If I’m still deciding what a scene should look like, I’d start with a text prompt. If I already have a composition I want to keep, I’d start with a still image.
Take a simple product clip: a ceramic mug on a kitchen counter. If the mug and framing already look right, the next prompt can focus on motion: “Keep the camera still. Let steam rise slowly from the mug.” Asking for a new angle, different lighting, and moving steam in the same pass makes it harder to tell what caused an unwanted change.
The first thing I’d check in the clip is whether the handle keeps its shape. A nice camera move doesn’t help if the product itself changes halfway through.
OpenMaker supports both text-to-video and image-to-video, so you can try both starting points in the browser.
When you animate a still image, what tends to drift first for you: the subject, the framing, or the lighting?
Top comments (0)