Split Face Motivational Typography Poster Prompt
Prompt
Based on the reference identity, construct a hyper-real black-and-white editorial poster divided by an exact vertical line through the center. On the right, show precisely half of a forward-facing portrait cut at the central axis. Give the subject a composed confident gaze into the camera and a matte-black crew-neck T-shirt. Resolve pores, separate beard hairs, naturally arranged textured hair, iris patterns, and catchlights at macro clarity. Use soft frontal light revealing epidermal detail without severe shadow. Leave the left half pure white. Stack heavy modern sans-serif lettering exactly as follows: 'IF\nYOU\nTHINK\nYOU CAN\nDO IT\nYOU\nCAN'. Treat every letter as a masking window containing faded, anatomically continuous grayscale pieces of the missing left face; skin and beard must align correctly across the dividing line. Deliver a flawless 8K editorial composition in vertical 4:5 with absolute 50/50 balance, crisp typography, and precise integration.
How to use this prompt
Paste the complete prompt into Gemini. Add any required reference image, generate the first version without rewriting the source prompt, and then request one precise revision at a time.
Best AI model
Gemini is the model identified by the original source page for this prompt.
Tips for better results
• Attach a clear, well-lit reference image and state which identity details must remain unchanged.
• Preserve the specified lighting direction and contrast before changing wardrobe, pose, or background.
• Check generated typography carefully and regenerate only the text area if lettering is inaccurate.
• Generate the complete prompt first, then refine one visual detail per revision.
• Keep the core subject and composition consistent while comparing variations.
Frequently Asked Questions
What does “Belief in Focus” create?
It creates the subject, composition, and visual treatment described in the original prompt.
Which model should I use?
The source identifies Gemini; follow the prompt's own model and reference-image requirements.
Can I customize the result?
Yes. Generate the original prompt first, then change one detail at a time while preserving the core composition.

