This specific prompt is designed to create a high-end, editorial fashion portrait while maintaining the exact likeness of a specific person. Because the prompt explicitly commands the AI to “use the subject in the image provided just exactly as he is,” standard text-to-image generation will not work. You need an “Image-to-Image” (img2img) or character referencing workflow.
Here is a step-by-step guide on how to execute this prompt for the best results:
Edit this image to show a men positioned in a close-up portrait shot, face tilted slightly upward at approximately 15-20 degrees with her chin gently lifted, creating a confident, aspirational angle. His head is centered in the frame with his gaze directed straight toward the camera from behind the sunglasses. He has short, raven-black hair, he's wearing nice vintage-inspiredround eye sunglasses with molten black-to-grey gradient lenses and matte black frames with delicate etched details, positioned perfectly on the bridge of his nose. His body is angled slightly (about 30 degrees) to create dimension, with shoulders relaxed and one shoulder subtly closer to the camera. he exudes magnetic confidence with a sultry pout. he's wearing a luxurious black blazer. The background is a rich, saturated golden-yellow that transitions to deeper amber tones at the edges. Dramatic directional lighting from above-left creates sculptural shadows along his neck and cheekbones, with warm backlighting creating a subtle halo effect. The composition is a Style/fashion portrait style with the face taking up roughly 60% of the frame. use the subject in the image provided just exactly as he is
Step 1: Select the Right AI Platform You must use an AI tool that supports strict facial consistency and image referencing.
- Midjourney: Best used with the Character Reference (
--cref) parameter. - Stable Diffusion: Ideal if you use FaceID or ControlNet extensions to lock in the facial structure.
- Leonardo AI: Works exceptionally well using the “Image Guidance” feature.
Step 2: Upload a High-Quality Reference Photo Choose a clear, high-resolution photo of the man you want to feature. Since the prompt dictates a 15-20 degree upward face tilt and a 30-degree body angle, uploading a base photo where the subject is already in a roughly similar pose will yield the most realistic, unwarped results.
Step 3: Correct the Pronoun Mix-up (Crucial Step) Before pasting, you need to clean up the text. The prompt currently mixes pronouns (e.g., “a men,” “her chin,” “His head”). AI engines take these words literally, which could result in a gender-mixed face or a confused output.
- Quick Fix: Change “a men” to “a man,” and “her chin” to “his chin.” Ensure all pronouns are masculine to keep the generation focused.
Step 4: Apply the Prompt and Parameters Paste your corrected text into the generator. If you are using Midjourney, your final input should look like this:
[Corrected Prompt Text] –cref [Link to your uploaded image] –cw 100
The --cw 100 (character weight) tells the AI to copy the face exactly from your reference image rather than inventing a new one.
Step 5: Refine the Atmosphere This prompt relies heavily on a saturated golden-yellow background and dramatic directional lighting. If the final image feels too flat, add negative prompts to guide the AI away from bad lighting. Use negative keywords like: “flat lighting, dull colors, white background, soft contrast.”