Trim Silence from Audio

Remove edge silence using an explicit threshold; it does not perform speech-aware noise removal.

Choose audio file

Silence threshold (dBFS)

Padding to keep (seconds)

Detected keep range

—leading/trailing trim only

Choose an audio fileThis removes quiet audio only from the beginning and end; it does not delete pauses inside the recording.

Trim Silence Audio: Threshold & Cut Audit

Audit silence trimming with dBFS threshold, padding, detected active range, removed duration and retained-percentage evidence.

Source duration (s)

Detected active start (s)

Detected active end (s)

Audio processing cluster

Verify the source before trusting the render

Decode locally, confirm sample rate, channels, duration and level evidence, then apply the transformation. Keep codec, browser decode and export-format limits explicit.

Start: Audio Editor Inspect a File

All 24 audio tools

Sample RateResample rate, Nyquist and output frames Stereo → MonoDownmix, correlation and cancellation risk Speed ChangerDuration, pitch and resampling behavior Loop MakerLoop boundaries, seam and repeated duration Frequency AnalyzerFFT resolution and spectral evidence Bass BoosterLow-shelf gain and clipping headroom Audio JoinerClip order, transitions and output duration Loudness AnalyzerLUFS planning, peak and dynamics

Interior silence remover and cut preview

The original edge trimmer above remains the fastest start/end cleanup. This workspace uses the same selected source file to detect longer quiet gaps inside the recording, preview the cut plan, render a browser-local WAV, and audition it before export.

Silence threshold (dBFS)Minimum interior silence (s)Padding kept at each side (s)Join fade (ms)

#	Detected silence	Removed range	Removed

Choose a source file in the primary trimmer above first. Detection is amplitude-based, not speech-aware.

Codec and measurement boundaries

Browser decoding differs across Safari, Chrome, Firefox, iOS, and Android. Rendered edits are exported as WAV unless a tool explicitly states another supported container. Loudness analysis is a practical BS.1770-style estimate rather than a certified meter, and sample clipping checks do not replace true-peak oversampling.

Audio render integrity

Compatible files are decoded locally. Source and rendered signal metrics distinguish the file container from decoded PCM, and generated exports are treated as current only for the settings that produced them.

Practical guide and verification

Use the tool first, then apply these checks to verify the inputs, interpret the result, and hand it off without displacing the primary workflow.

Decide whether you need edge trim or interior pause removal

The original workflow above trims only quiet material before the first active sample and after the last one. The new gap-removal workspace targets longer quiet runs inside the recording. Keeping these operations separate prevents a simple head-and-tail cleanup from unexpectedly changing pacing in the middle of speech or music.

Tune threshold and minimum-gap duration together

A high silence threshold can classify quiet speech or reverb as silence, while a very short minimum gap can remove natural phrasing. Start conservatively, inspect the detected gaps, and listen to the rendered preview. The detector uses amplitude rules rather than speech understanding, so context still requires human review.

Keep padding and a short fade around cuts

Cutting exactly at a threshold crossing can sound abrupt and can create clicks when waveform values jump. Retaining a small amount of padding around active material and applying a short fade at joins makes the result safer to audition. Very large padding can erase the time savings of gap removal, so verify the projected duration.

Preserve the original and audit the WAV render

Browser decoding support depends on the source codec, and the cleaned export is rendered as WAV rather than preserving every original codec parameter. Keep the untouched source, compare original versus projected duration, preview the result through the important transitions, and export only after the detected gap list matches the intended edit.