What the tech

Search

/ opens search from anywhere. Esc closes it.

Tools

Get the newsletter
Example result of this prompt The made-up input photo InputExample result

Example Example made on our own Mac with open models: Z-Image made the made-up input photo, FLUX.2 klein ran this prompt on it. Your result in the AI above will look different. How we make examples

Same portrait in four lighting setups

One 2x2 grid of your portrait in four classic lighting styles, so you can pick a favorite. You need one clear photo.

You need
1 photo of yours
Written for
Gemini, ChatGPT
Result
Image

Prompt

Using my attached photo, create one image with a 2x2 grid of the same studio portrait of me, shown from the shoulders up against a plain dark grey backdrop. Keep my pose, expression, outfit and framing identical in all four squares; only the lighting changes. Top left: soft, even beauty light from the front. Top right: split light, one side of my face lit and the other in shadow. Bottom left: warm side light with a small triangle of light on my shadowed cheek. Bottom right: fourth. Add a thin white line between the squares and labels. Keep my face exactly as it is in the photo: same identity, features, skin tone, age, hairstyle and body proportions. My face must look the same in all four.

Fill in, or keep the brackets and change them later

fourth
labels

How to use it

  1. Open Gemini or ChatGPT and choose the image model.
  2. Attach a front-facing photo and paste the prompt.
  3. Like one square? Ask for it alone, full size, in a new message.
What to expectKeeping one face identical across four panels is hard, so small differences between squares are likely.

Tips

  • Useful before a real shoot to show the photographer the look you want.
  • If the faces differ between squares, ask for two at a time instead of four.
  • Your photo is uploaded to Google or OpenAI. Read their data policy before you share a face.

Editor: Ali Dezhami · Updated November 18, 2026