Riesige städtische Werbetafel mit Identitätssperre

This is a highly specific and complex image prompt. Here is a breakdown of the prompt, followed by the refined prompt itself, structured for maximum effectiveness with AI image generators like Midjourney, DALL-E 3, or Stable Diffusion. ### Prompt Analysis & Refinement Strategy The original prompt is excellent but contains some conflicting or redundant elements. My refinement strategy focuses on: 1. **Clarifying the Subject:** The core subject is the *giant person* and their interaction with the *car*. The prompt must prioritize this. 2. **Resolving the Outfit Conflict:** The prompt specifies "black suit" but then lists "different outfit, changed clothes, gown, dress, suit" in the negative. I will make the "black suit" a strong positive element and keep the negative prompt focused on *other* types of clothing. 3. **Strengthening the "Forced Perspective" Concept:** The illusion is key. I'll use terms like "forced perspective," "scale contrast," and "interaction" to guide the AI. 4. **Optimizing for Vertical Composition:** I'll explicitly state the aspect ratio and composition at the beginning. 5. **Consolidating Redundancies:** Terms like "photorealistic," "cinematic," and "8K" are repeated. I'll consolidate them into a powerful, concise block. --- ### Refined Prompt (Positive) **For Midjourney / DALL-E 3 / Stable Diffusion:** ``` Vertical 9:16 composition. A photorealistic, cinematic, surreal scene of a colossal, giant man in a perfectly tailored black suit, kneeling in a bustling city street. He is using his index finger to gently touch the roof of a glossy, black luxury supercar parked below him. The forced perspective emphasizes the immense scale contrast between the giant's detailed, realistic hand and the small, highly detailed car. The giant's face is sharp, focused, and realistic, looking down at the car. The scene is set in a hyper-detailed, futuristic cityscape with towering skyscrapers and glass buildings reflecting natural daylight. A crowd of tiny, realistic people watches from a safe distance. The lighting is dramatic and realistic, with natural shadows and glossy reflections on the car and building windows. The image is an ultra-high-end, premium advertisement style, with sharp focus, perfect depth of field, and seamless integration between the giant, the car, and the environment. 8K, ultra-detailed, hyper-realistic, architectural photography, surreal realism. ``` --- ### Refined Negative Prompt ``` cartoon, anime, CGI, 3D render, illustration, painting, fake face, changed identity, face distortion, unrealistic body proportions, deformed hands, extra fingers, extra limbs, blurry face, plastic skin, low quality, unrealistic lighting, bad perspective, distorted architecture, duplicate people, warped car, text errors, watermark, logo, different outfit, changed clothes, gown, dress, casual clothes, jacket, t-shirt, jeans, multiple cars, other vehicles, daytime, night time, overexposed, underexposed, cluttered composition. ``` --- ### Tips for Best Results with AI Generators * **Midjourney:** Use the `--ar 9:16` parameter (e.g., `--ar 9:16 --v 6.0`). The prompt is long, so it will work well. You may need to run it a few times to get the perfect "touching" interaction. * **DALL-E 3:** This prompt is ideal for DALL-E 3 as it excels at following complex instructions and understanding scale. It will handle the "giant hand touching car" concept very well. * **Stable Diffusion:** If using SD, you might want to use a model like `Juggernaut XL` or `Realistic Vision` for the best photorealistic results. You can also use ControlNet with a depth map or OpenPose to better control the composition and the giant's pose. **Key to Success:** The phrase "forced perspective" and "seamless integration" are your best friends here. They tell the AI that the scale difference is intentional and must look physically plausible.