Help center7 articles
Video
Making clips from words, frames and references, working on a clip you already have, and finding B-roll.
- 8 minHow to make a videoOpen Video, pick a tab (Text / Image to Video, Reference to Video or Edit a video), pick a model, add the frames or references it takes, describe the shot, set Output, and press Generate. The price shows on Generate before you press it.
- 8 minStart and end frames, and reference clipsOn Text / Image to Video, a start frame begins the clip on your picture and an end frame makes it travel to a second. On Reference to Video, the model works from pictures, clips and sounds named @Image1, @Video1 and @Audio1, with an opening scene if you like.
- 4 minExtend, edit or upscale a clip you haveOn Video, open Edit a video, pick a model that works on a video and put one of your clips, or an upload, in its video slot. Carry it on from its end, change it the way your words say, or have a character move the way it moves. To make a clip sharper, upscale it in Edit & Enhance.
- 3 minCaption a videoOpen Captions and add a clip with speech. What it says is heard and set out as lines. Fix any word, pick a style, size, place and shadow, and press Apply captions. The price, by the minute, is shown before you press, and the captioned clip lands in My captions.
- 6 minWhich video model should I use?Seedance 2.0 is the default and takes frames, references and sound. Pick another for what your clip needs, such as a long take, a storyboard, a still to animate or a clip to work on. The model menu filters by what each takes and prices them all at your settings.
- 3 minFind B-roll for your editB-Roll finds free stock footage and sound effects two ways. Search stock costs nothing and looks up a subject on Pexels, Pixabay and Freesound. Match to audio reads a voiceover, song or video for the moments that want a cutaway, and finds footage and sounds for each.
- 6 minSound in your video: voices and lip-syncSome models make sound with the picture; on Seedance, a Music switch turns background music off. Add a recording as an audio ref and say in your prompt what it is for: a voice, a song or a sound effect. To lip-sync a face to your voiceover, use a talking model on a Creative Board.
Still stuck?
A person reads every message.
Usually within 4 hours on a working day.