← all guides

free guide workflows

Five Photo Codes for One Portrait

Make five distinct visual concepts from one portrait you own using full, copyable prompts.

Five Photo Codes for One Portrait

Use one full-body portrait to try five visually different scenes.

What this helps you do

Make five distinct visual concepts from one portrait you own using full, copyable prompts.

Before you start

Use a clear full-body image you own or have permission to edit. Remove private location information from the upload. Open an image-capable ChatGPT session. If you are following the Higgsfield workflow shown with this set, install its official ChatGPT plugin, sign in to Higgsfield, and check your available credits before generating.

Terms to know

Each slash label below is shorthand for a full prompt in this guide. It is not a hidden ChatGPT command.

Step-by-step

  1. Upload the same portrait for every experiment.
  2. Run /CLONEWALK to make a sequence of the person crossing one street.
  3. Run /LEVITATE and check that the shadow matches the lifted figure.
  4. Run /CITYGIANT, /COLORRAIN, and /AURORA as separate concepts.
  5. Compare each face, outfit, and important detail with the original before sharing.

Copy this

Paste this in the tool named above. Replace each bracketed detail with your own information.

Copy this

/CLONEWALK: Use my uploaded full-body portrait. Show five versions of the same person walking across one city street, like five moments from a motion sequence. Keep the face and outfit consistent.

/LEVITATE: Lift the person a few feet above the sidewalk. Add a believable shadow below and keep the face unchanged.

/CITYGIANT: Make the person appear taller than a city block while preserving realistic perspective and recognizable clothing.

/COLORRAIN: Put the person under vivid colored rain at night. Keep the face clear and the person separate from the background.

/AURORA: Put the person beneath a dramatic aurora. Keep the portrait recognizable and make the lighting match the sky.

Things to know

Generated images can alter faces, logos, text, or product details. These are creative scenes, not documentary images. If using Higgsfield through a connector, check its current access and credit terms before generating.

If something goes wrong

If the face drifts, shorten the scene description and restate the traits to keep. If the crop cuts off the body, specify full-body framing and enough room around the subject.

Final check

A pass is five clearly different scenes with a recognizable subject in each one.