# Zombie Couple Video Prompts: Roles, Actions and Boundaries

Write an original zombie couple video prompt with distinct character roles, one observable action and a controlled camera. Adapt it to your tool’s actual inputs.

Use this guide to plan your scene, then open the studio. Two-character creation supports two separate portraits at 720p, 4–15 seconds.

## Begin with the cast

Write who each reference represents before describing the world. Include a simple visual marker and screen position for both characters. Avoid relying only on partner, girlfriend or boyfriend when the tool needs to associate two distinct images. The prompt is a casting instruction, not a substitute for a mode that supports two characters.

## Describe visible movement

Replace loves them despite everything with looks up, recognizes the scarf, lowers their shoulders and reaches out. Keep those actions in the order you want to see. Ask one person to initiate and the other to respond. A tool may compress or omit steps, so review the result before treating the full sequence as successful.

## Bound the appearance

Use precise theatrical details: pale makeup, subtle dark circles and worn clothing. If you want a non-graphic scene, explicitly avoid visible injury and blood. Keep that boundary consistent in the reference preparation and all shots. Adding conflicting horror details later can change the focus of the entire scene.

## Specify framing separately

State medium-wide fixed shot or a small slow push, then stop. A long list of camera movements competes with the embrace. Aspect ratio belongs in the tool’s format settings when available; writing vertical in the prompt does not necessarily override those settings. Keep important faces and hands away from the edges.

Reference A is the adult survivor in a dark coat on the left. Reference B is the adult partner on the right, with subtle theatrical zombie makeup. In a quiet rainy street, A recognizes B and opens their arms. B pauses and takes one slow step closer. Fixed medium-wide camera; both faces stay visible and distinct. No shooting, injuries, blood, extra people, text or logos.

## Change one variable per draft

Keep a small record of the photo pair, roles, selected mode and prompt version. Compare a fixed shot against a slow push before changing the setting as well. If the faces fail, rewrite the casting or references; if the hands fail, simplify the interaction. This makes revision decisions understandable without pretending that retries guarantee a better clip.

## Use the prompt in a compatible tool

The copyable text here is original editorial direction. A preset-only studio may not accept arbitrary text, and a start/end-frame mode may not interpret reference labels. Confirm the actual fields before pasting. Review the complete output for identity, anatomy and the stated scene boundaries; keep unsuccessful versions out of your published examples.

## Questions before your first scene

### Can I paste this into the studio?

Only if it offers a custom prompt field. The inherited studio uses preset directions.

### Should I write a long negative prompt?

Use the fields the selected tool supports. Keep essential scene boundaries clear and relevant.

### Can a prompt fix unsupported inputs?

No. Confirm the mode supports your references before relying on the text.

### Is this a tested recipe?

It is an original planning example, not a guarantee or a measured generation result.