open — MP3, AAC, WAV, FLAC, Ogg — and nothing else, which is stated rather than discovered halfway through. Decoding gives raw float samples, so trimming is an array slice at a sample boundary rather than a re-encode — the cut itself loses nothing. One caveat worth stating plainly, because most tools in this category leave it out: decodeAudioData resamples to the AudioContext's own rate, so a 44.1kHz file decoded on a 48kHz context comes back converted. For a WAV the rate is read out of the RIFF header first and the decoder created to match, so no conversion happens at all; for a compressed file the rate is whatever the browser's audio engine runs at, and the page shows which rate was used rather than implying the original survived untouched. The waveform is drawn as a minimum and a maximum per horizontal pixel rather than one sampled value per pixel. That distinction matters more than it sounds: a three-minute track across a thousand pixels means each pixel stands for roughly eight thousand samples, and picking one of them draws a shape that changes every time the window is resized, and that hides short transients entirely. Fades are linear in amplitude, which over the short fades these tools produce is what people expect a fade to sound like; a logarithmic curve seems to hold on too long at the end. Export is WAV, at 16, 24 or 32-bit float. WAV rather than MP3 because encoding MP3 needs a bundled encoder of several hundred kilobytes, and because re-encoding lossy audio to lossy audio loses quality a second time — a trimmed WAV handed to whatever you are using next avoids both. The 32-bit float option writes format tag 3 rather than 1, since a float file labelled as integer PCM decodes as noise rather than refusing to open. Sample conversion scales negative and positive by their own limits, because signed 16-bit runs from -32768 to +32767 and treating both as 32768 overflows a full-scale peak into a loud click. Nothing is uploaded, and a long file is limited only by what your browser will hold in memory.

Audio Tools

Audio Tools

  • Audio Merger
  • Audio to WAV Converter
  • Audio Speed & Pitch Changer
  • BPM Detector
  • Tone Generator
  • Voice Recorder
  • Metronome
  • Chromatic Tuner