Replace Video Audio Online

1. 1

   Add your video

   Drop in an MP4, MOV, WebM or MKV file. It is read on your device and nothing is uploaded.

2. 2

   Add the new audio

   Choose an MP3, WAV, M4A, AAC, OGG or FLAC file. One soundtrack covers every clip you queue.

3. 3

   Replace or mix

   Replace drops the original track entirely. Mix keeps it underneath, with its own level, so music sits behind dialogue.

4. 4

   Fit it to the picture

   Loop the audio to fill the video, play it once, or cut the video where the audio ends. Add a fade, nudge the sync, or trim to the part you want.

5. 5

   Listen, then export

   Press play to hear the actual mix against the picture. When it sounds right, export and download.

Swap narration, lay in music, or fix bad room tone without re-exporting an entire edit. The picture you shot is the picture you keep — this tool rewrites the soundtrack around it and leaves every frame exactly as it was captured, so a quick social cut, a language swap or a mastered mix costs you nothing in image quality.

Benefits

• New soundtrack: Lay in music, a voiceover, or a cleaned-up mix in one pass.
• Picture untouched: The video is copied frame for frame, never re-encoded, so nothing softens and nothing blocks up.
• Mix, don't just replace: Keep the original underneath at a level you choose, so music sits behind dialogue instead of erasing it.
• Hear it first: Press play and listen to the real balance before you commit to an export.
• No watermark: The file you download is the file you made — nothing is stamped on it.
• Nothing is uploaded: Both files are read on your device, so a private recording stays private.

Most tools only replace. That is the right answer when the camera audio is unusable, and the wrong one whenever the clip has dialogue you want to keep. This tool offers both, and the difference is one control.

Which mode to use

• Replace: The video's own track is dropped and the new file becomes the only audio. Use it for a voiceover, a dub, or a clip whose original sound is noise.
• Mix: The original stays, and the new file plays over it at whatever level you set. Use it for background music under dialogue — 25% to 40% is the usual starting point.
• Both levels are independent: The new track and the original each have their own volume, from silent to 200%, so you can lift a quiet interview as easily as you can duck a loud track.
• Mix needs an original to mix with: If the clip has no audio track, or its track uses a format this device cannot decode, the option is shown as unavailable rather than silently doing nothing.

The two files almost never have the same length, and what should happen when they differ is a decision, not a default. There are three, and each one produces a different file.

The three fit modes

• Loop to fill: The audio repeats until the video ends. A 30-second music bed covers a three-minute clip without a gap. The output is the video's length.
• Play once: The audio plays through and the rest of the video is silent. This is what you want for a voiceover that ends before the picture does. The output is the video's length.
• Cut video to audio: The video ends where the audio ends. Useful when the soundtrack defines the edit — a song determines the cut, not the footage. The output is the audio's length.

A soundtrack that starts on the wrong frame is worse than no soundtrack at all, so the three controls that place the audio in time sit together in the rail and every one of them is audible in the preview as you move it.

The placement controls

• Offset: Nudge the new audio later or earlier against the picture. Positive delays it, negative advances it. The range tracks the two files, so you cannot push the whole track past the end of the clip.
• Trim: Choose the part of the audio file you actually want — the chorus rather than the intro — without cutting the file first in a separate tool.
• Fade in and fade out: Up to ten seconds each on the free plan. A fade-out is what stops a looping music bed ending mid-bar when the video does.
• The ranges are the real limits: Each slider's maximum is computed from the clip and the track you loaded, so a value that would produce silence or an impossible fade cannot be selected.

You provide a video and a replacement audio file. The audio is decoded, placed against the picture according to the fit, level, fade and offset you chose, and written into a new file next to the video's original frames — which are copied across untouched. One soundtrack can cover a whole queue of clips.

Workflow (overview)

• Video: Load the clip, or several with Pro.
• Audio: Load one soundtrack for the whole queue.
• Balance: Replace or mix, and set the levels.
• Placement: Fit, offset, trim, fades.
• Listen: Play the real mix against the picture.
• Export: The frames are copied, the new audio is written, you download.

Worth knowing before you export.

Key points

• The picture is copied, not re-encoded, so the exported video is visually identical to your source — there is no generation loss at all.
• The audio IS encoded, because a new track has to be written; that is the one part of the file that changes quality.
• The original audio is not recoverable from the exported file. Keep your source video if you might need the old mix.
• Output containers are offered only when they can carry your video as-is. A container that would force a re-encode is shown as unavailable, with the reason.
• The free plan handles videos up to 200 MB and 10 minutes, with soundtracks up to 50 MB. Pro is limited by your device instead.
• Free exports audio at 44.1 kHz and 192 kbps; Pro exports at 48 kHz and 256 kbps.

Cleaner results, with less friction.

Audio and quality

• Use a mastered file at consistent loudness so the swap does not introduce a jump.
• For music under dialogue, start the bed at about 30% and adjust while listening to the preview rather than after exporting.
• Add a one-to-two-second fade-out whenever you loop a music bed — it is what stops the track ending mid-phrase.
• Prefer a lossless or high-bitrate source; the export encodes once, and a poor source cannot be recovered afterwards.
• Keep an unmodified copy of the video for future changes.
• If lip-sync matters, time the audio to the picture in your editor first, then use the offset here for the final few frames.

Common use cases

• Quick fixes: Replace noisy camera audio with a studio take.
• Music: Drop a licensed track under a vertical cut, looped to length.
• Localization: Swap a voiceover for another language.
• Podcast video: Mix a cleaned-up microphone track over the camera's own sound.
• Social: Refresh the audio on a repost without re-rendering effects.
• Batch: Put one music bed across a series of clips and download them as a single archive.

The video's encoded frames are read from the source container and written straight into the new one, so no picture data is ever decoded or re-compressed. Meanwhile the replacement audio is decoded to samples, placed and shaped according to your settings, and encoded once into a track the chosen container can carry. The two are written together and the file is finalised.

Process

• Read: Both files are opened on your device; nothing is sent anywhere.
• Decode audio: The new track — and the original too, if you are mixing — becomes samples.
• Place and shape: Trim, offset, loop, level and fades are applied to those samples.
• Copy the picture: The video's frames move across as they are, in one pass.
• Write: The new soundtrack is encoded alongside them and the file is finalised.
• Verify: The result is read back to confirm it holds a real video track, a real audio track and the expected length before it is offered to you.

Hardware-accelerated, and the picture is copied rather than re-encoded — a ten-second clip finishes in well under a second.

Replacing a soundtrack is usually one step in a longer job. These handle the rest of it, and every one of them runs on your device too.

• Mute video
• Extract audio from video
• Video cutter
• Video compressor
• Video resizer
• Video cropper