Frames
First and last frame to change a spicy H3 position
Going from kneeling-behind to rider in one H3 prompt is how you get a third body. The model is being asked to invent the in-between. First and last frame is the tool that exists for that in-between: you supply both stills, you describe one motion, H3 fills the path. The scenario page is first & last frame video.
When to use it
- A position change (from-behind → rider, wall → kneel, edge → face-to-face).
- A wardrobe state change you already have stills for (shirt on → shirt over the head).
- A stand-up or sit-down that would smear in a text-only take.
When not to: a locked-camera look, a single push-in, afterglow. Those are one-motion takes from a single still or from text. Mixing a last frame into a shot that does not need a destination just fights the start frame.
How to shoot the stills
Same adult, same marks, same room, same ratio. Change pose, not identity. The end frame is the first frame of the next block on spicy positions. If the two stills disagree about the scar, the path in the middle will invent skin. Pull both stills from the body bible plus one blocking still, not from two unrelated generations.
H3 image-to-video takes an optional last frame; Kling, Vidu and Wan have dedicated start-end modes listed on that tool page. Crop both stills to the target ratio before you upload. Adults only; undressing photos of real people is not offered.