Use @image1 as the ONLY reference image for both the environment and the exact flight path. Create a 15-second ultra-realistic FPV drone video captured exactly as if filmed by a real professional FPV drone pilot. The entire sequence must be one continuous, uninterrupted drone shot with authentic FPV physics, realistic acceleration, momentum, banking, and natural camera stabilization. It must feel identical to real drone footage captured in real life. LOCATION LOCK (ABSOLUTE REQUIREMENT): The location in @image1 MUST remain 100% identical throughout the entire video. Do NOT modify, redesign, regenerate, or reinterpret the environment. Every building, road, tree, park, bridge, canal, TV Tower, shadow, skyline, and city layout must remain exactly where it is in @image1. This is the Tashkent TV Tower and its surrounding district. Do not replace any architecture. Do not add new buildings. Do not remove existing buildings. Do not change the weather. Do not change the lighting direction. Do not change the city layout. Treat @image1 as a locked real-world aerial photograph. START POSITION: The drone starts exactly from the white marked point shown in @image1 (bottom-left corner near the canal and road). Camera height is low, just above the trees. The drone immediately accelerates smoothly toward the TV Tower. FLIGHT PATH (FOLLOW THE RED LINE EXACTLY): 0:00–0:03 The drone starts from the white marked point and flies low over the canal and road toward the TV Tower. Smooth FPV acceleration. Maintain realistic altitude. Natural forward motion with slight FPV banking. 0:03–0:06 Approach the tower base. Begin a smooth clockwise spiral around the lower structure while gradually climbing. Keep the tower centered in frame. Maintain consistent speed. 0:06–0:10 Continue spiraling upward around the tower shaft exactly following the red path. Fly very close to the tower to emphasize scale while maintaining realistic safety distance. Camera movement must remain fluid with natural FPV momentum. No impossible movements. 0:10–0:13 Reach the antenna section. Perform one smooth circular orbit around the upper antenna exactly as marked in red. Capture the city skyline naturally. No sudden rotations. 0:13–0:15 Begin a wide descending spiral around the tower. Slowly pull away from the tower while keeping it visible in frame. End with a beautiful wide aerial view of the entire TV Tower and surrounding Tashkent city. CAMERA STYLE: Real FPV drone footage. Professional cinematic drone operator. Authentic drone physics. Natural stabilization. Slight FPV vibration. Realistic inertia. Realistic momentum. Realistic banking. No robotic motion. No CGI feeling. CAMERA SETTINGS: Ultra realistic. 4K UHD. 16:9. 30fps. Natural motion blur. Real drone lens. Wide-angle FPV lens (approximately 16–20mm equivalent). Real daylight exposure. No artificial HDR. ENVIRONMENT: Golden hour sunlight exactly matching @image1. Natural atmospheric haze. Realistic shadows. Natural tree movement. Realistic roads. Real moving cars. Natural city life. Everything must remain identical to the reference. IMPORTANT: The drone itself must NEVER become visible. This is footage captured FROM the drone. Only the drone camera perspective is visible. NEGATIVE PROMPT (STRICT): Do NOT change the location. Do NOT regenerate the city. Do NOT alter the TV Tower. Do NOT change the tower proportions. Do NOT move any buildings. Do NOT add buildings. Do NOT remove buildings. Do NOT modify roads. Do NOT modify parks. Do NOT modify trees. Do NOT modify the canal. Do NOT modify the skyline. Do NOT modify the horizon. Do NOT change weather. Do NOT change sunlight direction. Do NOT change time of day. Do NOT add fog. Do NOT add clouds different from the reference. Do NOT create fantasy architecture. Do NOT create CGI-looking visuals. Do NOT create synthetic textures. Do NOT create game-engine graphics. Do NOT create AI-generated city layouts. Do NOT zoom digitally. Do NOT use teleporting camera. Do NOT use impossible FPV movements. Do NOT use sudden acceleration. Do NOT use abrupt camera shakes. Do NOT use jerky transitions. Do NOT use cuts. Do NOT use scene transitions. Do NOT use speed ramps. Do NOT use cinematic fake camera moves. Do NOT show the drone. Do NOT show drone propellers. Do NOT show drone shadow. Do NOT add text. Do NOT add subtitles. Do NOT add logos. Do NOT add watermarks. Do NOT add UI elements. Do NOT add visual effects. Do NOT add lens flares. Do NOT add unrealistic reflections. Do NOT add artificial lighting. Do NOT distort perspective. Do NOT deform buildings. Do NOT warp the tower. Do NOT blur the city. Do NOT change the composition outside the actual drone movement. FINAL REQUIREMENT: The finished result must be indistinguishable from genuine FPV drone footage filmed by an experienced drone pilot around the real Tashkent TV Tower, following the exact red flight path drawn on @image1 while preserving the real-world location perfectly.

@image1 realistic cinematic celebrity arrival scene, single continuous one-take shot, handheld camera from inside a dense crowd, natural micro-shake, realistic inertia and physics, no cuts, 4K, anamorphic 2.39:1, film grain. Strict identity lock — main character’s face must exactly match the reference, no changes, no drift, consistent throughout all frames. Audio: natural environment only — loud crowd cheering, people shouting, camera shutter clicks, phone recording audio, airport announcements echoing, footsteps, fabric movement, distant engine sound. Lighting: natural daylight coming through airport glass panels, reflections, realistic shadows, slight haze for depth. Main character: calm and confident presence, subtle controlled smile, natural walking behavior, face unchanged. Outfit (strict lock): beige/cream blazer worn open, clean white t-shirt inside, beige pants, white premium sneakers, realistic cloth simulation. Accessories: transparent black square-frame luxury sunglasses with gradient lenses (Tom Ford style), no changes. Camera: handheld realism, micro jitter, no gimbal stabilization, no drone feel, natural focus breathing. TIMELINE: 0–3s — camera starts inside the crowd behind barricades, partially blocked view, phones raised recording with visible screens, loud chaotic crowd, camera searching for a clear view toward the arrival gate. 3–6s — camera lifts slightly above shoulder level, slightly shaky but controlled, focus shifts between foreground heads and the arrival gate, media flashes, anticipation builds. 6–10s — security steps in and pushes the crowd back, camera reacts with natural shake, through gaps the celebrity appears in the distance, slightly blurred at first, then becomes sharp as focus locks, walking forward with bodyguards. 10–13s — natural push-in (physical movement, not zoom), celebrity centered, confident walk, raises hand and gives a subtle wave, camera follows while maintaining framing in chaotic crowd. 13–15s — camera shifts and tilts to follow, celebrity reaches the vehicles, Lamborghini Aventador + Mercedes G63 + Innova visible, bodyguard opens the door, celebrity enters, door closes, engine sound rises, vehicles begin moving, camera lifts slightly to capture the final moment. NEGATIVE PROMPT: no face change, no identity drift, no extra characters, no duplicates, no cartoon style, no low quality, no blur, no unrealistic physics, no over-smooth camera, no artificial motion

Meni va parichkamni 6–7 yoshli bola va qiz holatiga kichkinalashtirib ber

Cinematic one-take long shot, 15 seconds, vast windy daisy field beneath a bright blue summer sky with large soft clouds. Natural daylight, soft wind through flowers and fabric, atmosphere calm, nostalgic, poetic, airy, visually elegant. Opening frame @Main image : A @BOY Image young man wearing a loose soft-blue Y2K linen shirt over a white tee and a @GIRL IMAGE young woman wearing a flowing soft-blue Y2K summer dress stand facing each other approximately two meters apart, posture calm and graceful. The man holds a bouquet of white daisies behind his back. Camera movement is divided into three continuous phases, single continuous motorized crane arc, perfectly smooth and constant velocity from 0 to 15 seconds, no acceleration, no deceleration, no shake, no jitter, as if mounted on a precision robotic arm on a frictionless track, rightward crane-arc orbital: Phase One (0–5s): Camera begins at frontal waist height with both figures fully framed. It slowly arcs rightward through their side profile and comes to rest directly behind the woman. Her flowing soft-blue dress naturally and completely fills the entire frame. Foreground daisies blur softly around the edges. Phase Two (5–10s): Camera continues rightward arc emerging from behind the woman. As the frame opens, the man's position is now occupied by a @BOY KID Image 3 6-year-old boy wearing the same soft-blue shirt, holding a bouquet of white daisies in front of him, maintaining the same standing position and eye-line toward the woman. Without pause the camera continues its rightward arc while slowly lowering below the boy's waist into a low-angle upward perspective pushing close toward his back. The boy's soft-blue shoulders and the back of his head naturally fill the entire frame with wide-angle distortion. Phase Three (10–15s): Camera continues rightward arc emerging from behind the boy. As the frame opens, the woman's position is now occupied by a @GIRL KID image 6-year-old girl wearing white headband and the same soft-blue dress, the same height as the boy. The boy now holds a single white daisy. Both children stand facing each other with the same elegant posture. Camera settles into a frontal composition of both children then performs a slow gentle pull-back, ending wide and open. Visual Style: 35mm wide-angle lens, shallow depth of field, Kodak film grain texture, Fujifilm overexposed 35mm slide-film look, bright late-afternoon sunlight, soft atmospheric haze, powder-blue and white summer palette, pale sage-green field tones, painterly blue sky with cream-white clouds, medium contrast, natural highlight roll-off, soft cinematic bloom, natural lens flare. Ambient sound: soft wind through flowers, soft breeze

@image1 = MAN CHARACTER REFERENCE @image2 = WOMAN CHARACTER REFERENCE [CINEMATIC SETUP] Film Style: Ultra-realistic cinematic smartphone footage, captured on iPhone 17 Pro Max video camera. Natural computational photography look, realistic smartphone dynamic range, subtle digital sharpening, authentic mobile sensor grain in darker areas. Cinematic but grounded in reality — not stylized Hollywood CGI. Slight shallow depth of field from smartphone portrait-video processing. Visual realism inspired by real viral disaster footage captured on phones. Color Grade: Natural midday color palette with realistic daylight tones. Soft cold-neutral urban colors, pale blue sky mixed with grey clouds, realistic skin tones, balanced exposure. Slight desaturation from atmospheric haze and smoke after impact. No cinematic teal-orange grading. Lighting: Realistic noon daylight entering through massive panoramic skyscraper windows. Soft natural ambient skylight illuminating faces. Reflections from glass windows physically accurate. Explosion outside creates occasional warm orange flicker on skin and window edges. No artificial studio lighting. Camera: Handheld but controlled iPhone 17 Pro Max footage. Mostly static framing with subtle natural human micro-movements and breathing motion. Occasional autofocus breathing and realistic exposure adaptation from bright explosion flashes. No cinematic crane shots, no drone movement, no impossible camera physics. Realistic smartphone motion blur and stabilization behavior. Lens Feel: iPhone 17 Pro Max main lens equivalent, realistic smartphone field of view. Slight edge softness and digital HDR behavior. Clean realistic reflections on panoramic glass windows. [CHARACTER REFERENCE LOCK] Use @image1 as the exact male character reference. Use @image2 as the exact female character reference. STRICT CHARACTER LOCK: Maintain exact facial structure, hairstyle, skin texture, eye shape, ethnicity, proportions, clothing silhouette, and identity from both reference images throughout the entire sequence. Do not beautify, stylize, or alter facial anatomy. No character drift. No age change. No hairstyle change. No body proportion change. [SUBJECT] The man from @image1 and the woman from @image2 inside a luxury American skyscraper penthouse with floor-to-ceiling panoramic windows overlooking Manhattan skyline. The woman has tears forming in her eyes during emotional close-ups. The man also becomes emotional with watery eyes and subtle trembling lips. Both characters remain fully realistic with natural facial muscle movement and believable emotional acting. [SCENE] Interior of a modern ultra-luxury skyscraper apartment high above New York City during midday. Massive panoramic glass windows reveal the Manhattan skyline in full daylight. Empire State Building visible prominently in background along with dense American skyscrapers and distant haze. Sky is partly cloudy with realistic noon lighting. Atmosphere initially calm and ordinary before disaster escalates. Distant traffic and urban movement visible far below. Interior lighting entirely natural from windows. A meteor enters Earth atmosphere and strikes the city in the distance during the argument. [ACTION TIMELINE] [0–3s] Medium two-shot captured on iPhone 17 Pro Max. The woman and man argue emotionally near the panoramic window. Realistic human speaking movement, overlapping dialogue, emotional frustration. The woman raises her voice with hurt expression while the man responds defensively. Natural smartphone framing. Slight handheld micro-motion. Realistic room tone and distant city ambience. [3–5s] Same handheld phone framing. A bright meteor streak suddenly appears outside the window descending toward Manhattan skyline. Both characters stop arguing mid-sentence. The man slowly turns toward the window first. The woman follows half a second later. Realistic natural reaction timing. [5–7s] Wide smartphone shot toward the skyline. Massive explosion erupts between skyscrapers in the distance. Huge realistic shockwave cloud expands outward with physically accurate smoke, debris, and fire. Buildings subtly shake from impact vibration. Window glass rattles realistically. Interior objects tremble naturally. No exaggerated Hollywood fireball — grounded realistic large-scale urban destruction. [7–9s] Close-up on the woman’s face shot on iPhone 17 Pro Max. Extremely realistic emotional acting. Tears forming in her eyes. Breathing becomes unstable. Skin texture, pores, moisture, eyelashes, and eye reflections highly realistic. Background softly blurred with destroyed skyline behind her. She looks toward the man in complete shock. [9–11s] Close-up on the man’s face. His eyes become watery. Lips trembling slightly. Fear mixed with helpless acceptance. Reflection of smoke and fire visible subtly in his eyes. Natural smartphone autofocus shift between eyes and background. [11–13s] Medium side-profile shot. Massive dust cloud and expanding debris wave rapidly approaching the skyscraper outside the panoramic windows. The man pulls the woman into a desperate emotional embrace. Both hold each other tightly. Fear, sadness, and surrender visible in body language. Exterior visibility rapidly disappears behind smoke and debris. [13–15s] Same handheld iPhone framing. The dust wave slams into the building exterior. Entire frame consumed by dense grey dust, shattered particles, smoke, and darkness. The couple becomes barely visible silhouettes through the dust-filled room while still embracing each other. Audio becomes muffled and nearly silent. Final frame fades into dirty grey dust haze. [SOUNDS] Beginning: Distant traffic ambience, muffled city sounds, emotional argument dialogue, subtle room echo. Meteor: Deep atmospheric rumble building slowly. Impact: Massive realistic low-frequency explosion boom, distant structural collapse echoes, glass vibration. Shockwave: Pressurized roaring wind, debris impacts, building creaks, alarms faintly activating in distance. Emotional Close-Ups: Breathing, heartbeat, subtle crying, near silence. Final: Heavy dust impact, muffled low-frequency rumble, complete audio suppression. [NEGATIVE] no sci-fi CGI look, no superhero aesthetic, no blue hour, no sunset, no orange cinematic grading, no dramatic Hollywood lighting, no impossible explosion physics, no stylized destruction, no futuristic UI, no camera drones, no cinematic crane shots, no smooth gimbal movement, no unrealistic fireball, no fantasy meteor, no overexposed highlights, no artificial beauty filter, no anime emotion, no character drift, no face deformation, no extra people, no military elements, no unrealistic building collapse,

MANI 2 YOSHLI BOLA QILIB BER

@image_1 @image_2 Используйте @image_1 как взрослую версию персонажа. Используйте @image_2 как детскую версию того же персонажа. Создайте 15-секундное эмоциональное студийное видео. Взрослый из @image_1 и ребёнок из @image_2 сидят вместе в чистой профессиональной студии с простым бесшовным бежевым циклорамным фоном, монохромным и минималистичным. Атмосфера должна быть тёплой, мягкой, эмоциональной, элегантной и глубоко человечной. ВАЖНЫЕ ПРАВИЛА ПЕРСОНАЖА: Взрослый должен выглядеть точно так же, как @image_1. Ребёнок должен выглядеть точно так же, как @image_2. Они должны ощущать себя одним и тем же человеком в двух разных возрастах. Сохраняйте индивидуальность лица, причёску, цвет волос, оттенок кожи, стиль одежды и общий внешний вид на всех кадрах. Не меняйте их одежду, причёску, пропорции или атмосферу от кадра к кадру. Сохраняйте оба персонажа стабильными и последовательными на протяжении всего видео. ЭМОЦИОНАЛЬНЫЙ ТОН: Эмоции должны быть трогательными и искренними, но естественными и умеренными. Не преувеличивайте свою игру. Взрослый улыбается тепло и эмоционально, но сдержанно и правдоподобно. Ребёнок ведёт себя как настоящий игривый мальчик — улыбается, делает милые детские жесты и реагирует невинно. СТИЛЬ ОПЕРАТОРА: Используйте плавное движение камеры только для кинематографа. Никаких резких скачков, никаких неуверенных переходов, никаких неловких изменений. Все переходы между кадрами должны быть мягкими, плавными и эмоционально плавными. Используйте профессиональное кинематографическое кадрирование с крупными планами, средними планами, средними крупными планами и деталями. Камера должна мягко скользить между кадрами. Резких изменений позы между монтажами не было. Никаких резких изменений в гардеробе или прическах. Нет дрейфа атмосферы. ОСВЕЩЕНИЕ: Профессиональное студийное освещение. Мягкий рассеянный кинематографический ключевой свет. Мягкий заполняющий свет. Лёгкий тёплый обод или разделение краёв. Освещение должно ощущаться премиальным, элегантным, мягким и погружающим. Бежевый фон студии должен оставаться чистым, минималистичным и визуально спокойным. МУЗЫКА: Добавьте только глубоко эмоциональную инструментальную фоновую музыку. Без вокала. Используйте мягкую эмоциональную кинематографическую музыку с мягким фортепиано и тёплыми оркестровыми струнными. Музыка должна вызывать у зрителя эмоции и эмоции, но не слишком драматично. Музыка должна мягко подняться во время объятий и последнего взгляда. ТЕКСТ НА ЭКРАНЕ: Используйте простой чистый чёрный текст. Above the adult from @image_1, show: "2026" Above the child from @image_2, show: "2003" The text must appear in the opening shot and return again in the final emotional shot. Do not add any other text. TIMELINE — 15 SECONDS TOTAL: 00:00 - 00:01 — MEDIUM SHOT Both characters are visible in frame. They are sitting across from each other at a table, looking at one another. The camera is straight-on at eye level. The frame is calm, balanced, and emotional. In the background / upper part of the frame, clean black text is visible: Above the adult: "2026" Above the child: "2003" 00:01 - 00:03 — CLOSE-UP The camera focuses only on the child’s face. Camera angle is at the child’s face level. The child smiles softly, makes a cute little playful expression, and reacts naturally like a small boy. Keep the emotion innocent and moderate. 00:03 - 00:05 — MEDIUM CLOSE-UP Show the adult’s face and upper chest. He looks toward the child and begins moving toward him emotionally. The camera is at the adult’s face level, slightly from the side to reveal the facial emotion more clearly. He smiles warmly, softly, and naturally. 00:05 - 00:07 — DETAIL / CLOSE-UP SHOT The hugging action begins. Frame mainly the adult’s hands and the back side of the child. Focus on the movement of the hands and the start of the embrace. The hug must feel gentle, sincere, and emotionally meaningful. 00:07 - 00:09 — MEDIUM SHOT / SIDE PROFILE Show both characters hugging from the side profile. The camera frames both of them clearly in the embrace. The hug must feel warm, touching, and natural. Keep the body language believable and physically correct. 00:09 - 00:10 — CLOSE-UP Close-up of the adult’s face. Сосредоточьтесь на его закрытых глазах и спокойном эмоциональном чувстве на лице. Угол прямой, не низкий и не высокий. На его лице выражены спокойствие, тепло и тихая радость. Эмоции должны оставаться естественными и контролируемыми. 00:10 - 00:11 — СРЕДНИЙ КАДР Персонажи аккуратно отделяются от объятий и возвращаются в свои прежние положения, снова лицом друг к другу. Камера возвращается к более раннему более широкому формату кадрирования. Движение должно ощущаться плавно и непрерывно, без визуального прыжка. 00:11 - 00:14 — СРЕДНИЙ КАДР Кадр остаётся в том же формате. Освещение становится чуть теплее и эмоциональным. Взрослый и ребёнок улыбаются друг другу в одной рамке. Взрослый из @image_1 смотрит на себя с любящей, мягкой улыбкой. Ребёнок из @image_2 невинно улыбается в ответ, как маленький мальчик, счастливо реагирующий на игру с ним. В верхней части кадра снова появляется чёрный текст: Над взрослым: «2026» Над ребёнком: «2003» 00:14 - 00:15 — ФИНАЛЬНЫЙ ХОЛД Держите финальную эмоциональную композицию. Оба персонажа продолжают мягко улыбаться друг другу. Инструментальная музыка сохраняется мягко. Завершите на тёплом, трогательном, кинематографическом настроении. ФИНАЛЬНЫЙ СТИЛЬ: Ультрареалистично. Эмоционально сильный, но тонкий. Профессионального студийного качества. Мягкий бежевый минималистичный фон. Одинаковый гардероб, причёска, логика поз и атмосфера во всех кадрах. Плавные кинематографические переходы. Мерцания нет. Никакого отклонения личности. Нет стильного дрейфа. Никакой преувеличенной актёрской игры. Без лишних людей. Только взрослая версия и его детская версия в трогательном моменте воссоединения в студии.

Use the uploaded face as the exact facial identity reference. Ultra-realistic Korean live sports broadcast aesthetic during an intense FC Barcelona vs Real Madrid football match inside a massive packed stadium. The person is accidentally captured by the stadium camera during an emotional pause in the game. Authentic Korean idol-like visual styling, soft glossy skin with realistic texture, subtle under-eye shadows, cinematic natural makeup, authentic unposed expression, slightly tired and emotionally distant mood. The subject is not looking at the camera. Their gaze is fully focused on the football match off-screen, looking toward the pitch with a calm, mysterious, emotionally unreadable expression, as if lost in thought. Natural candid behavior, completely unaware of the camera, subtle cinematic eye reflections from the stadium floodlights, quiet magnetic presence. Natural loose hair strands illuminated by stadium lights, subtle silver accessories, natural effortless styling. The subject is wearing an oversized Real Madrid football jersey layered over a dark inner shirt, relaxed casual fit, realistic football fan fashion. No beer cup in hand. Hands relaxed naturally or holding a white Real Madrid scarf. Surrounded by other football spectators wearing mixed Barcelona and Real Madrid jerseys. Authentic El Clásico atmosphere, crowd holding phones, scarves, towels, glowing LED lights, realistic audience behavior, shallow depth of field, creamy cinematic background blur. Massive football stadium environment during nighttime match, cinematic cool indoor floodlight lighting mixed with warm natural skin tones, slight atmospheric haze from stadium lights. Realistic football TV broadcast overlays visible on screen: - ZO'R TV LIVE watermark - Match scoreboard overlay - Time: 39:00 - Score: Barcelona 3 - 1 Real Madrid - FC Barcelona logo and Real Madrid logo visible on scoreboard graphics - Uzbek sports TV broadcast aesthetic - subtle live television compression artifacts - faint interlaced scanlines - muted blue-gray cinematic sports broadcast color grading - soft cinematic contrast Sony professional sports broadcast camera realism, documentary-style image quality, photorealistic skin detail, accidental viral crowd-cam aesthetic, Korean Twitter fancam energy mixed with UEFA Champions League atmosphere, cinematic K-drama mood, authentic live sports television screenshot, ultra realistic, emotionally cinematic, highly detailed skin pores, natural stadium reflections, realistic lens compression, broadcast zoom camera look, realistic football audience depth, premium sports cinematography. Aspect ratio 16:9.

Zoom-in medium shot, ultra-realistic Korean live sports broadcast during a major FC Barcelona vs Real Madrid football match inside a packed stadium. A mysterious spectator is accidentally captured by the stadium crowd camera during a quiet pause in the match. Calm, distant, emotionally unreadable, fully absorbed in the game and unaware of the camera. No posing or exaggerated emotion — just a real ordinary fan lost in thought. The subject wears layered Korean stadium fashion: an oversized football jersey over a dark long-sleeve shirt, relaxed casual fit, subtle accessories, holding blue inflatable thunder sticks naturally in both hands. Medium-shot waist-up framing with the stadium environment visible around the subject. Slow cinematic zoom-in through the sequence. The subject remains the only thing in sharp focus while the crowd stays heavily blurred with creamy telephoto bokeh. Minimal natural movement only: slow blinking, subtle breathing, tiny eye tracking, small head movement, slight hand shifts. Shot 1 (0–3s) — Medium-wide shot: The subject sits among cheering fans during the Barcelona vs Real Madrid match. Eyes fixed on the off-screen game. Slow blink. Slight breathing movement. A loose strand of hair shifts naturally. Slow zoom-in begins. Feels like a candid live-TV crowd-cam moment. Shot 2 (3–6s) — Tight medium close-up: Eyes slowly follow the game action. Natural blink. Lips slightly parted with soft breathing. Tiny swallow. Small eye focus shift before returning to the match. Expression remains restrained and mysterious. Slow zoom continues. Shot 3 (6–10s) — 3/4 side medium shot: The subject quietly watches the match. One slow blink. Thunder sticks shift slightly in the hands. Tiny head tilt. No smile or dramatic reaction — only calm focus on the game. Final zoom-in ends softly. Authentic football stadium ambience: layered crowd noise, fan chants, thunder stick sounds, distant shouts, ball kicks, referee whistles, stadium announcer ambience, occasional crowd surges, and match atmosphere. No dialogue. Ultra-realistic skin texture, cinematic stadium lighting, shallow depth of field, creamy telephoto blur, Korean tvN Sports / SBS Sports broadcast realism, Sony sports broadcast camera look, subtle live-TV compression artifacts, documentary realism, 24fps, smooth cinematic cuts, realistic sports crowd-cam aesthetic. Aspect ratio: 9:16 vertical. Ultra-realistic, cinematic vertical sports broadcast look.

Use @image1 as the main base scene and environment reference. Use @image2 as the exact 1:1 identity reference for the only new man entering the scene. Create an ultra-realistic cinematic action scene in one continuous shot. Start exactly from @image1: the armed man is standing inside the elevator holding a prop gun, and the red-shirt man is standing in the hallway facing him. Preserve the elevator, hallway, walls, lighting, camera angle, clothing, faces, and cinematic quality exactly as in @image1. Only the man from @image2 enters the scene and performs the action. The two characters from @image1 must not turn into different people, must not change faces, must not change clothes, and must stay visually identical to the original frame. The man from @image2 enters from the opposite side of the red-shirt man, from the side angle, moving fast and confidently. His face must look exactly like @image2, 1:1 facial identity match, same hairstyle, same body shape, same burgundy polo shirt, same white pants, same proportions. No face drift, no morphing, no different person. He quickly approaches the elevator, grabs the armed man by the arm and shoulder, pulls him out of the elevator in a realistic staged movie-combat motion, and knocks the prop gun safely away from his hand. Then @image2 gives the armed man one sharp staged slap to the face. After that, @image2 turns toward the red-shirt man and gives him one sharp staged slap too. The action should look like a dramatic movie scene, intense but non-graphic. No blood, no injury, no gore. The armed man reacts with shock and steps backward. The red-shirt man also reacts with surprise and slightly steps back. @image2 remains in control of the scene, standing between both characters with a serious and dominant expression. Camera movement: one continuous handheld cinematic shot, slight pan toward @image2 as he enters, small natural reaction shake during the staged slaps, quick reframe to keep all three characters visible, then subtle push-in on @image2 at the end. No hard cuts, no sudden zoom, no random camera shake. Identity lock: @image2 must remain exactly the same person as the reference image, 1:1 face match, same outfit, same hair, same body shape. The two characters from @image1 must remain exactly the same as in the original scene. No face morphing, no identity drift, no changing clothes, no extra people, no background changes. Quality: ultra-realistic cinematic drama, realistic body physics, natural hand contact, believable staged combat, stable lighting, sharp details, movie-like tension. Negative prompt: low quality, blurry face, face morphing, identity drift, wrong person, changed face, changed clothes, changed hairstyle, bad anatomy, deformed hands, extra fingers, broken limbs, floating limbs, robotic motion, unnatural physics, teleporting, extra people, random objects, background changes, flicker, glitch, frame jitter, plastic skin, hard cuts, fast zooms, extreme camera shake, cartoon, anime, blood, gore, graphic injury, real violence, unsafe gun handling.