Graphics behind a person.
Combine the person matte with layered graphics for depth.
Try it in your agent.
Use this prompt as a starting point. Review the result, then ask for the details you want to change.
A prompt to start from
Use Tesseract. Create an eight-second silent landscape Tesseract depth-title example using the included coastal walking footage. Keep the continuous original video as the background. Fade and lift the large Bone Exposure -30 word “Outside.” into the scene from 0.5 to 1.9 seconds, behind the woman’s head and shoulders. Use a full-width 382 px title so both ends remain visible around the subject. Implement the foreground with a duplicate video and a person-matte luma source, keeping their source timing and transforms identical; use the packaged RVM model. Do not draw a manual silhouette or use externally generated segmentation. Retain small Denim INK labels with 96 px safe margins. Required inputs: the coastal MP4, packaged fonts and the local segmentation model listed in media-manifest.json. Keep the title, footage and matte separately editable. Inspect motion and fine hair edges, and refine the title size and placement after rendering.
Tesseract plugins and runtime downloads are coming soon.
A few useful details.
How do the graphics stay behind the person?
A person matte separates the subject from the background. Put the graphic between those layers and review the matte edges in the actual footage.
How do I try this with my agent?
Tesseract plugins and runtime downloads are coming soon. Save the sample prompt for launch. Once Tesseract is installed, your local agent can use it to create, preview, and refine an editable project.
Where does the video render?
Tesseract previews and renders locally on a Mac. The local engine is free; your agent’s subscription or usage costs are separate.
Try these next.
Your next video, made with your agent.
Professional tools for editing footage, creating motion graphics, and shaping sound.
Free on your Mac. Your agent’s costs are separate.