Veo prompt for the Ramalama trend: copy it and go

Labelled block for Veo

Subject: the person in the reference photo, with face, hair and skin tone unchanged, dressed as a classic witch in a pointed black hat, a black dress with a ragged hem, black-and-purple striped stockings and pointed black ankle boots.
Setting: a quiet residential road after dark, black asphalt, warm orange lamps, trees and hedges.
Camera: vertical 9:16, 8 seconds, one unbroken shot. A phone held just off the ground glides backwards in front of the walker.
Action: from 0 to 5 s only boots, stockings and the dress hem are in frame, a bold runway stride landing on every beat, face out of shot. From 5 to 5.6 s a fast tilt up from the shoes to the face reveals the whole costume. From 5.6 to 8 s the witch cups the face with both hands, smiles, leans into the lens and laughs.
Audio: instrumental electro-pop with a stomping kick on each step, faint night air. No dialogue, no singing, no lyrics.
Look: photoreal, with no on-screen text, captions, logos or watermark.

What is different in Veo

• Keep the Audio label. Veo adds its own sound, and without a ban it can bring in speech, a song or a crowd.
• Eight seconds is the usual ceiling, so the beats keep their order and only their times are squeezed.
• Add your photo as a reference image if your Veo option has one, and keep each label to a single idea.