Prompt overview
-
Produces a highly realistic, cinematic street‑style portrait captured from the side.
-
Keeps background people blurred and preserves a moody, fashion‑forward documentary vibe.
-
Uses short, literal clauses so free models parse identity and outfit details reliably.
Quick specs
-
Media: Image
-
Use case: Generation
-
Techniques: Shot grammar, Parameter block, Negative prompts
-
Models: Nano Banana, ChatGPT Images, Midjourney v6
-
Estimated time: 3–6 minutes
-
Skill level: Intermediate
Variables to fill
-
Aspect ratio: {ar} (3:2 or 4:5)
-
Seed/version (optional): {seed}
Prompt template
Paste one of the model‑specific blocks below without edits to the core text. Keep capitalization and punctuation simple.
Original prompt text (unchanged)
“A highly realistic cinematic portrait of tho person from tho upload-ed retorence photo. recreated with the user s own face and hair-style, The subject is weoring Stylish sunglasses identical to the ref-oronco, drossod in a dark shirt with an oversizod gray wool coat layered on top. He stands outdoors in a candid street-photography moment, captured from the side with dramatic lighting and shadows. Tho expression is sorious, confidont, and masculino.
The background shows blurred people and a photographer holding a camera. keeping the sarne street-style documentary atmosphere, Tho overall mood is moody, cinematic, and fashion-forward, pro- sarving the same pose, outfit, and vibe as the reference image”
Model‑specific blocks
-
Nano Banana (Google Gemini)
“[Paste the original prompt text above exactly as is.] Aspect ratio {ar}. Keep proportions and identity consistent with the uploaded reference. No text, no watermark, no extra limbs, no face changes.” -
ChatGPT Images (free)
“[Paste the original prompt text above exactly as is.] Shot on 50–85mm look, f/4–f/5.6, side angle. Aspect ratio {ar}. Do not alter face or skin tone. No text, no watermark, no background clutter beyond soft blur.” -
Midjourney v6
[Paste the original prompt text above exactly as is], 50–85mm lens look, f/4, side angle, soft crowd bokeh, neutral street backdrop, moody cinematic grade —ar {ar} —v 6 {seed}
Sample Output
Input Image:
Result:
How to use
-
Upload the person’s reference photo, then paste the model block for your chosen tool.
-
Keep the core text unchanged to match your request; only fill {ar} and {seed} if desired.
-
If hands or proportions warp, add a follow‑up instruction: “natural proportions; correct limb count; hands relaxed.”
-
For stronger separation, add: “subtle rim light from behind; background remains blurred.”
FAQ
-
Can wardrobe colors shift?
If the coat tone changes, add “oversized gray wool coat stays gray; no color shift.” -
The mood looks too bright.
Add “moody contrast; overcast feel; soften highlights.” -
Background becomes sharp.
Add “background stays blurred; people and camera only as soft shapes.”
Compliance and notes
-
Use only images you have rights and consent to use. Avoid altering identity when working with real persons.
-
Avoid including logos or private locations without permission.
Revision history
-
v1.0 – Website‑ready post using the exact user‑provided prompt text; added model blocks and usage tips – 2025‑10‑19