Voice and captions
Captions that land on the word being spoken.
When someone watches with the sound off, the captions are the video. Every spoken word gets its own start and end time, whether it comes from a narrator, a character talking on camera or your own take, so the captions follow the voice exactly and the word being spoken lights up.
Narration per scene
Listen to a voice before you pick it, and every scene is narrated in it. Regenerate one scene, change speed and volume, or upload your own recording.
Captions in bold styles
Ready styles with the spoken word highlighted, boxed or popping, then position, font, size and colours. One style for the whole video.
Captions for talking characters
When a character says a line on camera, the clip is listened to and its words are timed, so the captions follow the dialogue too.
Another voice for your take
On a video from your own take, give the whole voice track a new voice: the same message retold in fresh words in any voice, or your exact words and timing in another voice through your ElevenLabs account.
Fix a word, not the audio
Every word is a chip. Correct a brand name's spelling in the captions or hide a word, without regenerating the voice.
Music that steps back
One background track generated to the length of your video, or your own, turned down whenever someone speaks.
Burned in or SRT
Burn the captions into the MP4 on export, or leave them off, and get an SRT file either way.
Your ElevenLabs voices
Optional. Add your ElevenLabs key and every voice in your ElevenLabs account shows up in the voice picker, with previews.
Where captions come from
Narration is timed as it is made. A character talking on camera is transcribed from the clip and matched to the lines you wrote. A video made from your own take is captioned from its voice track. A film in the Cutting Room is captioned from the words of each finished video in it, so several videos share one caption style. The one thing that is not captioned is sound you add yourself without words to match, like a music track.
ElevenLabs is optional. The standard voices run through your AtlasCloud account. ElevenLabs bills you directly, per 1000 characters of narration, if you choose to connect it, and it is what keeps your exact words when you give your take another voice.
One script, every platform
Captions and voice carry over when you duplicate a video into another format.