Café Portrait Prompt Description

JSON-Bildprompt zur präzisen Steuerung von Pose, Kleidung und Details eines Café-Porträts.

Der Prompt

{
  "subject": {
    "description": "A young, attractive blonde woman with sleeked-back hair styled into a loose side braid, resting her right cheek on her hand and looking directly at the camera with a calm, natural, slightly pensive expression. Her facial features are balanced and aesthetically pleasing, with clear and smooth skin.",
    "position": "Seated at a wooden table in a cafe, facing the camera.",
    "pose": "Head resting gently on right hand, elbow on table; left arm relaxed on the table surface.",
    "expression": "Calm, natural, slightly pensive, soft gaze.",
    "clothing": {
      "top": "Black spaghetti strap tank top with a minimal, fitted look."
    },
    "accessories": "Multiple small gold hoop earrings, thin rings on fingers, minimal jewelry, a small script tattoo on the inner left forearm (text: 'no pain').",
    "hair": "Blonde hair, neatly slicked back and styled into a loose braid falling over the left shoulder with slight natural flyaways.",
    "skin_details": "Clear, smooth, healthy-looking skin with subtle natural texture, minimal blemishes, no heavy retouching"
  },
  "scene": {
    "description": "Interior of a modern cafe/bar during daytime. The bar counter and shelves are clearly visible, filled with liquor bottles and glassware, but the space feels clean and not overcrowded.",
    "location": "A modern cafe in Istanbul, Turkey.",
    "setting": "Indoor cafe with daylight.",
    "background_elements": "Bar shelves with bottles, glassware, wooden textures, large windows with daylight entering, very few or no visible people; if present, only soft blurred silhouettes without distinguishable features.",
    "lighting": "Soft natural daylight coming from windows combined with gentle indoor ambient light.",
    "atmosphere": "Relaxed, calm, modern urban setting, not overly busy."
  },
  "technical_details": {
    "shot_type": "Medium close-up.",
    "perspective": "Eye-level, natural handheld perspective as if taken by another person sitting at the table.",
    "focal_length": "Smartphone wide lens (~26mm equivalent).",
    "depth_of_field": "Shallow depth of field, subject sharply in focus, background softly blurred with natural bokeh.",
    "composition": "Subject slightly off-center, balanced composition with vertical lines from shelves and soft background structure.",
    "colors": "Neutral and natural tones, warm wood browns, soft gold from jewelry, realistic color balance.",
    "camera_type": "iPhone 13 rear camera",
    "camera_behavior": "Natural smartphone processing, slight edge sharpening, realistic HDR, no artificial filters",
    "resolution": "Standard mobile photo quality, not ultra sharp, slightly softened details",
    "image_characteristics": {
      "grain": "Very subtle fine digital grain",
      "dynamic_range": "Balanced HDR with controlled highlights and shadows",
      "sharpness": "Moderate, not overly crisp",
      "compression": "Minimal compression artifacts, close to original capture"
    }
  },
  "constraints": {
    "background_people": "Avoid clearly visible or detailed people; allow only indistinct blurred shapes",
    "focus_priority": "Face must be the sharpest element",
    "avoid": "Artificial faces in background, over-processed skin, Instagram-style filters, excessive sharpness, cinematic DSLR look"
  }
}

So nutzt du es

Den Prompt-Text kopieren (Button oben) und als System-Prompt bzw. erste Nachricht in ChatGPT, Claude oder einem lokalen Modell einfügen. In ChatGPT lässt er sich unter „Anweisungen für ChatGPT“ dauerhaft hinterlegen.

Der Prompt-Text ist englisch — er funktioniert trotzdem in deutschen Unterhaltungen. Für deutsche Antworten einfach am Ende ergänzen: „Antworte auf Deutsch.“

Im Detail

Ein JSON-strukturierter Bildprompt, der ein Café-Porträt bis ins Detail festlegt: Pose, Mimik, Frisur, Kleidung, Schmuck und sogar ein Tattoo mit Text. Der Vorteil gegenüber einem Fließtext-Prompt: Einzelne Felder lassen sich gezielt austauschen, ohne die gesamte Beschreibung neu zu schreiben – das macht Ergebnisse reproduzierbarer und Varianten leichter erzeugbar. Nützlich für alle, die eine wiederkehrende Figur konsistent in mehreren Bildern darstellen wollen. Zu bedenken: Solche fotorealistischen Personenprompts funktionieren nur mit Modellen, die strukturierte Prompts gut interpretieren, und sollten nicht für täuschende oder identitätsnahe Zwecke verwendet werden.

Praxis-Tipp

Ändern Sie gezielt einzelne JSON-Felder wie „clothing“ oder „accessories“, um Varianten derselben Figur zu erzeugen, statt den ganzen Prompt neu zu formulieren.

Lizenz & Quelle

• Lizenz: CC0-1.0 (Public Domain)
• Quelle: awesome-chatgpt-prompts (GitHub)

Lade …