🎬 Делаем из одной фотографии настоящий сториборд
🎬 Делаем из одной фотографии настоящий сториборд Хочется, чтобы персонаж выглядел как герой фильма? Тогда вместо одной картинки сделайте сразу серию кадров с разных ракурсов. Просто загружаете фото в GPT Image, вставляете промпт - и получаете набор кадров: общий план, крупный план, вид со спины, сверху, сбоку и другие ракурсы. После этого их можно объединить в полноценную анимацию или использовать как основу для генерации видео. Промпт 👇 Create one final image as a clean 3×3 storyboard grid (9 equal panels), using the uploaded reference image as the one and only visual source. Strict continuity: preserve identical subject(s), face identity, body proportions, hairstyle, clothing, accessories, age, skin tone, expressions, location, architecture, environment, weather, lighting direction, color palette, key objects and spatial layout across all panels. Never add, remove or replace characters, props, locations or scene elements. The storyboard must depict only a realistic continuation of the original moment. Story arc: setup → subtle change → rising tension → escalation → turning point → climax → immediate aftermath → de-escalation → logical resolution. Every generation must create a completely different story while remaining fully consistent with the original scene. Camera language: P1 extreme wide establishing shot, eye level, 24mm. P2 medium shot, 3/4 angle, 50mm. P3 tight close-up, eye focus, 85mm. P4 over-the-shoulder, 70mm. P5 low-angle hero shot, 35mm. P6 dutch-angle medium close-up, 50mm. P7 first-person POV. P8 top-down bird's-eye view, 35mm. P9 cinematic wide symmetrical ending shot, 28mm. Every panel must have a unique framing, camera height, distance, angle, focal length and composition. Include variations such as profile view, rear view, foreground framing, negative space, leading lines, centered composition, off-center composition, layered depth, foreground occlusion, rack focus, shallow DOF, deep DOF, dynamic perspective. No repeated camera setups. Ultra photorealistic only. Looks like real DSLR/cinema camera footage. Natural skin texture, visible pores, fine wrinkles, realistic eyes, individual hair strands, authentic fabric texture, physically accurate lighting, global illumination, realistic reflections, volumetric light where appropriate, natural shadow falloff, optical lens behavior, cinematic color science, HDR, extremely high detail, film-quality realism. No CGI, no 3D render, no illustration, no anime, no cartoon, no painting, no stylization, no plastic skin. Single final image, 3×3 grid, 9 equal panels, thin black borders, no captions, no text, no numbers, no speech bubbles, no watermark. Every panel must look like a frame from the same live-action film with perfect identity and environmental continuity. #промпт_дня #промпт@ii_razum #chatgpt #изображение #ai #нейросети