Put text behind you
The classic depth effect: a big title that sits behind the person in your footage but in front of the scene — like the subject is standing in front of the type. Doing this by hand means rotoscoping; your agent does it with a matting model in a couple of minutes.
What you need
Section titled “What you need”- A clip (or photo) with a clear subject — you at your desk, a presenter, a product shot. Already in the project or just sitting on your disk.
- A connected agent.
What to say
Section titled “What to say”Segment me out of intro.mp4 and put the title “MAKE IT LOUD” behind me — big, bold, slightly above shoulder height.
Variations that use the same machinery:
Put me on a clean gradient background instead of my office.
Make the captions look like they pass behind my head.
What happens
Section titled “What happens”The agent runs Robust Video Matting locally over your clip to separate subject from background, builds a single muxed asset (color + mask side-by-side — one file, so nothing can drift), uploads it, and authors the sandwich: original background at the back, your title in the middle, the subject cutout on top. For photos it uses an image matting model instead — same result, same layering.
Make it yours
Section titled “Make it yours”Everything arrives as ordinary layers: the title is a normal text layer you can restyle, retime, and animate in the properties panel, and the cutout is a media layer like any other. Ask for a follow-up (“track the color grade”, “animate the title in on the beat”) or tweak it by hand — both work on the same document.