Add Audio to Video: Replace the Soundtrack

How to add audio to a video

1. Choose the video above. The file card shows its length, so you know how much audio you need.
2. Under "Audio file", choose the sound: MP3, M4A, AAC, WAV, OGG, Opus or FLAC.
3. Leave "Start the audio at" at 0 to start the sound with the picture, or type a time such as 5 or 0:12.5 to start it later.
4. Press "Add the audio" and download the MP4. Its name ends in "-new-audio", and the result card says whether the sound was copied or encoded.

Replace, not mix

The audio file takes the place of the video's own sound. Whatever the video had before, whether speech, room noise or music, is not in the result. This is the right tool when:

• the original sound is unusable and you have a better recording made at the same time, for example from a separate microphone;
• the video was silent, such as a screen capture, a slideshow or an animation;
• you have recorded a voice-over that should be the only sound;
• you have already prepared a finished mix in an audio editor.

If you want music under the original speech, the two need mixing first. One way that uses the free tools here: extract the audio from the video, put that and the music on two tracks in an audio editor such as Audacity, lower the music until speech is clear, export one file, and add that file back with this tool. A video editor with a timeline does the same in one place.

Timing and length

The result always has the video's length. The audio is fitted to it like this:

Audio longer than the video

cut off where the video ends

Audio shorter than the video

silence after the audio ends, until the end of the video

"Start the audio at" set to 0

sound and picture start together

"Start the audio at" set to 5

five seconds of silence, then the audio from its own beginning

To start part-way into the audio file instead (skipping the intro of a song, say), cut the audio first with the audio cutter. To make the video shorter to match a short song, trim the video first.

What gets copied and what gets encoded

The picture is always copied into the new MP4 without re-encoding, so its quality doesn't change. The sound depends on the file:

• If the sound is in a format an MP4 can hold as it is, such as AAC from an M4A file, and it starts at 0, it is copied too, so the whole job takes seconds.
• Otherwise, including any file with a start offset, the sound is encoded to AAC at 192 kbps, which plays in practically every MP4 player.

The result card tells you which route was taken. If the sound needs encoding and your browser can't encode AAC, the tool says so; use an M4A or MP3 file, or try Chrome or Edge. Everything happens on your device; neither file is uploaded.

Examples

Music rights in plain words

Owning a copy of a song doesn't give you the right to publish it in a video. Platforms scan uploads for known recordings, and a match can mute the video, block it or send its earnings to the rights holder. Use music you made, music licensed for video use, or a royalty-free library, and keep the licence terms. The article on music licensing for translated videos covers what to check before publishing a version with new sound. If you only need the sound gone, mute the video instead.

Recording a voice-over

A phone voice memo app or a computer microphone is enough for narration. Test the level first with the microphone test, record in a quiet room, and leave a second of silence at the start so you can line things up with the offset.

A voice in another language

If what you need is the same video speaking another language, AI dubbing produces the translated speech in a copy of each speaker's own voice, in 19 languages. It costs 50 credits per minute (minimum 2 minutes), and for it only the audio track is uploaded and then deleted within 30 minutes of the job finishing. To offer that language on YouTube alongside the original rather than replacing it, see YouTube multi-language audio.