Cinematic AI Video Generation with Seedance 2.5

A cheerful Korean girl with a high ponytail, wearing a fitted white crop T-shirt and blue jeans, records a casual lifestyle vlog in English using her smartphone in selfie mode. She starts at home, tying her ponytail, doing light natural makeup, checking her outfit in the mirror, grabbing her bag and saying, "Good morning, everyone! Come spend the day with me!" She leaves her apartment and walks through lively Seoul streets, smiling at the camera while chatting naturally: "The weather is amazing today, so I thought I'd take you guys with me." As she continues walking, she points out cafés, shops, and interesting streets while speaking fluent English in a relaxed, conversational tone. She laughs, jokes with the audience, and occasionally waves at the camera. After a few minutes, she notices several people curiously watching her film. She becomes slightly embarrassed, laughs naturally, covers her mouth for a second, and says, "Okay... I think everyone is wondering who I'm talking to!" She giggles, shrugs, then confidently continues the vlog: "It's a little awkward, but that's part of being a vlogger, right? Let's keep going!" Natural handheld smartphone footage, realistic selfie perspective, authentic English lip sync, spontaneous expressions, subtle camera shake, slight motion blur, real city ambience, everyday lifestyle creator aesthetic, realistic skin texture, neutral iPhone color science, documentary-style realism, social media vlog quality, 9:1 6 vertical format.

OPERATION: Helicopter refueling stop at offshore oil platform. DECK CREW MEMBER: Male, 35. Fire-retardant coveralls. Life vest. Ear defenders. Alert, methodical safety behavior. SHIFT TIME: 1:30 PM. Midday operational window. WORK ZONE: Offshore platform helideck. Parked helicopter with rotors idling. Refueling hose station. Wind sock visible. Open ocean horizon in background. CAMERA STYLE: Shoulder-level documentary following the deck crew member. Handheld stability with wind movement visible. Occasional wide shots showing ocean scale. Zooms only during fuel connection moments. WORKFLOW: Helicopter lands and rotors set to idle. Deck crew approaches with safety clearance signal. Refueling hose connected and fuel transfer begins. Visual inspection of connections during fueling. Hose disconnected, helicopter cleared for takeoff. AMBIENT AUDIO: Helicopter rotor idle hum. Ocean wind gusts. Fuel pump mechanical sounds. Radio safety communications. Distant wave sounds. REALISM NOTES: Strict offshore aviation safety protocol. No one acknowledges the camera. Crew maintains constant situational awareness throughout.

@image1(完全按照该参考图呈现,不进行任何重新设计)年轻女性,黑色长发,白色衬衫外搭黑色背心,牛仔短裤 场景:现代化室内空间,一面巨大落地窗正对着晴朗的城市天际线,午后阳光大片洒入室内,窗外蓝天白云清晰可见。 画面风格:电影级写实质感,慢动作玻璃破碎特效(风格化处理,强调美感而非危险血腥),高对比度光影,粒子特效浓郁,杜绝任何血腥或写实伤害画面。 总时长:15秒 | 画面比例:16:9横屏 | 分辨率:4K | 帧率:24帧 分镜脚本(每2秒切换一次运镜): 0-2秒 — 静谧日常(平视中景镜头): 镜头平视拍摄,她静静站在巨大落地窗前,阳光洒满全身,神情放松地望向窗外晴朗的城市天际线,一切祥和明亮。 2-3秒 — 震动前兆(正面反应特写镜头): 镜头切至正面特写,她脚下地面开始传来细微震动,身旁悬挂的窗帘轻轻晃动,她眉头微蹙,神情从放松转为警觉。 对白(女声,疑惑): 「あれ、地震?」 3-4秒 — 剧烈摇晃(手持晃动镜头,声音锚点): 一阵剧烈的地面震动骤然袭来,伴随低沉的隆隆声响——镜头手持剧烈晃动,她惊慌地伸手扶住墙壁试图站稳,窗户玻璃开始出现细密裂纹。 对白(女声,惊呼): 「きゃっ!」 4-5.5秒 — 玻璃碎裂启动(超慢动作特写镜头): 镜头切至极致慢动作,玻璃裂纹以风格化的美感迅速蔓延扩散,如蛛网般覆盖整片落地窗,她惊恐地转身后退,长发因动作甩起。 5.5-7秒 — 碎片飞舞(360°环绕慢动作镜头): 镜头以360度环绕她展开的慢动作,巨大的玻璃瞬间爆裂成千万晶莹碎片,每一片碎玻璃在阳光下折射出彩虹般的光芒,如钻石雨般悬浮飞散于空气中,她惊讶地抬手护住面部。 7-9秒 — 碎片悬浮凝视(仰角特写镜头): 镜头仰角特写她从指缝间惊愕地望着眼前这片美丽而危险的碎片风暴,阳光穿透飞舞的玻璃碎片形成无数细碎光斑洒在她脸上,发丝因气流飘扬。 对白(女声,惊叹又害怕): 「うそ…」 9-11秒 — 碎片穿越瞬间(慢动作侧向追踪镜头): 镜头侧向追踪捕捉数片玻璃碎片以慢动作从她身侧擦身而过却奇迹般未曾伤及她分毫(风格化处理),她下意识闭眼侧身,气流吹动衣摆。 11-12.5秒 — 尘埃落定(俯拍广角镜头): 镜头俯拍拉远,玻璃碎片如雪花般缓缓飘落满地,阳光透过破碎的窗框缺口洒入室内形成斑驳光影,她惊魂未定地站立在碎片之中。 12.5-14秒 — 惊魂回神(正面缓慢推近镜头): 镜头缓慢推近她的正脸,她大口喘息,望向破碎的窗户与满地晶莹碎片,眼神从恐惧逐渐转为庆幸的复杂神情。 对白(女声,松了口气): 「大丈夫…生きてる。」 14-15秒 — FREEZE FRAME(定格收尾): 镜头拉远至广角,她站立在阳光洒落的破碎窗前,满地碎玻璃反射着午后阳光如星辰般闪烁——画面在此刻凝固定格,渐暗淡出。 运镜手法:平视中景、正面反应特写、手持晃动、超慢动作特写、360°环绕慢动作、仰角特写、慢动作侧向追踪、俯拍广角、正面缓慢推近、freeze frame定格——每1.5-2秒切换一次运镜,地震发生后运镜节奏刻意放缓强化慢动作美感。 特效强化:窗帘晃动前兆、玻璃裂纹蛛网状蔓延、碎片阳光下彩虹反光、碎片如钻石雨悬浮飞散、气流吹动发丝衣摆、破碎窗框斑驳光影、满地碎玻璃星辰般闪烁。 音效设计:0-2秒平静室内环境音→2-3秒细微震动隆隆声渐强→3-4秒剧烈地震声(声音锚点)→4-9秒玻璃碎裂清脆声效延展拉长(慢动作音效处理)→9-12秒气流呼啸声→12-15秒渐归平静的喘息声与环境音。 负面提示词:禁止任何血腥、割伤或写实伤害画面、禁止玻璃碎片显得危险致命、禁止角色比例前后不一致、禁止连续重复同一运镜角度、禁止freeze frame画面失焦模糊。

严格按参考分镜顺序制作，不要把整张网格当成一张图。每个面板都是独立镜头，生成一段15秒、手持手机自拍感的韩国情侣旅行 Vlog。 男生长相参考@@@，女生长相参考@@@。全程严格保持两人的脸、发型、穿搭、身材比例和情侣关系一致，不换脸、不变人。 视频像旧 iPhone 随手拍下的私人旅行记录：轻微晃动、偶尔失焦、自然变焦、运动模糊、镜头拉丝、曝光变化和不完美构图。不要广告感、电影感或过度稳定。 内容按旅程自然推进：出发、白天游玩、吃东西、牵手散步、互相拍照、打闹、看风景，最后在夜晚街道、车内或房间温馨收尾。 甜蜜感要自然，不刻意摆拍。加入回头等对方、拉衣袖、分享食物、整理头发、靠肩、牵手、笑着躲镜头、互相看对方等细节，让人看完很羡慕。 可穿插路人、街景、车窗、食物、手部、背影、侧脸、影子和环境细节。人物不要一直看镜头，要有真实聊天、笑声和自然互动。 整体质感：低对比度、轻微褪色、自然光、真实噪点、旧手机画质、温柔韩剧情侣日常氛围。 BGM为温馨轻快的韩剧恋爱 Vlog 风格，保留风声、脚步声、街道声、海浪声、餐具声和自然笑声。 禁止字幕、文字、水印、Logo、拼贴边框、静态照片轮播、夸张亲密动作、机械摆拍、CGI和商业旅拍感。

Create a premium cinematic character bible sheet for a Wing Chun master named IP MAN. Use the uploaded character sheet as the strict visual reference for his exact face, body, proportions, hair, skin tone and outfit. Do not change his appearance. LEFT SIDE: Massive bold brushstroke painted text IP MAN top left corner, rough expressive hand-painted brush font in deep black with subtle red accent. Below it small text: WING CHUN MASTER / THE SILENT FIST. One large full height dynamic hero pose of Ip Man directly below — standing in classic Wing Chun ready stance, one hand forward open palm, other hand pulled back at hip, completely composed, black changshan flowing slightly, traditional Chinese ink wash splash behind him, deep blacks and subtle reds radiating outward fading into white, no box no frame, standing directly on the white page. CENTER: Section labeled TURNAROUND in small clean text. Three full body views: Front, Side, Back. Same character same outfit same proportions. Below: TRAINING EQUIPMENT with clean illustration of the Muk Yan Jong wooden dummy — tall post, three arms, diagonal leg, wooden cross base. RIGHT SIDE: Section labeled DETAILS. Grid of four clean close-up panels: Face front calm and composed, Face side profile, Eyes authoritative and focused, Hands in Wing Chun stance close-up showing finger position. Below: COLOR PALETTE — black, white, dark wood brown, stone grey. Tagline bottom right: SIFU. OVERALL: Clean white background, black and red ink wash behind hero pose, minimal clean typography, cinematic cel-shaded 3D anime, semi-realistic CGI, hand-painted textures, digital painting feel, hard-edge brushwork, not cartoon not Disney not Pixar, print ready.

SCENE CONTEXT In a frozen post-apocalyptic wasteland, an armored snow-truck races down a white highway, fleeing a massive blizzard-avalanche rolling in behind it. Inside the churning wall of snow, the colossal face of an ice monster forms and dissolves. The shot opens on a vast white vista where the truck is a tiny speck, then pushes in one continuous move to the cab, revealing the male driver and a woman in the passenger seat. Grounded, photoreal, high-budget fantasy realism — every surface, particle, and performance reads as captured live, not generated. LOCATION MAP Wide frozen plateau. A snow-packed highway cuts diagonally from deep background (upper-left, x 30%, y 38%) toward foreground (lower-right). Flat tundra and dead, ice-crusted utility poles flank the road. Behind the truck, north horizon, a 200-meter-tall wall of churning white blizzard/avalanche advances. Weak sun glow hidden behind the storm wall acts as backlight from screen-rear. Lighting direction: diffuse top-down with cold backlit haze from behind the truck. FIRST FRAME AND SPATIAL BLOCKING First visible frame is a super-wide aerial-height vista. The snow-truck sits deep background, center-frame, x 50%, y 46%, scale tiny (under 4% of frame width), driving toward lower-right and toward camera along the diagonal road. The blizzard wall fills the background behind it, x 0% to 100%, upper half of frame, advancing. No empty plate — the truck is present and moving in frame one. As the camera pushes in, the truck grows to fill frame; by the end the cab and windshield dominate, x 50%, y 50%, large scale, driver (screen-left seat) and woman (screen-right seat) clearly visible through the glass. Driver: torso faces forward down the road, hands gripping the wheel, eyes flicking to the side mirror toward the storm. Woman: torso forward, head turned back over her shoulder toward the storm wall, then snapping front. Both bodies face direction of travel (screen-lower-right). FORMAT MODE Single continuous take. Real-time motion. No cuts. No subtitles, no music. OPTICS One continuous push-in shot with a prime-lens optical character — clean, sharp center field, gentle natural falloff, no zoom-breathing, fixed-focal clarity. Begin on a 107° wide rectilinear character: immense environment spread to all frame edges, straight horizon, truck a tiny element in vast space, no fisheye curve. The camera then dollies forward continuously, the lens character settling toward an 84° classic wide as it closes on the cab — strong foreground presence, environment and storm still visible at frame edges, straight lines rectilinear, immersive close perspective on the windshield. One smooth optical move, no hard stop; perspective expands naturally as distance collapses. Prime-lens rendering throughout: crisp micro-contrast, organic shallow falloff on the cab at the end. CAMERA ARRI Alexa large-format cinema-camera character — wide tonal latitude, gentle highlight roll-off in the bright snow, clean shadow detail, true filmic color science. Camera starts at high aerial height looking down-forward over the road, far from subject, then descends and pushes in along the road axis, accelerating to match and slightly overtake the truck, ending at windshield height roughly 2 meters off the truck's front-left quarter. Movement is a powerful, weighted aerial-to-chase glide — fast forward velocity, slight motion-settle as it locks onto the cab, subtle real operator weight, no random jitter, no drone float. Focus rides the truck throughout; background storm stays slightly soft, truck razor-sharp by the end. ACTION TIMING 0.0s to 2.0s: Super-wide vista, tiny truck speeding down the diagonal road, blizzard wall looming and advancing, snow streaming off the road surface. 2.0s to 5.0s: Continuous push-in, truck grows, snow tracks throwing white spray, exhaust steam ripping off the stacks, the ice monster's face surfacing in the storm wall behind. 5.0s to 8.0s: Camera locks onto the front-left quarter and cab; through the windshield the driver grips the wheel and checks the mirror, the woman turns from looking back at the storm to facing forward, both tense, the truck roaring past camera-near. PERFORMANCE (Hollywood-grade acting) Naturalistic, restrained, high-stakes performances captured like seasoned A-list film actors. The driver's fear reads through micro-expression, not theatrics: tight jaw, a hard swallow, eyes darting between the road and the mirror, white-knuckle grip, controlled breathing. The woman plays escalating dread — a sharp glance back at the monster, pupils widening, a flinch, then forcing composure as she snaps front and braces a hand on the dash. Real eye-line continuity, genuine weight behind every reaction, no stiff or doll-like faces, no exaggerated mugging. Believable adrenaline and chemistry between two survivors who have done this before. PHYSICS Vehicle: heavy armored mass on massive studded rear snow-tracks and oversized spiked balloon front tires, a steel snowplow blade angled at the nose. Throws continuous white snow-spray and powder rooster-tails from the tracks; suspension compresses over road ridges, body rocks with inertia, exhaust stacks vent hot steam shearing backward in the wind. Snow particles travel in correct depth planes and accumulate on hood and cab roof. Storm: blizzard wall churns with turbulent rolling motion, snow and ice particles driven forward by wind, denser at the base. The monster face forms from compacted swirling snow — jagged ice teeth, hollow pale-glowing eye-sockets, frost-crystal features — coalescing and partly dissolving, never fully solid, made of moving snow and ice shards. Bodies: driver's arms carry the wheel's weight with road micro-corrections; the woman's hair and scarf lag and whip with the cab's motion. LIGHTING Cold overcast daylight, flat and diffuse, exposed for the bright snow field with Alexa-style highlight roll-off so the snow holds detail. Hidden sun glow behind the storm wall backlights the blizzard so the monster face reads as a brighter churning silhouette against the gloom. Truck surfaces catch cool top light with hard cold speculars on wet metal and ice; cab interior falls into cooler shadow, faces revealed by soft windshield bounce and dash glow. Palette stays white, steel-grey, pale cyan — no warm key, no studio fill. WARDROBE (Mad Max, snow-logical) Driver: scavenged cold-weather warlord gear — fur-lined battered leather greatcoat over a quilted thermal parka, riveted metal pauldron on one shoulder, fingerless gloves over thermal liners, frost-fogged ski goggles pushed up on a fur-flap ushanka cap, respirator scarf at the neck, bone and scrap-metal ornaments. Weathered, ice-crusted, lived-in. Woman: hooded fur-collared shearling jacket cinched with leather harness straps and buckles, knitted balaclava pulled down, goggles around the neck, fingerless gloves, scavenged-metal arm guards. Practical, insulated, battle-worn — same scrap-and-fur aesthetic, never impractical for sub-zero cold. STYLE Large-scale fantasy-realism in the vein of George Miller's kinetic wasteland chase staging, Denis Villeneuve's monumental scale and cold restraint, Peter Jackson's tactile creature physicality, and Zack Snyder's heroic backlit silhouettes. Photoreal texture, real grain, true film color science, grounded physical cinema — epic scope that still feels captured on a real set. OUTPUT SETTINGS IMAX-grade 8K capture, high bitrate 80–120 Mbps, ARRI Alexa large-format sensor character, prime-lens rendering, real-time motion, single continuous take. POSITIVE LOCKS Truck present and moving in the very first frame; no empty establishing plate. One continuous push-in from super-wide vista to cab close-up; no cuts. The monster is made of churning snow and ice and stays inside the blizzard wall, never a solid creature on open ground. Driver remains screen-left, woman screen-right inside the cab. Cold white/grey/cyan palette only; faces stay in cool shadow, no warm front fill. Performances stay photoreal and restrained; no stiff or artificial faces.

[CINEMATIC SETUP] Shot on Arri Alexa 65, 35mm lens. Grounded, hyper-realistic cinematic daylight setting in a bustling city center. Overcast natural lighting, muted realistic color grade. No video game effects or glowing lights; super-speed is represented purely through physical temporal distortion and heat waves. Audio style: immersive spatial sound, bustling city ambiance instantly cutting to deep vacuum sub-bass and tearing wind. [IMAGE REFERENCES / LEGEND] @[Image 1](image_1) = [main character: 30-year-old man, movie-level realistic facial features, no deformation, stable throughout. Wearing a solid black t-shirt (the T-shirt with no logos on the back), black cap, blue oversized jeans, Nike Air Force 1] [TIMELINE SECOND BY SECOND] 0-3s: [Steadicam tracking medium shot] Pablo (@[Image 1](image_1)) walks casually down a busy, photorealistic city sidewalk at normal speed. [Physics] Pedestrians walk by naturally, the fabric of his original clothing moves subtly with the normal breeze. [Audio] Normal bustling city traffic, chatter, and regular footsteps. 3-5s: [Sudden fast push-in to medium close-up] Pablo stops suddenly, looks straight ahead, and violently takes off. [Physics] A hyper-realistic sonic boom distorts the air around him. Extreme heat waves and temporal distortion follow his trajectory. [Audio] Sudden explosive sonic crack, instantly transitioning into a low-vacuum hum; all city noise vanishes completely. 5-8s: [Ultra-high-speed tracking shot] The world freezes entirely. Pedestrians and cars are suspended perfectly mid-motion. Pablo sprints down the middle of the avenue at impossible speed. [Physics] His exact original clothing ripples aggressively from extreme wind velocity. The sheer force of his passing flips suspended newspapers and shatters a nearby car window in hyper-slow motion. [Audio] Heavy, rapid footsteps echoing in isolation, violent tearing wind resistance. 8-11s: [Dynamic low-angle orbit] Pablo rounds a corner, his sneakers digging into the concrete. [Physics] Asphalt cracks and explodes upward from the friction, but the debris hangs completely frozen in the air. Real-world rigid body physics apply to the destruction. [Audio] Sharp concrete crunching, deep sub-bass frequencies. 11-15s: [Slow down to wide shot] He skids to a halt, the temporal air distortion fading away. Time snaps back to normal speed instantly. [Physics] The suspended shattered glass and concrete debris crash to the ground realistically. Pedestrians jump in shock at the sudden noise. Pablo resumes walking normally. [Audio] Instant rush of loud city traffic returning, debris hitting the ground heavily. [STYLE & QUALITY BOOSTERS] Photorealistic 8K, Arri Alexa 65 footage, ultra-detailed textures, cinematic lighting, hyper-realistic rigid body destruction, no glowing effects, grounded VFX, coherent multi-character scene, movie-level realistic facial features, strict adherence to original clothing, stable throughout. [NEGATIVES] video game style, CGI look, glowing trails, colored lightning, deformed faces, blurry motion on main character, changing clothes, low resolution, text artifacts, inconsistent lighting, watermark, extra limbs.

A detailed, 3D Pixar-style animation featuring an exaggerated, baby-faced caricature of football star Erling Haaland with a prominent blonde mullet and chubby, rosy cheeks. Wearing a bright red Norway national team jersey with white and blue trim, he confidently strides out of a dark stadium tunnel onto a vibrant, sunlit pitch. Surrounded by a roaring crowd and flashing cameras from press photographers, he looks directly into the lens, making comical, dramatic pouty faces and flashing a two-finger peace sign. The scene cuts to a high-energy match where he dynamically sprints down the field in shiny golden cleats, tightly controlling a star-patterned soccer ball while swiftly dodging defenders under bright stadium floodlights amidst falling confetti.

Cinematic comedic stadium sequence, 16:9, 15 seconds, FIFA World Cup 2026 atmosphere, no on screen text, no logos, no watermarks, no scoreboard graphics. [0:00-0:02] Drone aerial establishing shot, quick descent over a packed World Cup stadium at dusk, stadium lights blazing, two distinct fan blocks visible yellow-green Brazil jerseys on one side, red and blue Norway jerseys on the other. Camera glides forward toward the Norway section. [0:02-0:05] Slow dolly in, ground level, on a lone fan in front of the Norway supporters' block holding a large drum, striking it twice, slow and deliberate, dust and confetti particles drifting in the floodlight beams behind him. [0:05-0:09] Wide tracking shot along the stands: thousands of Norway fans sit down in perfect unison, forming tight rows like a longship's benches, arms drawing back in unison as if pulling oars through water, rhythm building in speed with each beat. [0:09-0:12] GoPro POV shot from inside the crowd, low angle looking up along the rowing motion, fast rhythmic arm-pulls, jerseys rippling, joyful chanting faces in unison, camera subtly shaking with the crowd's rhythm. [0:12-0:15] Cut to the Brazil supporters' section, quick dolly-in, fans standing frozen and stunned, a few holding their heads in disbelief, contrasting stillness against the roaring motion blurred Norway section still rowing behind them. Final beat: rack focus from one dejected Brazil fan's face to the blurred chanting Norway crowd, holding for a comedic freeze beat. Technical requirements: Anatomically accurate crowd members throughout correct hand/finger count and natural arm articulation during rowing motion, no melted or extra limbs, no distorted faces in close up or wide crowd shots. Realistic cloth physics on jerseys and scarves during synchronized movement. Confetti, dust, and floodlight particle behavior must respond naturally to air movement and crowd motion, no floating/static particles. Warm stadium floodlight color grading with deep dusk blue sky, high contrast between lit crowd and shadowed upper tiers. Smooth, natural crowd choreography synchronized but not robotic, slight individual variation in timing. No on screen text, subtitles, scoreboards, sponsor logos, or watermarks anywhere in frame. Audio reactive drum beat pacing implied through motion timing (rowing speed increases with each cut).

SEGMENT 2 of 10 of one continuous music video — continues the morning block scene. @ Image1 is the performer — same identity: cornrow braids into long dark curls, silver chain-drop earring, gold pendant, glossy lips. @ Image2 is WARDROBE REFERENCE ONLY (ignore mannequin head): black cropped blazer, leopard tie, black bodysuit, denim ruffled micro skirt, black loafers — identical to segment 1. @ Audio1 is the master track — the only audio. COMPOSITION RULE — CRITICAL: NO centered frames. Lateral thirds, over-shoulder, foreground occlusion (car hoods, hydrants, crew shoulders), diagonal leading lines, asymmetric group blocking — crew clustered on one side, her on the other, negative space active. PRECISE LIPSYNC — TOP PRIORITY: 0.0–3.0 "...ring loud from the pavement all the way up through the vaults" / 3.0–5.5 "They clap in my comments, I'm the one they quote — you just type tough, I'm the one that wrote the code" / 5.5–6.5 "Sharp tongue" / 7.5–9.0 "cut through the back row" / 10.0–11.0 "Big shoes" / 12.0–13.5 "still make the map grow" / 13.5–15.0 "Whole crew — built out of backbone". Gaps: closed mouth. FILM CONTEXT: Act I, LOCATION A — same brownstone street, morning light, sedan visible in background passes of the frame. Same 3 girls, same outfits. Black girlie hustle vibe. Shot flow: 0–5.5s tracking shot from ACROSS the street, traffic and parked cars wiping the foreground: she walks the far sidewalk on the right third, girls trailing on the diagonal behind her; crash-zoom punch through a gap in traffic on "wrote the code", landing her face on a third. 5.5–9s "Sharp tongue... back row": tight over-shoulder past a girl's bamboo hoop earring (foreground bokeh, left edge) to her rapping in the right half; on the instrumental gap the crew hits a hip-cock freeze staggered in depth, nobody centered. 10–13.5s "Big shoes... map grow": ground-level frame — her loafers enter from the left, walk the bottom-third diagonal; tilt up along her body keeping her hugging the left edge, brownstone cornices cutting the sky on the right. 13.5–15s "Whole crew, built out of backbone": four abreast but framed OFF-CENTER — the group occupies the left two-thirds, empty street ahead of them as negative space; freeze mid-stride on "backbone". Continuity: same outfit, block, crew; gold pendant visible. Audio intent: @ Audio1 only; sneaker scuffs, traffic. Quality bar: editorial off-center cinematography, no AI gloss.

SEEDANCE 2.0 视频提示词 时长：15秒 画幅：9:16竖屏 参考：严格使用参考图中的成年女主角，保持清晰面部五官、发型、妆容、身材比例和整套穿搭完全一致，全程为同一人物，不更换服装。 整体风格： 高级网红时尚短片，冷艳、妩媚、自信，深灰摄影棚，黑色镜面地面，暗紫色与金色轮廓光，柔和美颜主光，肤质真实细腻，轻微胶片颗粒，节奏感强但动作自然。 SHOT 1｜0—5秒｜怼脸眼神展示 主体：参考图女主。 动作：开场就是女主的面部超近景。她先微微侧脸看向画面外，随后缓慢转动眼睛直视镜头，轻轻眨眼，略微挑眉，嘴角露出若有若无的妩媚笑意；一只手自然拨动耳边碎发，身体轻微靠近镜头，目光自信、有吸引力。 环境：背景虚化成柔和的暗紫色光斑。 摄影机：85mm人像镜头，眼睛清晰锐利，浅景深，镜头缓慢推近并小幅环绕面部约15度，捕捉眼神、高光和真实皮肤细节。 约束：面部不能变形，不要夸张嘟嘴，不做低俗暗示动作。 SHOT 2｜5—10秒｜拉远展示完整造型 动作：女主抬手从镜头前自然划过，形成流畅遮挡转场。镜头迅速而平稳地从面部拉远到全身，她踩着高跟鞋向前走两步，交叉步停下，一只手扶腰，轻轻转换重心；随后转成侧身姿态，展示整体身材比例与服装轮廓，再顺势转身，回头看向镜头。 摄影机：由85mm近景无缝衔接至35mm全身镜头，轻微低机位但保持端正高级，镜头跟随后退；动作停顿瞬间出现一次时尚摄影闪光灯效果。 声音：清晰的高跟鞋脚步声、相机快门声、低沉时尚电子节拍。 约束：始终完整保留参考图服装、丝袜、高跟鞋及其花纹细节；镜头展示整体造型，不长时间聚焦单一身体部位。 SHOT 3｜10—15秒｜环绕炫姿收尾 动作：女主连续完成正面、侧面、背面三个自然模特姿势，动作之间使用短促流畅的速度变化。摄影机围绕她旋转约270度，她的裙摆、发丝和耳边碎发产生真实轻微摆动；环绕结束时她快速转回正面，向镜头走近，微微俯身贴近镜头，轻轻挑眉并眨一下眼，露出俏皮又妩媚的半笑，最后用指尖轻点镜头，画面瞬间切黑。 摄影机：50mm镜头稳定环绕，镜面地面产生柔和倒影；最后快速推近至面部特写，准确对焦双眼。 风格：高级、冷艳、妩媚、有网红爆款展示感，动作丰富但不杂乱，人物始终拥有镜头掌控力。 负面限制： 禁止人物身份变化、换脸、服装变化、蕾丝花纹漂移、丝袜接缝移动、身材忽胖忽瘦、比例夸张、肢体扭曲、额外手脚、多余手指、手指融合、五官融化、塑料皮肤、过度磨皮、走路滑步、高跟鞋穿模、衣物异常透明、走光、低俗姿势、私密部位特写、裙底机位、镜头角度漂移、人物瞬移、背景突然变形、过度慢动作、闪烁、字幕、文字、Logo、水印。

A photorealistic cinematic scene of a muscular grey bulldog chained to a tree in a lush green backyard on a sunny day A sleek grey cat hits the bulldog with a red metal tray shattering its teeth into white shards The cat sweeps the shards onto the tray and mocks the furious dog The bulldog lunges violently but a thick rope violently snaps him back to the ground A tiny brown mouse appears and points at the dog laughing The cat dances just out of reach The bulldog leaps forward and grabs the cat squeezing tightly but the cat magically slips out of a fuzzy grey sweater leaving the empty sweater in the dogs paws The cat is now wearing a tight red onesie The bulldog freezes completely still mid lunge like a statue holding the empty sweater The camera zooms into a black circular iris transition opening up into a lavish dimly lit royal dining room with velvet drapes The grey cat is now lounging on an opulent red throne ringing a small golden handbell with a smug expression The tiny brown mouse struggles to walk while balancing a gigantic silver serving platter loaded with massive colorful domed food dishes on its head Cinematic lighting volumetric dust particles 8k resolution highly detailed fur textures shot on 35mm lens lively orchestral comedy music playing throughout Seedance 20 15 seconds 16 9 aspect ratio