Remove, silence or re-level the soundtrack without re-encoding a single frame. The picture is copied bit for bit, and the output keeps your container with an audio codec it can hold. No uploads, everything runs in your browser.
Initializing in your browser…
Cut a clip out of a video, choosing whether the cut lands on the exact frame (re-encodes to MP4) or leaves the streams untouched (instant and lossless, but only at a keyframe). The result says which ran and reads the real length back off the file.
Convert a video between MP4, WebM, OGG, MOV, AVI and MKV, each naming the codecs it really writes and each producing a file that actually plays. One quality dial normalised into four different encoder scales.
Join clips end to end, saying in advance whether they can be stitched together untouched and, when they cannot, exactly which differences stop it. Clips of different sizes are letterboxed into one frame rather than squashed.
A clip has background chatter you do not want; you need it silent for a captioned post.
Input
clip.mp4, 41,791 bytes · remove the track
Output
clip.mp4, 8,570 bytes, every frame bit for bit the source, no audio stream at all
The video is copied with -c:v copy, so the picture is untouched and the file shrinks by exactly what the soundtrack occupied. Keeping a silent track instead gives 10,940 bytes, which is the choice when something downstream expects an audio stream to exist. A WebM keeps its own container and gets Opus rather than AAC, which is what used to produce a 264 byte file nothing could open.
Remove, silence or re-level the soundtrack of a video without re-encoding a single frame. The picture is copied through bit for bit in every mode, so the file is visually identical to the original, and the output keeps your container with an audio codec that container can legally hold.
Strip audio from clips that contain copyrighted songs before uploading to platforms with content ID systems.
Remove audio from recordings that captured private conversations or sensitive background audio before sharing.
Mute the original audio before importing into an editor where you plan to add a voiceover, music track, or sound effects.
Raise the level of a video recorded too far from its subject, with the tool warning you at the point where boosting starts to clip.
Pick 'Remove the track'. ffmpeg runs -c:v copy -an -threads 4, the audio stream is dropped, and the download is named <name>_no-audio with the source container kept.
Pick 'Keep a silent track'. ffmpeg runs -c:v copy -af volume=0 with the encoder the container accepts, so an MP4 gets silent AAC and a WebM gets silent Opus.
Pick 'Change the level' and set 250 percent. ffmpeg runs -af volume=2.5; the page warns first that anything already near full scale will clip.
Every mode passes the video with -c:v copy, so the picture stream is copied rather than re-encoded and your frames are bit for bit the source. Nothing about the resolution, the codec or the quality changes in any mode, which is why even a long clip finishes in a moment and why there is no generational loss to worry about. Only the soundtrack is ever touched, and the tool states which of three things it did to it.
Removing the track runs -c:v copy -an, which drops the audio stream from the container entirely; the file shrinks by exactly what the soundtrack occupied. Measured on a two second test clip, that took 41,791 bytes down to 8,570. Keeping a silent track re-encodes the audio at zero gain instead, for players and editors that expect a stream to be present: the same clip came out at 10,940 bytes, so silence is not free but it is close. Changing the level multiplies the waveform by the figure you set, and the slider runs to 400 percent rather than stopping at 100, because volume= is a gain and not a ceiling: boosting a quiet recording is the commoner request. Above 100 percent anything already near full scale clips, and the page says so before you run it rather than afterwards.
The container is the part that used to be wrong. The output keeps your source container, because the video stream is being copied and has to stay somewhere that can hold it, and the AUDIO codec is chosen to suit that container: AAC for MP4 and MOV, Opus for WebM, Vorbis for OGG, MP3 for AVI. Writing AAC into a WebM, which is what happened before, is not something Matroska will do: ffmpeg refuses to write the header, exits with an error, and leaves a 264 byte stub behind. That stub used to be handed over as a successful mute, because the exit code was never checked. It is checked now, and a WebM in gives a WebM out with its VP9 video copied untouched and its audio in Opus.
The tool reads the file before it starts and tells you what is in it: the container, the video codec, the audio codec and the size. A file with no audio stream at all is refused with that reason rather than run and failed. Afterwards the result panel prints the container that came out, the audio encoder used, the real size and the exact command that ran, and the finished file plays in the page beside the original.
No, in any mode. The video stream is copied with -c:v copy and never decoded or re-encoded, so the frames in the output are the frames in the input.
Removing it deletes the audio stream from the container, so the file shrinks by whatever the soundtrack occupied and a player finds no audio at all. Silencing it keeps a stream that is present and quiet, which some editors and upload pipelines expect. On a two second test clip, removing gave 8,570 bytes and silencing gave 10,940.
Yes. The level runs to 400 percent, because the underlying volume filter is a gain rather than a ceiling. Above 100 percent anything already close to full scale clips, and clipping cannot be undone afterwards, so the page warns before you run it.
It stays a WebM, with its VP8 or VP9 video copied through and its audio written as Opus. WebM cannot hold an AAC track at all: ffmpeg refuses to write the header and leaves a 264 byte file behind, which is exactly what this tool used to produce for every WebM it was given.
Video is decoded and re-encoded on your own device. Large files never leave your machine.