Prompt del canal

Toma a8

Copia el prompt, pégalo en ChatGPT y adjunta tus fotos. Hará esta misma toma contigo.

Prompt
Use uploaded photos as the ONLY identity reference. Generate the exact same woman from the reference photos. Preserve her real face, facial proportions, natural eye shape, eyelids, iris color, pupil size, gaze, skin tone, expression style, hair color, exact hair length, hair texture, body proportions and figure. Do not invent a new face. Do not beautify, retouch, smooth skin, reshape the face, slim the body or make her look more perfect.

Eyes are critical: copy the eyes from the reference photos only when visible. Keep the same eye shape, eyelids, iris color, pupil size and natural gaze. Eyes must look normal and human in a real iPhone flash photo, with natural shadows and small realistic catchlights. Do not make the eyes brighter, larger, sharper, glassy, stylized or doll-like.

Strict format: vertical 4:5 Instagram post crop, high-quality clean iPhone photo, sharp subject, real phone detail, no low quality, no blurry face, no compression artifacts.

Create a realistic iPhone flash photo collage / photobooth-style strip made of three stacked candid shots of the same woman in the same location. The mood is playful, edgy, urban and real, like a friend took several close flash photos at night and arranged them vertically.

Format and layout: final image is a vertical 4:5 Instagram post. Inside it, show three horizontal photo frames stacked vertically, each with slightly different expression and pose. The same woman appears in all three frames with the same outfit, same location and same lighting.

Scene: dark urban wall at night, black brick wall, metal window bars or railing behind her, street posters on the wall, slightly gritty nightlife atmosphere. Posters may have graphic shapes but no readable text or logos. Background should feel real and imperfect, not staged.

Outfit: dark oversized top or sweatshirt layered over a gray long-sleeve shirt, casual edgy streetwear. Small cross necklace or simple jewelry. Thin rectangular glasses with a delicate metal frame. Rings and natural manicure visible while holding the drink.

Hair: voluminous wavy hair with bangs / face-framing fringe, slightly messy and natural. If the reference photos do not have bangs, keep the natural hairline and adapt the front hair softly without changing identity. Preserve exact hair color, hair texture and realistic hair volume.

Action and expressions: she holds a clear plastic cup with a bright green drink and straw. 
Frame 1: she holds the drink near her shoulder and looks at the camera with a playful pout.
Frame 2: she smiles or laughs while sipping from the straw.
Frame 3: close-up, she drinks through the straw with a slightly pouty playful expression, looking into the camera.
The expressions should feel alive, funny and spontaneous, not posed.

Lighting: direct iPhone flash at night, realistic flash highlights on face, glasses, drink and hair. Dark background, slight flash shadow, true-to-life colors, subtle phone noise, no studio lighting, no cinematic grading.

Realism: natural skin texture, pores, small imperfections, natural under-eye texture, realistic eyes from reference photos, realistic hands, realistic glasses reflections, realistic straw and drink, minimal processing.

Add human imperfections: slight misalignment between the three frames, imperfect flash framing, tiny lens softness, street wall texture, casual nightlife clutter, not too polished.

Avoid: beauty filter, skin smoothing, face changes, body changes, changed identity, changed eyes, artificial perfect skin, distorted hands, extra fingers, professional camera look, studio lighting, CGI, illustration, readable text, watermark, logos.

El prompt está en inglés: así el modelo lo entiende con más precisión. No hace falta traducirlo.

Qué hacer con él

  1. 1

    Pulsa «Copiar»: el prompt entero pasa al portapapeles.

  2. 2

    Abre ChatGPT y pégalo en un chat nuevo.

  3. 3

    Adjunta dos o tres fotos tuyas y envía. La toma llega en un minuto.

Más prompts

En la biblioteca hay 494 fotos más

Treinta prompts nuevos cada semana. Cada uno con su ejemplo listo.

Del mismo set