AI Video for Beginners, Part 2: Generate a Usable Shot

Posted on Nov 2 • Originally published at ay3s.com

Your photo is ready. Now let's make it move. If you don't have a starting image yet, complete Part 1. This attempt focuses only on camera movement.

Use the coffee photo from the first lesson as your starting frame. Keep the aspect ratio at 9:16 and choose a short duration supported by the tool, such as five seconds. Start with a low-cost preview setting; button names and available durations vary by platform.

If the tool crops out half the cup, fix the crop before generating. A prompt can't restore a composition that has already been cut off.

A coffee cup on a table by a window. The cup and objects on the table remain still. The camera slowly pushes forward with a small range of movement. Preserve the cup's shape and color, the background arrangement and steady lighting.

This spells out three things: the subject, the movement and what must stay consistent.

Leave out “epic cinematic masterpiece” for now. If the tool offers a negative-prompt field, try “distorted cup, flickering image, sudden cut.” If it doesn't, don't worry about it.

If the image looks beautiful but the handle grows into a strange shape, set that version aside. Editing rarely hides distortion in the main subject well.

If the camera moves too much, try “a very slight forward camera movement.” If the cup keeps warping, try “locked-off camera” to reduce complexity first.

Don't change the photo, style and action all at once. If the next version works, you won't know which change helped.

Set a small trial budget, perhaps three generations. If all three are unstable, simplify the photo rather than keep paying for retries. Save the usable clip and its prompt, then continue with Part 3: edit a finished short. The beginner's guide lists the full sequence.

This article was prepared for DEV with AI assistance.