Photorealistic Portrait featuring Input Variable - Cinematic-lighting & Map via Nano Banana
Nano Banana1:1

Photorealistic Portrait featuring Input Variable - Cinematic-lighting & Map via Nano Banana

Photorealistic Portrait featuring Input Variable - Cinematic-lighting & Map—a Nano Banana prompt with map and portrait art direction.

#map#portrait#interior#miniature#weather

Best used for

Use for character exploration, portrait studies, costume variations, or visual-development boards.

Created by

Prompt Bookmarks @promptbookmarks Original source: x.com

The prompt

Copy it.
Then make it yours.

There is nothing special about these images other than the fact I am trying a crazy Nano Banana prompt to generate them 

Prompt: 
Input Variable: Give a single name
System Instruction:
Act as a Cinematic Set Designer. Create a photorealistic "Split-Diorama" capturing a moment of intense rivalry between the Input Subject and their Historical/Thematic Nemesis.
Phase 1: Dynamic Inference (The Casting):
The Protagonist (Input): Identify the input figure. Determine their iconic look (e.g., Wild hair, tweed suit) and their "Work Environment" (e.g., Cluttered Academic Office).
The Antagonist (The Rival): Identify their greatest intellectual or physical rival. (e.g., If Einstein→→
Niels Bohr or Oppenheimer; If Tesla→→
Edison). Determine their look and environment.
The Ephemera (The Clutter): Identify 3-4 specific items found in their workspace (e.g., Newspapers with specific headlines, Blueprints, Lab equipment, Instruments, Scattered notes).
 Conflict: The visual representation of their debate  .
Phase 2: The Narrative Topology (The Map):
Fill in the bracketed {VARIABLES} based on Phase 1.
codeText
<layout_engine>
|      [ BACKGROUND: {SHARED_ERA_SETTING} ]      |
|                                                |
|  ( {ATMOSPHERE_A} )        ( {ATMOSPHERE_B} )  |
|                                                |
| [ {FIGURE_A} ] <----Clash----> [ {FIGURE_B} ]  |
| ( {POSE_A} )      (Vision)      ( {POSE_B} )   |
|                                                |
| [ {CLUTTER_A} ]                [ {CLUTTER_B} ] |
| (Scatter: Papers/Tools)    (Scatter: Gear/Tech)|
|                                                |
|-----------------[ FLOOR ]----------------------|
</layout_engine>
Phase 3: Visual Translation:
{FIGURE_A} & {FIGURE_B}: Render Hyper-Realistic Human Figures. They must be dressed in era-accurate clothing. They are facing each other, engaged in debate or work.
{CLUTTER_A} & {CLUTTER_B} (Crucial): This is the "Mess."
Constraint: Do not leave the floor empty. Pile it with Contextual Debris.
Example: If Einstein, scatter Newspapers with headlines like "Theory of Relativity Proven!" and chalk-covered erasers. If Edison, scatter broken glass bulbs and patent documents.
<----Clash---->: A visual manifestation of their differing philosophies separating the room (e.g., A chalk cloud meeting steam; Order meeting Chaos).
{SHARED_ERA_SETTING}: The background environment that unites them (e.g., A University Hall, A Patent Office, A Stage).
Output: 8k Resolution, Cinematic Lighting, "Vanity Fair" Portrait Aesthetic, High Texture Density.

Detected placeholders

{VARIABLES}<layout_engine>[ BACKGROUND: {SHARED_ERA_SETTING} ]{ATMOSPHERE_A}{ATMOSPHERE_B}[ {FIGURE_A} ]<----Clash---->[ {FIGURE_B} ]

Refinement notes

Three useful ways to customize it

01

Subject and action

Replace the central subject first, then adjust its pose, action, expression, or relationship to the scene.

02

Composition and light

Change the viewpoint, crop, negative space, light direction, and contrast before layering on more style words.

03

Output shape

The saved reference uses 1:1. Change the aspect ratio to match the final placement before refining small details.