Unuploaded

Converter · No upload · No re-encode

Remove the audio from a video on your device

Drop an MP4, M4V or MOV and get the same video back without its audio, the picture not re-encoded.

Statusidle Files— File— Video track— Duration— Audio track— Audio bytes removed— File size— Network since file read—

No re-encode. Capture date, location and other metadata can remain in both modes; rows count other non-video tracks kept/removed, not metadata. Refused by name: Matroska and WebM, AVI, fragmented MP4, detected DRM-protected tracks, damaged index, no audio track, over 2 GB or indexes over 500,000 samples or 200,000 chunks across all tracks. Takes .mp4, .m4v and .mov.

How this tool works

An MP4, M4V or MOV file holds compressed media data and an index that lists the tracks and where each track's chunks sit in that data. This tool reads the index, deletes the audio tracks' entries, works out which byte ranges belong to the tracks that remain, and writes a new file with a rebuilt index and only those ranges. The audio is removed from the output bytes, and the readout shows how many bytes that was. “Audio tracks only” keeps subtitles, chapter tracks and other tracks; “Audio and every other non-video track” keeps only the video. Neither mode inspects metadata, so chapter metadata and other container or video-track metadata can stay in the file. The new file is assembled from ranges of the original, so a two-gigabyte video does not need a second two-gigabyte copy in memory, and the index is written before the media so playback can start before the whole file is read. After the page's code loads, the tool needs no decoder, encoder or worker library.

Another way to mute a video in a browser is to play it with the sound off and re-record the picture. That re-encodes the picture, which can lower its quality. It runs in real time, so a three-minute clip takes three minutes. And it writes WebM, a file type some phones, editors and messaging apps will not open. This tool rewrites the MP4, M4V and MOV container family without re-encoding, and it refuses other containers by name, with the reason on the row.

Codec, resolution, duration, channel count and sample rate are read from the file's index and reported as the file declares them. Nothing is decoded, so this page cannot verify that the video plays. Named refusals include non-ISO-base-media containers (Matroska, WebM, AVI, Windows Media, Ogg and Flash Video), fragmented MP4, detected DRM-protected tracks, a damaged or inconsistent index, files with no audio track and unsupported compact sample-size tables. The protection check visits each parsed track and sample entry for the encrypted entry codes encv, enca, encs, encf, encm and enct. It also looks for sinf protection boxes at fixed positions for standard video and version-0 audio entries. It does not inspect all protection layouts: extended audio entries and other track types do not get equivalent sinf checks. There is no duration cap. Files above 2 GB, or indexes above 500,000 samples or 200,000 chunks across all tracks, are refused with the measured count and the limit; kept and removed tracks both count. Files are not shortened to fit.

Frequently asked questions
Is my video uploaded anywhere?
No. The page reads the file and rewrites it in this tab, with no decoder or encoder library to download. The meter counts resource-timing entries from the first file read, so it does not include the page's own loading, and it is not a full record of network activity. Once the page has loaded, an offline conversion can check whether this operation runs locally.
Does this re-encode the video or reduce its quality?
No. The compressed picture samples are copied without being decoded or re-encoded, so they keep their codec, resolution and frame timing. The writer deletes the audio tracks' index entries and updates the offsets of the chunks it keeps. An MP4 stays an MP4 and a MOV stays a MOV. The time it takes depends on the file size and your device, since nothing is played back or recorded.
Why only MP4, M4V and MOV, and not MKV, WebM or AVI?
Because the reader and writer here handle the index structure that MP4, M4V and MOV share, and MKV, WebM and AVI index their media differently. Re-recording works for any video the browser can play, but it takes as long as the clip, can add another lossy video encode and turns the output into WebM, so this page refuses containers it cannot remux. For those, use a dedicated video tool that can remux them losslessly.
Is the audio really gone, or just hidden?
Gone from the new file. The bytes that belong to the removed audio tracks are not copied into it. Deleting only the index entry would leave the sound data behind, unreferenced, so this writer rebuilds the media data from the tracks it keeps. The result row shows how many audio bytes were dropped and the file size before and after. This covers the removed tracks' samples; the tool does not search the rest of the file for other hidden data.
Why did my file get refused?
Its row names the reason, and one refused file does not stop the rest of the batch. The usual causes are a different container with an .mp4 name, a fragmented MP4, a protected track, a damaged or incomplete index, a file with no audio to remove (no output is made), or a file over the 2 GB limit, which is refused and not shortened.

Related tools