Remove Silence from Audio Online
Review Every Pause Before Export
Built for remote standups, interview recordings, audiobook chapters, and voice memos: find the quiet stretches, decide how much to trim, and preview the result before you save anything.
You can also drag and drop an audio or video file anywhere on this page.•Audio and video files with a browser-compatible audio track: MP3, WAV, M4A, AAC, FLAC, OGG, WebM, MP4, MOV, MKV, and more
A Silence Remover That Shows Its Work
Detection runs on RMS energy, not speech recognition, so it behaves the same whether the recording is in English, another language, or has no speech at all. Nothing is discarded silently — every candidate section is marked in red first.
Every Cut Is Visible First
Red overlays mark the exact ranges scheduled for removal on the waveform before anything leaves the exported file.
Tune Three Independent Controls
Minimum duration, retained pause, and the dBFS threshold interact, so small adjustments reshape the plan without guesswork.
A/B the Result as Often as You Want
Toggle between the trimmed preview and the untouched original before committing to an export.
How to Remove Silence from Audio
A short workflow for cleaning long pauses without cutting quiet words or intentional breaks.
Open a Recording
Choose a file from your device or use the sample audio. The editor decodes it locally and draws the waveform before running silence analysis.
Set What Counts as Silence
Enable removal, then adjust the minimum duration, retained pause, scope, and dBFS threshold. Confirm that the red regions contain pauses rather than quiet speech.
Preview and Export
Play the edited timing, compare it with the original, and export when the pacing sounds right. Only the output omits the approved red sections.
What the Silence Remover Actually Does
Each item below is a real control in the editor — there is no speech transcription or AI voice classification behind the result.
Minimum Silence Duration
Brief dips are ignored; only regions that stay below the threshold for the chosen duration make it into the plan.
RMS-Based dBFS Threshold
The analyzer classifies short windows of audio energy against a threshold you set in decibels.
Retained Pause at the Edges
A small buffer is kept around every removed section so breaths and consonants are not cut abruptly.
Scope: Everywhere or Edges Only
Run detection across the full selection, or restrict it to the beginning and end of the file.
Export With Boundary Fades
Approved ranges are omitted from the output and short fades are applied at each join to avoid clicks.
Non-destructive Preview
Playback skips planned removals while the original decoded audio stays available for comparison at any time.
Recordings Where This Actually Saves Time
Automatic detection works best when dead air is predictable. In these workflows it speeds up the first pass, though a human listen-through is still worth doing before you publish.
Edges Only, Nothing in Between
Use the edges-only scope to strip lead-in and trailing silence while leaving the body of the recording untouched.
Interviews and Panel Recordings
Shorten dead air between answers while keeping enough space for the exchange to still sound like a conversation.
Voice-over and ADR Takes
Close long gaps between separate takes before moving on to volume, EQ, or compression passes.
Recorded Lectures and Webinars
Trim the pauses where a presenter was thinking or switching slides, without hunting for every gap by hand.
Fast Iteration on Threshold
Loop the preview repeatedly while dialing in threshold and retained pause until the joins sound seamless.
Staging for Further Edits
Clean up timing first, then reuse the same session to cut, fade, adjust gain, or reshape tone.
Continue Editing the Recording
Use a focused tool for the next change, or open the full editor to combine several adjustments in one session.
Detection Runs in Your Browser, Not on a Server
For a file selected on this device, decoding, level analysis, preview, and export all run in the current browser session — that audio is not uploaded to the application. Normal page services can still make network requests, and importing from Google Drive uses a separate transfer flow.
Local audio analysis
RMS measurements and time-range detection use the decoded samples already held in browser memory.
Original file preserved
The source file is never overwritten. Removal only applies to the new file created during export.
Device-dependent limits
Large or long recordings need more memory and processing time, which shows up most on mobile devices.
Visible detection plan
Analyze, review, preview, export
const audioCtx = new AudioContext();
// Quiet ranges are calculated in this tab
redRanges = detectByRms(threshold)
preview(redRanges) → ready
Questions About Removing Silence from Audio
Practical answers about detection, speech pacing, formats, privacy, and export behavior.
Does this only work well on English speech?
No — detection is based on RMS energy, not on recognizing words, so it behaves the same on any spoken language or on audio with no speech at all, such as music or ambience.
How does the tool decide what is silence?
It measures RMS audio energy in short windows and compares it with the dBFS threshold. A region must remain quiet for at least the selected minimum duration before it is marked in red.
Can it remove silence only at the beginning and end?
Yes. Choose the edges-only scope to target leading and trailing silence. Choose all pauses when you also want to review quiet gaps inside the recording.
Will it remove breaths or quiet words?
It can if the threshold is too high or the retained pause is too short. Detection is based on level, not meaning, so inspect every red region and compare the preview with the original before exporting.
Does it permanently change my original audio?
No. Settings and preview are non-destructive. The original file remains unchanged, and silence is omitted only from the new file created when you export.
Can I use the silence remover on a phone?
Yes, in a modern mobile browser with enough available memory for the recording. The controls and waveform adapt to touch screens, although long files may process faster on a computer.
Which output formats are available?
The editor can export MP3, WAV, M4A, or FLAC. Compressed formats are re-encoded, and actual import compatibility depends on the browser decoder or the local fallback processor.
Product evidence
What This Silence Remover Does — and Does Not Do
The page describes the same level-based workflow exercised by the repository tests. Clear limits help you choose safe settings instead of expecting speech recognition or automatic editorial judgment.
Verified behavior
- • Detects quiet ranges at the beginning, middle, and end of synthetic recordings.
- • Updates the red removal plan when threshold and pause settings change.
- • Skips removed ranges during preview and joins the remaining pieces during WAV export.
- • Keeps the playhead synchronized when automatic repeat reaches removed trailing silence.
Important limitations
- • It cannot distinguish an intentional dramatic pause from unwanted dead air.
- • Background noise above the threshold can prevent a gap from being classified as silence.
- • An aggressive threshold can remove quiet speech, room tone, or musical tails.
- • Input decoding and practical file size depend on browser and device capabilities.
Silence Removal Is Different from reduce rumble and hiss
These solve two different problems. Silence removal shortens ranges whose level stays below a threshold — useful for dead air, long gaps between spoken passages, and empty space at the edges of a recording. Noise reduction targets the hiss or hum that persists while someone is actually speaking, and this tool does not do that.
The best result usually comes from a controlled reduction, not from eliminating every quiet moment. Natural speech includes breaths, filler words, and short pauses that help a listener follow along. Start conservatively and treat the red overlays as a proposed edit list that still needs a listen, not a final decision made for you.
Settings Guide for Natural-Sounding Results
The three controls work together: the threshold defines how quiet a region must be, minimum duration keeps a brief dip from becoming a cut, and retained pause decides how much space is left at each join.
- 1
Pick the right scope first
Use edges-only when the goal is just trimming setup and teardown silence. Use all pauses for interviews or lectures only when you can review the internal edits afterward.
- 2
Start at the default threshold, then nudge it
Enable removal and watch the red overlays appear. If a quiet word gets marked by mistake, lower the threshold until only genuine silence qualifies.
- 3
Raise the minimum duration if speech feels choppy
A longer minimum ignores quick dips and leaves the short conversational pauses that keep pacing natural.
- 4
Keep a slightly wider retained pause for conversational audio
Interviews and panel discussions often have overlapping breaths and filler words — a wider buffer avoids joins that sound abrupt.
- 5
Listen twice before exporting
First the edited version, then the original. Pay close attention to soft openings, word endings, and any room tone that should have stayed.
💡 For recordings made in a noisy room — traffic, a fan, echo — don't just keep raising the threshold. Reduce the background noise first, then come back to the silence remover for a more conservative pass.
Examples of Safe Silence-Removal Workflows
Remote standup or meeting recap
Trim the dead air while someone was unmuting or thinking, so a recap stays short without losing context.
Interview and panel editing
Shorten unusually long pauses between answers while keeping the exchange sounding like a real conversation.
Voice-over and ADR
Tighten gaps between separate takes before applying volume, EQ, or compression.
Recorded lecture or webinar
Shorten setup delays and long pauses in explanations without manually selecting every gap.
Voice memo cleanup
Remove empty time from a long memo before sharing a more concise copy.
Music and ambient recordings
Prefer edges-only detection and carefully protect reverb tails, quiet intros, and intentional rests.