Magical Day in Hidden Japanese Town

30 seconds | 16:9 | live-action supernatural horror drama Fictional dramatization. Documentary-style visual realism. SETTING An isolated old farmhouse surrounded by dense forest and fields. Time: late afternoon transitioning into night. Weather: overcast, cold and slightly windy. Aged wooden exterior, neglected rooms, narrow staircase, old family photographs, dusty furniture, dim corridors and a locked basement door. Natural environmental activity, distant wind, creaking wood and subtle household sounds. CHARACTERS A young couple in their early 30s and their 8-year-old daughter. Maintain consistent faces, clothing, hairstyle, physical appearance and character identities throughout the sequence. Human behaviour remains grounded and restrained. STORY 00–04s — ARRIVAL Wide observational shot establishing the isolated farmhouse. The family carries boxes from their vehicle toward the front door. The daughter stops and looks toward an upstairs window. The parents continue walking, unaware. Natural wind moves the trees and clothing. 04–08s — THE HOUSE Handheld perspective follows the family through the dusty interior. The father explores the living room while the mother opens curtains. The daughter quietly looks toward the staircase. A faint creaking sound comes from upstairs. She asks, "Is someone up there?" Her parents dismiss it as the old house settling. 08–12s — THE FIRST SIGN The family moves through the hallway. The daughter suddenly stops. At the far end of the corridor, a bedroom door slowly moves open by itself. The parents turn toward it. The camera operator hesitates and slowly approaches. The room appears completely empty. Avoid exaggerated supernatural movement. 12–16s — NIGHT Cut to the farmhouse after dark. The family is preparing for bed. The daughter suddenly looks toward the hallway. A faint silhouette appears at the far end for only a moment before disappearing when the mother looks. The daughter quietly says, "She's there again." The parents exchange an uneasy glance. 16–20s — THE FOOTSTEPS Handheld camera follows the father through the dark hallway. Slow footsteps can be heard above him. He looks toward the staircase. The footsteps stop. He cautiously climbs the stairs. The camera follows behind with natural handheld movement. A bedroom door at the end of the hallway is slightly open. 20–24s — THE LOCKED ROOM The father discovers the daughter standing outside a locked basement door. She is completely still. She quietly says, "She told me not to open it." The father looks at the door. The handle slowly moves from the other side. The camera operator instinctively steps backward. 24–28s — THE BASEMENT The door suddenly opens slightly. Cold darkness fills the doorway. The mother grabs her daughter and pulls her backward. A faint female voice comes from inside the basement. The camera struggles to maintain focus in the darkness. No visible monster. No graphic violence. 28–30s — FINAL REVEAL Camera remains in the hallway. The family backs away from the basement. The daughter suddenly looks past her parents toward the upstairs landing. A pale female figure is standing silently at the top of the stairs. Nobody notices her. The camera slowly turns toward the figure. The image abruptly cuts to black. CAMERA Grounded supernatural-documentary cinematography. Mix: wide environmental coverage medium handheld following shots brief close reaction shots slow observational hallway perspectives Natural handheld instability. Realistic operator movement. Plausible camera positions. Autofocus hunting in darkness. Small framing corrections. Natural motion blur. No impossible camera movement. No floating camera. No excessive slow motion. No polished Hollywood horror cinematography. LIGHTING Lighting originates from practical and environmental sources. Natural overcast daylight during arrival. Warm household lamps and weak practical lighting after dark. Natural exposure changes when moving between illuminated rooms and dark hallways. Deep but detailed shadows. Realistic highlight roll-off. No artificial rim lighting. No supernatural glow. PHYSICS Real gravity and momentum. Feet maintain contact with floors and stairs. Furniture has believable weight. Doors move through physical hinges. Curtains respond to airflow. Dust reacts naturally to footsteps. Clothing responds to movement and wind. Wood, doors and floorboards behave realistically. No floating objects. No impossible movement. No exaggerated supernatural physics. HUMAN PERFORMANCE Restrained, believable fear. Quiet breathing. Small eye movements. Hesitation before entering dark spaces. Natural glances toward unexplained sounds. Parents instinctively protect their daughter. The daughter behaves frightened but not theatrically possessed. No screaming without cause. No exaggerated facial expressions. No heroic behavior. No theatrical horror acting. AUDIO Natural farmhouse ambience. Wind through trees. Distant insects. Wood creaking. Footsteps on wooden floors. Floorboards reacting under weight. Quiet household sounds. Distant birds during daylight. Low nighttime ambience. Soft breathing. Subtle door movement. Faint footsteps upstairs. A barely audible female voice from the basement. No music. No jump-scare sound effects. No oversized Hollywood trailer effects. No artificial bass drops. VISUAL CHARACTER Photorealistic live-action footage. Observational supernatural-documentary character. Natural skin texture. Realistic fabric and materials. Authentic old farmhouse textures. Subtle sensor noise. Natural motion blur. Imperfect handheld framing. Practical lighting. Atmospheric darkness. The supernatural presence should feel physically present within an otherwise completely realistic environment. Avoid: video-game aesthetics CGI appearance excessive HDR oversaturated colors floating ghosts

Create a 30-second ultra-realistic video that feels like a genuine forgotten home-video recording from the early 2000s. It must NOT look cinematic, staged, polished, or AI-generated. MAIN SUBJECT: Young Korean woman, early 20s, natural everyday appearance. Faded charcoal-grey sleeveless crop top, loose high-waisted light-wash jeans, black canvas sneakers, black cord necklace. Black wavy hair tied in a messy side ponytail with wispy bangs. Minimal makeup, realistic pores and skin texture, subtle imperfections. Warm, approachable personality. Maintain EXACTLY the same identity, face, hairstyle, clothing, proportions, and appearance throughout. LOCATION: Authentic quiet Korean residential neighborhood on a warm late afternoon. Narrow concrete lanes, low-rise houses, small gates, potted plants, old bicycles, utility poles, tangled overhead wires, parked scooters, laundry hanging from balconies, mature trees, uneven concrete, faded walls. No shops, signs, advertisements, cafés, crowds, or tourist landmarks. CAMERA: Early-2000s consumer MiniDV camcorder recorded casually by her friend. Heavy natural handheld shake, imperfect framing, occasional accidental cropping, autofocus hunting, lens breathing, exposure pumping when entering shade, slight rolling shutter, motion blur during quick movement, mild digital compression, subtle sensor noise, faded consumer-camera colors, soft contrast. No stabilization. No cinematic camera movement. No gimbal. No modern HDR. No artificial film look. 00:00–00:04 The recording begins abruptly while the camera is already moving. She is walking out through a small metal gate carrying a reusable grocery bag. She notices the camera and gives a slightly confused smile, as if she didn't realize her friend had started recording. The operator walks backward awkwardly and briefly loses her face from frame. 00:04–00:08 They turn into a narrow residential alley. She suddenly notices a small delivery box left near a neighbor's gate. She picks it up and looks around, trying to figure out whose it is. The camera moves closer too quickly, autofocus briefly locks onto the box instead of her face. 00:08–00:12 A middle-aged Korean neighbor appears from a doorway and gestures that the package is hers. The woman laughs softly, hands it over, and gives a small polite bow. The interaction feels spontaneous and unscripted. The camera shakes slightly as the person filming laughs behind the camera. 00:12–00:16 She continues walking and stops beside an old vending-style outdoor water dispenser near a residential entrance. She takes a small sip from her reusable bottle, wipes her mouth with the back of her hand, then notices the camera still pointed at her and playfully covers the lens for a moment. 00:16–00:21 The camera drops slightly as her hand moves away, revealing her walking ahead beneath large trees. Sunlight flickers naturally across the pavement and her clothing as leaves move in the breeze. A bicycle passes slowly in the distant background. She casually kicks a small fallen leaf forward while walking. 00:21–00:25 She reaches a tiny neighborhood convenience-style residential rest area with a concrete bench, but there are NO commercial signs or branding. She sits down, places the grocery bag beside her, and checks her phone. She suddenly looks toward something off-camera and smiles naturally, reacting to her friend rather than posing. 00:25–00:28 The friend walks closer. She looks directly into the lens and says casually, "Why are you still filming me?" with a small laugh. Her expression should feel completely spontaneous, not performed. 00:28–00:30 She stands and starts walking toward the camera. The operator instinctively backs away, causing noticeable shaky footage and brief focus hunting. She reaches toward the camera as if trying to stop the recording. The image abruptly cuts to black while her hand is still approaching the lens. AUDIO: ONLY authentic location sound. No music. No cinematic sound design. No narration. Include distant birds, leaves moving, footsteps on concrete, fabric rustling, a faint scooter passing, bicycle wheel sounds, neighborhood voices far away, subtle traffic in the distance, the neighbor's brief greeting, her natural laughter, and the actual sound of the camcorder operator moving and breathing. Dialogue should sound naturally recorded through a cheap built-in microphone, slightly compressed and imperfect. REALISM REQUIREMENTS: The entire video must feel accidentally captured by a real person, not directed for a film. Avoid beautiful composition, dramatic lighting, perfect framing, slow motion, smooth transitions, excessive depth of field, exaggerated facial expressions, perfect skin, artificial bokeh, cinematic color grading, CGI-looking environments, or overly clean textures. Small imperfections are essential: missed focus, exposure fluctuations, awkward framing, slight camera shake, realistic motion blur, occasional clipped highlights, compression artifacts, imperfect timing, natural pauses, and genuine micro-expressions.

Create a 15-second ultra-realistic cinematic pizza-making sequence featuring the girl from the uploaded reference image as the exact character reference. Preserve her facial identity, green-hazel eyes, long jet-black hair, skin tone, facial structure, and overall appearance consistently throughout every shot. Dress her in a stylish black apron over her white fitted T-shirt, with her hair tied into a practical low ponytail. The video opens with an intense close-up of her holding a freshly baked pizza slice toward the camera. She takes a confident bite as molten cheese stretches dramatically from the slice, steam rising around her face. She immediately turns toward the counter and begins making another pizza. She spins a round piece of dough high above her head, catches it perfectly, and slaps it onto the counter as flour bursts into the air. The camera follows the movement with a fast whip-pan. She rapidly spreads bright-red tomato sauce across the dough, then throws fresh mozzarella, ricotta, and pepperoni onto the pizza with energetic precision. Use fast macro cuts of sauce swirling, cheese landing, and pepperoni bouncing onto the dough. She slides the topped pizza into a blazing stone oven. Extreme macro shots show the crust rapidly puffing into huge leopard-charred bubbles, pepperoni curling into crispy cups, cheese bubbling and melting, and heat distortion shimmering around the pizza. She pulls the finished pizza from the oven, scatters fresh basil over the bubbling surface, then lifts one slice high. A huge molten cheese pull stretches between the slice and pizza as she gives the camera a confident satisfied look. Finish with a dramatic close-up of the golden blistered crust, melted cheese, crispy pepperoni, fresh basil, and steam filling the frame. Style: Premium cinematic food commercial, female chef protagonist, aggressive handheld camera, whip pans, smooth tracking, extreme macro food photography, dramatic blue-and-warm-amber lighting, realistic dough and cheese physics, natural steam, oven heat distortion, detailed skin and hair, shallow depth of field, cinematic film grain, photorealistic, ultra-detailed, 4K HDR, 24fps, 16:9. Audio: Natural diegetic sounds only — dough slapping, flour scattering, sauce spreading, toppings landing, oven roar, bubbling cheese, crust crackling, basil flutter, and final bite crunch. No music, dialogue, subtitles, logos, text, or watermark.

CORE MESSAGE: "What if today, you could use the break to find your focus?" — a KitKat break creates a moment to pause, reset, and return to what matters with renewed focus. Feel like a premium global chocolate ad with emotional family storytelling — NOT a generic product commercial, NOT a fitness ad, NOT a children's cartoon. Warm, aspirational, elegant, believable. SCENE 1 — MORNING BEGINS (0:00–0:03) Luxury family home, soft morning light. Camera moves through the elegant kitchen (marble island, warm wood, pendant lighting, garden view). Mother (Ref 1) prepares breakfast in elegant cream/beige homewear; son (Ref 3) enters sleepily; father (Ref 2) preps coffee in background. Red KitKat (Ref 5) sits naturally on the island; camera pushes toward it — integrated, not a catalog shot. VO (warm female): "Every day brings a new challenge…" SCENE 2 — WAKE UP, CHAMPION (0:03–0:06) Mother wakes son, opens curtains. MOTHER: "Good morning, champion." BOY: "Five more minutes…" Natural, warm. Mother stays in casual homewear — no office clothes. SCENE 3 — BREAKFAST / KITKAT INTRO (0:06–0:09) Family at the island. Son notices the KitKat, picks it up, looks at it (doesn't eat yet — introduction moment), places it by his school bag. VO: "Sometimes, all you need is a moment…" SCENE 4 — READY FOR SCHOOL (0:09–0:12) Son dressed in navy blazer/uniform. Mother straightens his collar; father nods. KitKat visible in his open bag. He heads out the front door into daylight. VO: "...to reset." SCENE 5 — SCHOOL / THE CHALLENGE (0:12–0:16) Realistic classroom. Son looks focused but slightly overwhelmed at his desk, tries to work through a problem, briefly distracted. Natural classroom ambience, no music yet. SCENE 6 — THE KITKAT BREAK (0:16–0:20) (key product moment) Bell rings. Son takes a breath, pulls the exact KitKat (Ref 5) from his bag. Close-ups: opens wrapper, snaps off a finger, bites, pauses — visibly relaxes and refocuses. VO (slow, deliberate): "What if today, you could use the break to find your focus?" Music lifts gently. Only one bite — no repeat consumption. SCENE 7 — BACK TO FOCUS (0:20–0:22) Match cut to him at his desk, calm and writing confidently; teacher notices. VO: "Pause. Reset. Go again." SCENE 8 — FOOTBALL FIELD (0:22–0:26) Match cut to his foot kicking a ball. Realistic child-level play — receives, passes, shoots, scores; teammates celebrate. School bag with KitKat visible on the sideline (no second KitKat eaten — visual continuity only). VO: "And get back to what matters." Music turns energetic; football ambience. SCENE 9 — FATHER AND SON (0:26–0:28) Home, cream sofa. KitKat nearby on coffee table. FATHER: "Proud of you, kid." BOY: "Thanks, Dad." Warm, intimate two-shot. SCENE 10 — FINAL KITKAT HERO SHOT (0:28–0:30) Premium product shot on the marble island: unopened KitKat, a few wafer fingers, chocolate crumbs, warm natural light, home softly visible behind. VO: "Have a break, have a KitKat." ON-SCREEN TEXT: same line. End clean on product.

Create a 30-second authentic 1980s Japanese game-show broadcast sequence inspired by classic Japanese obstacle-course television. Realistic live-action footage, broadcast-camera aesthetics, natural handheld movement, analog film grain, slightly washed colors, chaotic energy, practical stunt physics, and no modern visual artifacts. No reference images provided; generate everything from this description. 0–5s — Wide Establishing Shot: A young Japanese man in his late 20s, wearing a red tracksuit and white safety helmet, sprints onto a massive horizontal rotating log above a deep muddy pit. The log spins continuously beneath him while several giant padded wrecking balls swing across the path. The camera tracks him with imperfect handheld broadcast movement. 5–10s — Medium Action Shot: The first wrecking ball swings toward him. He ducks desperately, rapidly shuffling his feet to maintain balance as the rotating log accelerates. His arms flail naturally, and muddy water splashes below where other contestants have fallen. 10–15s — Close Action Shot: A second padded wrecking ball strikes his shoulder, sending him spinning off balance. He drops onto his knees and desperately hugs the rotating log as his body slides around its wet wooden surface. Strong motion blur and realistic momentum. 15–20s — Dynamic Tracking Shot: He ends up hanging underneath the rotating log, gripping it with both hands while his legs kick through the air. He struggles to crawl toward the opposite platform as the log continuously rotates. His grip gradually weakens and his fingers slip. 20–25s — Major Impact: He finally loses his grip and falls into the deep muddy pit. He hits the water with a huge, physically realistic explosion of brown mud and water. The camera reacts with chaotic handheld movement as the splash briefly fills the frame. 25–30s — Comedic Ending Shot: He resurfaces completely covered in mud, helmet crooked and face stunned. He slowly raises one defeated fist toward the camera while the rotating log continues spinning behind him and a wrecking ball swings overhead. Maintain the authentic chaotic game-show atmosphere. Visual requirements: High-fidelity realistic live action, accurate rotating-log physics, believable human movement and impacts, realistic mud and water simulation, practical-stunt appearance, coherent character identity throughout, 1980s Japanese television broadcast grain, subtle analog noise, imperfect exposure, natural motion blur, period-accurate camera movement, no CGI-looking surfaces, no modern cameras, no futuristic elements, no text overlays, no subtitles.

Ultra-realistic summer café scene with subtle frozen-time VFX. One continuous 20-second shot. Natural handheld camera with a smooth backward track and one gentle side move. Bright daylight, warm terrace atmosphere, realistic shadows, soft film grain, no dramatic action styling. One adult man seated at an outdoor café table, wearing casual summer clothes. One adult waitress carrying a glass of beer on a tray. A few adult customers remain in the background. Sunny café terrace with small tables, umbrellas, chairs and a quiet city street. Photorealistic pause-time effect, simple body motion, stable anatomy, realistic glass handling, coherent background extras, natural summer atmosphere, seamless camera continuity, no fall, no injury, no direct body repositioning. 0-5s: Medium-wide tracking shot - The man waits calmly at his table. The waitress approaches with a cold beer on a tray. 5-8s: Gentle push-in - Her foot catches slightly on the edge of a chair. She loses balance for a moment. The beer glass begins sliding off the tray. 8-11s: Close-up - The man snaps his fingers. Time pauses instantly. The waitress becomes motionless in a tilted but stable pose. The beer glass and a few droplets remain suspended close to the tray. 11-16s: Slow side move - The man stands, calmly takes the suspended beer glass, places it on his table, then moves the chair slightly out of the waitress's path. He does not touch her. 16-20s: Push-in to medium close-up - He sits back down and snaps again. Time resumes. The waitress regains her footing naturally and looks confused when she sees the beer already on the table. The man smiles and says in English: "Perfect timing."

15 second vertical comedy clip, 9:16, photorealistic, shot like a real phone video that turns into a fast fashion montage. Cut driven, handheld, no music. MAYA. 27, olive skin, long chestnut brown wavy hair with sun lightened strands, dark brown eyes, gold hoops, layered gold necklaces, thin bracelets. Wears a fitted ivory sleeveless square neck top, high waisted colorful print resort trousers, thin tan belt, barefoot. Very expressive face. LENA. 28, tall, golden skin, long dark espresso brown glossy waves, hazel brown eyes, gold earrings, calm and confident. Same face, same hair, same body, same age in every shot. Only her clothes change. SETTING: a white motor yacht anchored in a calm turquoise Mediterranean bay. White sailboats in the distance, low rocky green coastline on the same side the whole time, cloudless blue sky, bright late morning sun from the same direction in every shot. White deck, steel railings, cream cushions, coiled ropes. Light sea breeze moves hair and fabric. STYLE: natural sunlight, real skin texture with pores, correct hands and fingers, real fabric movement, clear turquoise water, no beauty filter, no plastic skin, no heavy HDR. CAMERA: handheld with small natural shake. Fast whip pans and hard match cuts between outfits. Normal lens, no slow motion, no floating drone moves, no heavy blur. 0.0 to 1.6s Maya stands on the deck, medium close up, sea behind her. She stares off to the right, eyebrows up, mouth open, and leans forward. Maya says, flat and shocked: {No. Absolutely not.} Do not show what she is looking at. 1.6 to 3.0s Camera pushes in. Maya covers her mouth, then drops her hand and points hard to the right. Maya says: {You said ONE outfit!} On the word "outfit" the camera whip pans right along her arm. 3.0 to 4.6s The pan lands on Lena at the bow, wearing an oversized open white linen shirt over a black one piece swimsuit, black sunglasses, flat sandals, big woven tote. Beside her sit one large suitcase, a garment bag and two tote bags. Medium full shot, railing leading toward her. Lena says calmly: {It is one outfit.} Small pause, she lifts one finger: {For now.} 4.6 to 6.0s Hard match cut, same body position. Lena now wears a cobalt blue swimsuit with a white sheer sarong tied at her waist, hair damp at the temples, barefoot. She takes two steps toward camera and points at the water. Lena says: {Swim.} 6.0 to 7.4s A white towel sweeps past the lens, and behind it Lena now sits at a small deck table with fruit and sparkling water, wearing a coral orange halter dress, sunglasses on her head, woven handbag. She lifts the sunglasses slightly and smiles. Lena says: {Lunch.} 7.4 to 8.8s Hard match cut. Lena stands at the railing in an emerald green satin slip dress with gold statement earrings, same daylight. Wind lifts her hair, she turns her head toward Maya. Lena says: {Sunset.} 8.8 to 10.3s Fast whip pan across the cabin. Lena walks past camera in an ivory cropped blazer, matching wide leg trousers, champagne silk camisole, gold heels, small metallic clutch. Camera tracks back one second. Lena says: {Dinner.} 10.3 to 11.8s Hard match cut. Medium close up of Lena in a silver metallic evening mini dress with crystal earrings. She adjusts one earring, completely serious. Lena says: {Emergency.} 11.8 to 13.2s Smash cut to a tight close up of Maya, honestly confused, looking toward Lena. Maya asks: {What emergency?} 13.2 to 15.0s Cut back to Lena, medium shot, silver dress catching the sun, turquoise water, sailboats and green coast behind her. She slides her sunglasses on, gives a tiny smile to camera. Lena says: {Bad photos.} Maya laughs loudly off screen. Lena walks out of frame. Hard end, no fade. AUDIO: only real sound, meaning ocean, wind, boat movement, footsteps, fabric, bracelets, glass, laughter. Dialogue language: English, natural conversational delivery, never theatrical, never both women talking at once. No music at any point. No narration. No subtitles. Last frame: empty yacht deck at the bow, turquoise bay, sailboats and green coastline behind, bright sunlight, gentle water movement. No on screen text, no captions, no logo, no watermark. Negative prompt: music, score, subtitles, on screen text, watermarks, logos, brand names, extra people, two Lenas, two Mayas, changing face, changing hair, changing age, mixed or merged outfits, clothes morphing on the body without a cut, magical transformation, slow motion, drone shots, warped hands, extra fingers, plastic skin, beauty filter, oversaturated colors, moving coastline, changing sun direction, horizontal framing.

Use exactly 2 uploaded image assets. image1 = The adult female lead is the sole and highest-priority character identity reference. Strictly maintain her face, facial feature proportions, skin tone, hairstyle, body type, sense of age, overall temperament, clothing, and accessories. image2 = A Korean adult male wearing glasses is the sole and highest-priority character identity reference. Strictly maintain his face, facial feature proportions, skin tone, hairstyle, glasses, body type, sense of age, overall temperament, clothing, and accessories. Generate a 30-second, 16:9 landscape, 4K, 24fps, hyper-realistic live-action film-grade action comedy short. The scene is a high-rise modern top-floor office / luxury apartment-style space at night, with floor-to-ceiling windows overlooking the city nightscape. The interior includes an open living room, a long hallway, modern furniture, and mixed warm-cool lighting. No anime feel, no game CG feel, no cheap VFX, no slow motion, no pseudo slow motion. Action must be realistic, fast, clear, and cinematic. 【Premise】 This is a pre-choreographed, mutually consensual action comedy performance between two adult actors. The overall tone is light, playful, and competitive — no fear, threat, or coercion. It is not a life-or-death fight, not a vulgar skit, but a high-energy action comedy where "the female lead keeps attacking, and the male lead keeps dodging while disrupting her rhythm with playful cheek kisses." 【Core Dynamic】 The female lead is on the offensive the entire time. The male lead almost never truly fights back — he only dodges, repositions, moves in close, and backs away. His main "scoring method" is not punches or kicks, but using momentary gaps in the action rhythm to suddenly step in, deliver a quick, clear, playful cheek kiss, then immediately pull back. There are 5 cheek kisses total. Each one must be a clear comedic beat. After each kiss, the female lead becomes more annoyed and more serious, continuing the chase. 【Action Principles】 Female lead: Continuously presses forward, using jabs, hooks, elbow strikes, knee strikes, side kicks, high kicks, spinning kicks, turning attacks, and relentless pursuit. The further it goes, the more serious and aggressive she becomes. Male lead: Almost never throws a punch, never delivers real kick or punch counters, no heavy throws or slams. His main movements are only: dodging, retreating, sidestepping, ducking, leaning back, gliding, repositioning, redirecting attack paths, suddenly stepping in, cheek kissing, and immediately backing off. The structure must keep escalating repeatedly: Female lead attacks continuously → male lead dodges continuously → male lead seizes a tiny gap and suddenly delivers a cheek kiss → female lead gets angrier → she attacks again even more fiercely. This structure repeats 5 times. 【Camera Rules】 Action segments primarily use medium shots, medium-wide shots, wide shots, close follow cam, handheld feel, lateral tracking, and slight breathing-like camera shake, clearly conveying spatial movement. Before each cheek kiss, the camera first follows the rhythm of both performers' movements, capturing the male lead suddenly entering the female lead's close personal space. When each kiss happens, do not cut away suddenly, do not switch camera angles. Must use a fast, smooth camera push-in / dolly-in within the same continuous shot: Action medium shot → male lead suddenly steps in → camera rapidly pushes in to medium close-up / close-up → clearly showing the brief cheek kiss, the male lead's slightly smug expression, and the female lead's momentary shock or irritation → camera naturally pulls back as the two separate and continues following the chase. Each kiss lasts approximately 0.3–0.6 seconds, must be clearly visible but very brief. No stopping to strike a pose. 【Male Lead Expression & Personality】 The male lead maintains a composed, playful, slightly cheeky prankster vibe throughout. He is not angry and does not want to actually overpower the female lead; his enjoyment comes from continuously avoiding her attacks and then seizing a momentary opening to deliver a cheek kiss. After each successful kiss, he must show a very brief, natural, slightly smug expression: a subtle upward curl of the mouth, eyes carrying a hint of "you still haven't hit me" playfulness, occasionally a light eyebrow raise. Not lecherous, not sinister, no exaggerated villainous grin. First time: slightly pleased with himself. Second time: more obviously amused. Third time: starting to deliberately provoke. Fourth time: clearly knows the female lead is getting angrier, but still can't help showing a cheeky grin. Fifth time: after the kiss, clearly shows a guilty "I went too far" smile, then immediately runs away. 【Dialogue Rules】 The film has very minimal dialogue — only the following 4 Korean lines are allowed. Do not add any other dialogue, narration, or subtitles. After the first cheek kiss, the female lead says briefly, startled and annoyed: "야!" After the second cheek kiss, the male lead says while casually backing away with a playful tone: "또 실패." After the fourth cheek kiss, the female lead grits her teeth and suppresses her anger, saying: "너 진짜..." She doesn't finish the sentence before resuming her attack. After the fifth cheek kiss succeeds, the male lead immediately turns and runs away, shouting in a not-very-sincere casual tone: "미안!" The third cheek kiss has no dialogue at all — only a clear kiss sound, the female lead's expression, and the immediate chase create the comedic beat. All lines must strictly remain in the original Korean text above. No Chinese dialogue, no English dialogue, no auto-translation. 【Timeline】 0–4 seconds Action starts within the first 1 second. The female lead is already attacking relentlessly in the high-rise living room area, fast and sharp: jabs, spinning elbow strikes, mid-level kicks in continuous succession. The male lead barely throws any offense — he only sidesteps, ducks, glides, and leans back to dodge, moving very calmly. Camera follows closely, establishing the basic dynamic of "female lead keeps attacking, male lead keeps dodging." 4–6 seconds | First cheek kiss The female lead continues pressing forward, throwing a punch then a kick to close in on the male lead. After dodging, the male lead briefly steps in from roughly a 45-degree angle behind the female lead's side — not directly behind her. He then quickly delivers the first brief kiss on her right cheek. Must rapidly push in to a close-up of the face within the same shot, clearly showing the kiss. The female lead freezes for a moment, then says irritably: "야!" The male lead immediately backs off, showing his first smug, cheeky expression. 6–10 seconds The female lead is visibly angrier, raising her attack tempo. She uses low sweeps, high kicks, hooks, and spinning pursuits. The male lead keeps dodging, not counter-attacking, only retreating and evading, occasionally using very brief redirections to change the direction of her attacks, but never launching a real counter. 10–12 seconds | Second cheek kiss The female lead throws a high kick that the male lead barely dodges by ducking low. Seizing the half-beat as her leg comes down, he cuts in from the other side and delivers another kiss on her left cheek. Must again rapidly push in to a close-up of the face within the same shot, clearly showing the moment. After the kiss, he casually backs away saying: "또 실패." His expression is more obviously amused. The female lead gets angrier and immediately charges. 12–16 seconds The female lead begins chasing more fiercely, with bigger, faster, more aggressive movements. She continuously presses forward, spins, throws high kicks, chasing the male lead from the living room to the glass window and then turning toward the open walkway. The male lead almost only performs near-miss dodges: leaning back, slipping around close, ducking low, sidestepping away. The action must feel like "one more moment of hesitation and he would get hit." 16–18 seconds | Third cheek kiss After the female lead's two consecutive attacks miss, the male lead suddenly steps in at the instant she turns her head, delivering another shorter, more sudden, crisper cheek kiss. Still must rapidly push in to a close-up of the face within the same continuous shot, clearly showing the action. No dialogue here — only a brief natural "smack" sound and the female lead's expression. The male lead's expression now carries deliberate provocation and cheekiness. The female lead's eyes clearly turn fiercer, and she resumes the chase. 18–22 seconds The female lead's anger continues to build, and the pursuit becomes more vicious. She uses high kicks, spinning crescent kicks, and knee strikes to close in, almost without pausing. The male lead dodges throughout, retreating into the hallway, weaving around furniture edges, never counter-attacking — only reading her movements and narrowly avoiding them. The male lead's composure begins to carry a hint of comedic panic from being chased, but he still dodges very skillfully. 22–24 seconds | Fourth cheek kiss One of the female lead's attacks grazes past the male lead's face. Using the momentum of her attack, he suddenly steps in close and delivers the fourth cheek kiss. The camera again rapidly pushes in to a close-up within the same shot, clearly showing the action. The male lead still can't help showing a cheeky grin. The female lead grits her teeth, suppressing her anger, and says: "너 진짜..." Before she finishes the sentence, she immediately resumes the chase, not giving him a moment to breathe. 24–27 seconds The female lead enters her most intense pursuit state, with maximum speed and pressure. She throws high kicks and spinning attacks almost without pause, driving the male lead deeper into the hallway. The male lead still does not counter — only continuously retreating at the last moment, sidestepping, ducking, leaning back, like he's dodging for his life. 27–29 seconds | Fifth cheek kiss In the fastest chase and exchange of the entire film, the male lead looks like he's about to get kicked, but dodges at the very last instant, then suddenly and very quickly returns to the female lead's close range, completing the fifth — and most exaggerated, most cheeky — cheek kiss. Must use the clearest continuous push-in facial close-up, clearly showing the action. After the kiss, the male lead shows a guilty "I went too far" smile, then immediately turns and runs away, shouting with a hint of guilt and feigned casualness: "미안!" 29–30 seconds The male lead immediately flees deeper into the hallway. The female lead chases after him without hesitation, still ready to continue attacking. The camera follows both of them rushing out at high speed, and CUT TO BLACK on the clear sense that "this chaos isn't over yet." 【Performance Focus】 Female lead: On the offensive the entire time. Brief shock at the first kiss; increasingly irritated afterward; by the latter half she's essentially chasing him down. Her movements cannot be soft — they must carry real aggression. Male lead: Never truly counters — only dodges, steps in close, kisses once, then backs off. His expression goes from composed and playful to carrying a funny "I might have overdone it" look later, while still maintaining an expert-level ease of movement. Both must move like people who actually know how to fight — realistic movement, realistic momentum, realistic breathing, realistic fabric movement. 【Environment & Physics】 The space must be realistically usable: floor-to-ceiling glass, living room, open floor, long hallway, city nightscape. Light collisions with furniture edges, sudden stops and turns, gliding steps, and weaving around objects are allowed, but no large-scale destruction. No weapons, no guns, no bloody injuries. Action sound effects must be realistic: footsteps, air whooshes, fabric friction, breathing, spatial reverb. 【Sound】 No weird BGM, no piano-style scoring. An extremely subtle low rhythmic ambient bed is acceptable, but action and performance take priority. Key sound effects: footsteps, breathing, punch/kick air whooshes, fabric friction during close dodges, spatial reverb. Each cheek kiss must have a natural brief "smack" sound. The beat of stillness after each kiss must be clear, but no slow motion. 【Strict Prohibitions】 Do not portray the male lead as actively counter-attacking; Do not have the male lead throw consecutive punches or consecutive kicks; Do not shoot the action as mutual fighting; No prolonged standoffs; No slow motion; No animated feel; No game CG feel; No weapons; No blood; No extra characters; No subtitles; No watermarks; No daytime scenes; No old, worn-down, dilapidated scenes; No vulgar intimate scenes; No romantic long kisses; No mouth-to-mouth kissing; No kissing hair; No kissing ears; No kissing air; No facial clipping/merging; No character identity drift; No face distortion; No deformed limbs; No unclear action; Do not shoot the whole piece as a pure romance scene; Do not shoot the whole piece as a slapstick farce — the action itself must remain fast, professional, and cinematic.