Mute Video Online

1. 1

   Add your video

   Drop in an MP4, MOV, WebM or MKV file. Nothing is uploaded — it is read where it sits.

2. 2

   Choose what happens to the sound

   Remove every track, silence only the sections you pick, add a silent track, or keep one track and drop the rest.

3. 3

   Mute it

   The picture is copied across untouched, so the run takes a moment rather than a re-encode.

4. 4

   Download

   Play the result to check it, then save it. A batch comes down as one archive.

Sound is the first thing people ask to remove from a clip, and usually for one of four reasons: the recording picked up a room, a television or a conversation that should not be in it; a voiceover or a music bed is going over the top and the original audio would fight it; a platform wants silent autoplay footage; or the file is going into an edit where the audio track is being sourced somewhere else entirely. In every one of those cases the picture is fine and only the sound is in the way.

What muting gets you

• A clean bed for a voiceover. Recording narration over a clip that still has its own audio means fighting two sources in the mix. Removing the track first makes the edit a single decision.
• Nothing recorded by accident. Phone footage picks up whoever was standing nearby. Muting is the fastest way to be certain nothing private is in a file you are about to post.
• Footage that autoplays. Feeds mute autoplaying video anyway, and a file with no audio track is smaller and starts faster.
• A smaller file. An audio track is typically 128 kbps. Over an hour that is roughly 55 MB removed without touching a single frame of picture.

This is the part that makes muting different from almost every other video edit, and it is worth understanding because it decides whether your file comes back changed. Sound and picture are stored as separate streams inside a container. A frame of video does not know whether there is a soundtrack beside it — that relationship exists only in the container's index. So removing the audio does not require anything to be decoded, altered and encoded again: the video packets are copied across into a new container byte for byte, and the new container simply does not list an audio track.

What that means in practice

• No generation loss. A re-encode throws away detail every time it runs. A copy cannot, because the compressed data is never opened. The frames in your muted file are the identical frames that went in.
• It finishes almost immediately. There is nothing to decode. A ten-second clip at 1080p completes in a few thousandths of a second on a normal laptop; a full-length video is limited by how fast the file can be read, not by encoding.
• The file gets smaller, never larger. You are removing a stream and adding nothing. A tool that re-encodes can easily hand back a bigger file than you gave it.
• Rotation and orientation survive. A phone video stores its rotation in the container. Copying carries it across, so a portrait clip stays portrait.

The two times it does re-encode

• You chose a container that cannot hold your codec. WebM carries VP8, VP9 and AV1 — it cannot legally hold H.264. Ask for WebM from an H.264 source and the picture has to be converted. The format list says which choices re-encode before you pick one.
• You asked for an exact trim start. A copy has to begin at a keyframe, because a frame in between cannot be decoded without the frames it refers to. Ticking exact start re-encodes so the cut lands precisely where you asked.

Removing every sound is often more than you wanted. A cough over one sentence, a swear word, a phone ringing halfway through, a piece of background music that cannot be cleared — the rest of the audio is worth keeping. Add a range for each stretch you want silent, set its start and end, and everything outside those ranges comes through exactly as it was. Ranges that overlap or touch are combined into one, so drawing two spans across the same moment does not produce an odd blip between them.

Why there is a fade at each edge

• A hard cut is audible. Dropping a waveform to zero mid-cycle is a step change, and a step change contains every frequency at once. It is heard as a click, which on a tool whose whole job is silence is conspicuous.
• Ten milliseconds is the default. Short enough that no syllable is lost at the boundary, long enough that nothing snaps. You can take it to zero for a hard cut, or out to a quarter of a second for a gentle duck.
• Short ranges still reach silence. If a range is shorter than two fades, the fade shortens with it. A range you draw is always fully silent in the middle, however brief it is.

What partial muting costs

• The picture is still copied. Only the audio is decoded, scaled and written again. Your frames are untouched, exactly as in a full mute.
• The audio is re-encoded. Changing sample values means the audio cannot be copied. It is written back at a high bitrate, and the difference is not audible on speech or music.
• Length is not a problem. The audio is processed a block at a time and nothing is held in memory, so a two-hour recording costs the same working memory as a ten-second one.

These sound like the same thing and produce genuinely different files. Removing the track means the finished video has no audio stream at all — the container lists one video track and nothing else. Replacing it with silence means the container still lists an audio track; the samples in it are simply all zero. Most of the time you want the first. The second exists because some editing software, some broadcast specifications and some upload pipelines treat a video with no audio stream as malformed and either reject it or fall over on import.

Which one to pick

• Removing is the default. Smaller file, no ambiguity, and every player and platform handles it. Choose this unless something downstream has told you otherwise.
• Pick the silent track if an import failed. If an editor, a scheduler or a delivery spec refused your file for having no audio, this is the fix. The track matches the source's sample rate and channel count, so nothing has to be converted on the other side.
• The silence is real silence. Every sample is zero, not a very quiet noise floor. Turning the volume up finds nothing.

A film file, a screen recording with a separate microphone, or anything exported from an editor can carry several audio streams — an original language and a dub, a mix and a commentary, system sound and a headset. Most tools treat all of that as one thing called "the audio" and give you the choice of keeping every stream or none. Here each stream is listed with its language tag, its name and its channel layout, and you tick the ones that survive.

How it behaves

• The tracks you keep are copied too. Kept audio is moved across without being decoded, exactly like the picture. This is the one mode where both streams come out identical to the ones that went in.
• Language and name come with them. A track labelled French in the source is still labelled French in the result, so your player's audio menu still makes sense.
• It appears when there is something to choose. On a file with one audio stream the option is not offered, because keeping the only track is the same as changing nothing.

If you only need a section of the clip, there is no reason to export the whole thing and cut it afterwards. Turn on trim, set a start and an end, and only that span is written — still muted, still copied rather than re-encoded. It is the same work either way, so the trim is effectively free.

Where the cut actually lands

• A copied cut starts at a keyframe. Compressed video stores occasional complete frames and describes the rest as changes from them. A copy has to begin at one of those complete frames, so the start moves back to the nearest one at or before the time you asked for.
• You are told when that happens. If the start moved, the result says where it actually began. A cut in the wrong place that nobody mentions is worse than one you know about.
• Exact start is there when it matters. Tick it and the video is re-encoded so the cut is frame-accurate. It takes longer and the picture is no longer a copy — which is why it is a choice rather than the default.

Modern MP4, MOV, WebM and MKV files can be opened here, and any of the four can be written. What matters for a lossless mute is whether the container you pick can hold the video codec you already have, and the format list marks the ones that cannot before you choose. Keeping the source's own container is always safe, which is why it is the default.

Picking an output container

• Same as source. Always copies. Nothing is converted, and the result opens wherever the original did.
• MP4 and MOV. Both carry H.264, HEVC, AV1 and VP9, so almost any source copies straight into them. MP4 is the safest thing to hand to somebody else.
• MKV. Carries everything the others do and more, and it is the container to pick when a file has several audio tracks you want to keep.
• WebM. Only holds VP8, VP9 and AV1. An H.264 source has to be converted to go in, so pick it when you specifically need WebM.

What is refused, and why

• AVI, WMV, FLV, MPEG, OGV and 3GP. These carry codecs no modern device can decode without extra software, so the file is refused the moment you pick it rather than after a wait. Convert it first and the converted file mutes normally.
• Animated GIF. A GIF has no audio track to remove, so muting one would do nothing at all.
• The refusal names the file. You are told what the file is, what can be opened, and where to go next — never just "unsupported".

Muting is one of the few video operations where there is no quality trade-off to manage, so most of the advice is about picking the right mode and the right container rather than about settings.

Keep the picture untouched

• Leave the format on same as source. It is the one choice that is guaranteed to copy. Changing container only to change it costs you a re-encode for nothing.
• Prefer keyframe trimming. Unless the exact frame matters, let the start snap. It is instant and the picture stays identical.
• Mute before you compress, not after. Removing the audio first means the compressor has less to work on and one less stream to think about.

Common jobs

• Prepping footage for a voiceover. Full remove, same container. Drop the muted file into your editor and record over it.
• Taking one word out. Partial mute with a single range. Everything else in the recording is untouched.
• Fixing a rejected upload. Silent track mode, so the file has the audio stream the pipeline insists on.
• Dropping a commentary track. Track selection, keeping the main mix, output to MKV so every kept track fits.
• Clearing a batch. Add up to twenty files and download them as one archive rather than clicking through twenty save prompts.

A video file is a container — a wrapper with an index describing which streams are inside, where their data sits and what time each piece belongs at. Muting rewrites that wrapper. The compressed picture data is moved across unread; the index of the new file simply lists a video stream and, depending on what you chose, either no audio stream, one silent one, or the ones you kept. Nothing is decoded on the way through, which is why the operation is both exact and quick.

What happens when you press the button

• The file is read, not uploaded. The container's index is parsed where the file already is. Its duration, dimensions and every audio track's language and codec come from that index.
• A plan is made first. Whether the picture can be copied, what happens to each audio stream, and where a trim will actually land are all decided before any bytes are written.
• The work happens in the background. The page stays responsive, the progress figure is real progress, and cancel actually stops the job rather than hiding it.
• The result is checked before you see it. The finished file is re-opened and its audio streams counted. If a mute somehow left sound behind, you get an error instead of a file you would have trusted.

Hardware-accelerated, on your device: the picture is copied rather than re-encoded, so a ten-second 1080p clip finishes in a few thousandths of a second.

Muting is usually one step in a longer job. These are the tools most often reached for either side of it — pulling the original sound out to keep, putting a new track in its place, or trimming and shrinking the silent file once it exists.

• Extract audio
• Replace audio
• Video cutter
• Change video speed
• Compress video
• Video resizer