STYLE: High-end theatrical anime film style, ultra-fluid animation. Bright luxury department store flooded with natural daylight from a massive glass ceiling. Rich vibrant colors, cinematic contrast. Highly reflective marble floors, shattered glass, floating dust, sparkling debris. Luxury storefront lighting, colorful ads and digital screens. No music, no dialogue, environmental sounds only. LENS & CAMERA: Every shot uses a different focal length, height and framing. Alternate ultra-wide, macro, telephoto, POV, overhead and low-angle shots. Avoid repetitive compositions. Natural handheld movement only during heavy impacts. GLOBAL CONSTRAINTS: Strict character consistency throughout — never change either character's face, hairstyle, outfit or weapon. Character A always on the left, Character B always on the right. Never reverse screen direction. Maintain the 180-degree rule. No subtitles, no logos, no UI. Marble floor stays glossy and reflective. Destroyed objects react with believable physics. SETTING: A luxurious abandoned multi-floor department store — glass ceiling, fashion boutiques, perfume counters, cosmetic displays, jewelry showcases, designer handbags, digital billboards, escalators, glass elevators, central atrium, luxury cafés, broken fountains. The mall is empty. TOTAL LENGTH: 10 Seconds 0.0–1.8s — Macro insert into 16mm ultra-wide. Character A's grip tightens on both rapiers, heel twisting on polished marble. She launches forward at incredible speed, sliding across the glossy floor. Perfume bottles explode into colorful mist, display shelves collapse behind her. 1.8–3.3s — Medium shot. Character A unleashes a rapid dual-blade combo. Character B blocks with the chain whip, wraps one rapier and throws Character A through a luxury handbag display. Bags fly everywhere, glass cabinets explode. 3.3–4.6s — POV from Character A. Character B aggressively swings the chain whip, smashing nearby cosmetic counters. Lipsticks, mirrors and powders burst into the air. 4.6–5.3s — Macro insert. Character A notices the nearby escalator moving as emergency power briefly returns, its metal steps climbing upward. 5.3–7.0s — Character A jumps onto the moving escalator, Character B chases immediately. Both sprint upward exchanging attacks. The chain whip tears through hanging banners, broken lights rain down. 7.0–8.6s — Wide tracking shot. Character A leaps from the top of the escalator onto the second-floor glass skybridge. Character B follows without hesitation. The bridge shakes violently beneath their feet. 8.6–10.0s — Slow motion. Both fighters charge toward each other. Character B swings the chain whip; Character A slips beneath it and spins between floating glass fragments, sunlight shining through thousands of suspended shards.
Retro Anime Samurai Duel
A retro cinematic anime video showcases a legendary samurai duel in the style of 1990s hand-painted cel animation. The scene is imbued with an old manga aesthetic, creating a nostalgic and dramatic atmosphere.
Prompt
Retro cinematic anime, 1990s hand-painted cel animation, legendary samurai duel, old manga aesthetic.
How to Use This Prompt
Copy the prompt
Grab the cleaned-up prompt in your language, or send it straight to the editor with one click.
Pick your model
We pre-select the model that produced this result, but you can swap it for any model Vorla supports.
Generate and refine
Run it as-is to reproduce the look, then edit the wording to make the result your own.
More from Seedance 2.0
Cinematic stop-motion miniature cooking video. The scene is a handmade miniature summer backyard barbecue tabletop model. Real human hands appear in frame without showing any face, preparing a tiny summer cookout hot dog about 3 cm long. The visual style combines a premium food commercial with handcrafted stop-motion animation: warm, healing, delicate, and appetizing. Scene details: A miniature backyard grill sits on a handmade wooden table. Inside the grill are small flames and a subtle charcoal glow. Nearby are miniature hot dog buns, tiny plates, miniature tongs, ketchup, and mustard. The two sauces have a colorful, soft clay-like texture. All props look like handmade miniature models, with visible textures of clay, cardboard, wood chips, metal tongs, and tiny food. Action sequence: 1. Real human hands enter from the edge of the frame and use tiny tongs to pick up the 3 cm miniature hot dog. 2. The hot dog is gently placed onto the miniature grill. Small flames flicker, and subtle grill marks appear on the surface. 3. The tongs pick up the hot dog and place it inside the opened miniature bun. 4. A small amount of clay-textured ketchup and mustard is squeezed onto the hot dog, forming cute red and yellow lines. 5. Finally, the hands gently push the finished miniature hot dog toward the camera and pause for a moment, creating a food-commercial hero shot. An advertising tagline appears. Camera language: Macro lens, shallow depth of field. Focus should always prioritize the miniature hot dog, tongs, grill, and sauces. Start with a low-angle macro close-up and slowly push in. During grilling, the camera gently follows the movement of the tongs. End on a close-up of the finished hot dog. The motion must be stable, clear, and readable. Avoid fast shaking, sudden cuts, or chaotic movement. Visual style: Warm July evening light, golden side/back light, slight background blur, handcrafted stop-motion texture, 12fps stop-motion rhythm, while keeping materials and lighting realistic. The small flames flicker softly. Food movement should have a subtle handmade frame-by-frame animation feeling. The overall result should feel satisfying, like a refined miniature cooking commercial.
<<<image_1>>> is the exact snowman character reference - preserve his round proportions, friendly face, carrot nose, black coal buttons, twig arms, knitted red pom-pom hat and matching red scarf in every shot. <<<image_2>>> is the exact beach location reference - preserve the wide white sandy beach, calm turquoise ocean, deep blue sky, beach towels, vacation accessories and sunbathers relaxing in the background. Full-color polished 3D animated comedy short, soft detailed snow texture, expressive character animation, bright tropical sunlight, smooth 24 fps motion, exaggerated cartoon timing and physics, consistent rendering throughout. No subtitles, no text overlay, no captions. A cheerful snowman arrives at a crowded tropical beach for his vacation, relaxes with a cocktail and portable fan, but the fan battery dies and he melts in one sudden comic disaster. Opens on a wide establishing shot matching <<<image_2>>>: the snowman cheerfully walks onto the white sand from the edge of the frame, pulling a red suitcase behind him while the sunbathers continue relaxing peacefully on their towels. Medium tracking shot: he reaches an open spot near the middle of the beach, stops, proudly looks around at the ocean and sets down his suitcase. Medium shot, quick push-in: he opens the suitcase and rapidly pulls out a colorful folding sun lounger and a tropical cocktail with a straw and tiny umbrella. Wide shot, static: he reclines comfortably on the lounger, smiling and sipping the cocktail, but the blazing sun makes him sweat and soften slightly. Medium shot, fast comedic action: alarmed by the heat, he rummages through the suitcase and urgently pulls out a small battery-powered portable fan. Close shot: he switches it on; cool air blows across his face, scarf and pom-pom flutter, and he relaxes again with a deeply satisfied smile while sipping the cocktail. Extreme close-up: the fan slows down, sputters and stops with a weak electronic dying sound. Medium handheld shot: the snowman panics, desperately searches the empty suitcase for spare batteries, throwing harmless vacation items into the air, but finds none. Tight close-up, fast dolly-in: he freezes, realizes he is doomed, looks directly ahead in absolute horror and screams at the top of his lungs, "NOOOOO!" Hard cut to a wide shot: he instantly collapses into a large splash and puddle of clear water on the sand. The nearby sunbathers briefly lift their heads in confused surprise. Final comedic hold: only his orange carrot nose, red knitted hat, red scarf, black coal buttons and two twig arms remain scattered beside the empty lounger, silent fan and untouched cocktail. Audio: cheerful tropical vacation music, gentle ocean waves, distant beach ambience, footsteps crunching through sand, suitcase clicks, folding lounger snap, cocktail sipping, hot sizzling snow, spinning fan, weakening motor whine, frantic rummaging, loud despairing scream "NOOOOO!", sudden watery splash; music stops sharply after the melt. No ship, no boat, no island, no gore, no disturbing imagery, no character duplication, no extra limbs, no surviving snow body, no rebuilding after melting, no text, no labels, no watermarks, no logos, no UI elements.
一、核心定位电影级写实质感,纯古风中国仙侠美学。整体采用:高级冷面动作喜剧默片式视觉误会 “高手光环”引发的群体投射不断升级的集体脑补克制的互联网热梗节奏 Arri Alexa 电影机质感清晰稳定的面部微细节细腻胶片颗粒自然体积光本段喜剧必须严格按照三个清晰节拍推进:偶然躲过一剑所有人自动脑补成绝世身法第二次巧合进一步坐实高手滤镜最普通的真相当场拆掉整个传奇笑点不能来自人物变蠢。笑点来自:外界对“高手”的过度投射连续巧合造成的错误神话所有人一本正经地误判真相揭晓后仍有人嘴硬圆场二、参考控制 1. 人物身份锚点 @图片 1 严格锁定角色 ID A:剑仙师姐 @图片 2 严格锁定角色 ID B:小师妹 2. 环境参考本轮新上传的全部背景与地点参考图,共同决定同一套环境 DNA。正式构图之前,先无声整合彼此兼容的:真实地形建筑语言空间尺度材质植被水体天气云层山雾主光方向反光关系空气纵深真实可行走路线然后重新规划成一个:本轮独有完整统一空间合理物理连续的新空间。 3. 环境原则背景始终持续鲜活,但在叙事上绝对中立。允许自然持续运动的元素包括:风植被水体山雾倒影远处见证人物空间环境声但必须严格遵守:背景不能触发剧情背景不能制造笑点背景不能替人物解释误会背景不能配合包袱节奏背景只提供真实世界存在感三、角色设定角色 ID A|剑仙师姐同一个 @图片 1 剑仙师姐。 25–30 岁东亚女性,椭圆脸,白皙肤色,深色杏眼,黑色长发半挽,以白玉簪固定,身材高挑纤细。固定造型:白色刺绣真丝汉服半透明分层宽袖银色腰封玉佩白色布靴唯一一柄银色长剑人物气质:冷静端庄反应天然并不刻意装高手对别人为何误解自己毫无预设关键表演要求:第一次蹲下系鞋带时必须自然、日常、完全没把敌人当回事第二次低头检查玉佩时也必须顺手而为直到最后才真正抬头理解发生了什么全程不能故意耍帅,必须像她真的只是在处理小事角色 ID B|小师妹同一个 @图片 2 小师妹。 20–25 岁东亚女性,圆润灵动的脸,黑发编辫,身形娇小。固定造型:青绿色亚麻汉服深色腰带木簪黑色布鞋唯一一柄深色钢剑人物气质:聪明观察快全程最接近观众视角越往后越清楚大家误会了什么必须努力忍笑关键表演要求:第一次巧合后先看看师姐,再看看周围人第二次提醒“玉佩戴反了”要非常自然,完全不是故意配合后半段忍笑不能太夸张,只能靠嘴角压制、眼神变化、轻微肩膀控制完成说“你自己的脑补”时必须面无表情,像在宣布事实配角 1|敌方剑客一名极其自负先轻敌,后震惊,再崩塌负责误把巧合认成绝学关键表演要求:第一剑气势要足,但动作清楚安全第二剑开始明显紧张,又不愿承认自己被压制第三段深鞠躬“前辈,我服了”必须极其郑重最后信念崩塌时,反应要真实,不要卡通化配角 2|年长师父一名只作为远处见证者负责最后一次权威补刀不能抢戏,但每次开口都很重关键表演要求: “不看剑,也能避剑。” 必须像认真点评武学最后“连续两次……也算本事。” 必须像学术裁定最后要故意藏笑,但不能演成明显搞笑老人配角 3|远处弟子两名只承担见证与氛围反应位置较远不能抢主画面关键表演要求:第一次闪避后同步震惊第二次后压低声音交流表情必须像真的见证了传说级操作不能比主角更显眼四、核心道具与动作元素 1. 银色长剑属于师姐全程身份稳定本段不作为真正交锋核心,只用于身份与敌我关系确认 2. 深色钢剑属于小师妹全程身份稳定 3. 玉佩师姐随身佩戴第二次误会的关键原因必须清楚可见,且确实存在“戴反了”的合理视觉细节 4. 白色布靴鞋带第一次误会的关键原因必须真实可见地松开师姐蹲下系鞋带必须清楚可读五、喜剧结构第一拍|偶然闪避敌人出第一剑。师姐不是躲,而是因为鞋带松了,顺手蹲下系鞋带。剑从她头顶安全掠过。外界解读:她竟然不看剑也能避剑。第二拍|误会升级敌人第二次出手。小师妹提醒师姐玉佩戴反了。师姐顺手侧身低头检查,第二剑再次落空。外界进一步解读:她连看都不看对手。她是在用绝世身法戏耍敌人。第三拍|真相揭晓敌人正式认输。师姐一脸困惑地解释: “我刚才在系鞋带。” “第二次是在看玉佩。” 所有神话一瞬间坍塌。最后补刀当众人都僵住时,师父仍认真裁定: “连续两次……也算本事。” 笑点在于:真相极普通误会极宏大权威人物还在帮忙圆回来六、分镜脚本 0–5s|全景或远景角色 ID A 始终是同一个 @图片 1 剑仙师姐。角色 ID B 始终是同一个 @图片 2 小师妹。两人正前方只有一名极其自负的敌方剑客。同一个年长师父和两名弟子只在远处作为见证者。敌人突然大喝: “接我一剑!” 随后完成一次气势极足但动作清晰的横向试探剑击。就在完全相同的一瞬间,同一个白衣剑仙发现自己的白色布靴系带松开,于是极其自然地低身蹲下重新系鞋带;敌人的剑锋安全地从她头顶上方掠过。远处两名弟子同时震惊。师父极其郑重地低声说道: “不看剑,也能避剑。” 镜头要求先清楚建立四类人物位置关系敌人出剑动作必须清楚、流畅、安全师姐蹲下动作必须自然到像完全无视了敌人剑锋必须明确从头顶上方安全掠过远景见证反应要成立,但不能抢主画面 5–10s|中景或牛仔景保持同一个白衣剑仙、同一个青衣小师妹、同一个敌人和完全相同的地理空间。敌人明显开始紧张,却嘴硬喊道: “装什么松弛!” 随后再次发动一次简单清楚的斜向佯攻。就在这一刻,同一个小师妹突然提醒: “师姐,你玉佩戴反了。” 同一个剑仙师姐极其自然地侧过身体低头检查同一枚玉佩,第二剑竟然又从身侧落空。敌人立刻后退一步,彻底震住: “她连看都不看我?” 远处一名弟子压低声音: “这就是高手?” 同一个小师妹缓慢看看师姐,又看看周围所有人,已经开始拼命忍住笑意。镜头要求第二剑必须和第一次不同,形成递进小师妹提醒玉佩时语气必须自然,不带包袱腔敌人后退一步时要明显心态开始崩小师妹忍笑要非常克制背景继续自然运动,但绝不能和动作同步制造笑点 10–15s|特写或极近特写同一个敌人突然把长剑归鞘,极其郑重地向师姐深深行礼: “前辈,我服了!” 同一个剑仙师姐终于抬起头,是真的完全没理解发生了什么: “服什么?我刚才在系鞋带。” 全场绝对安静半拍。同一个小师妹非常好心地补充: “第二次是在看玉佩。” 再次完整安静半拍。所有人的表情同时凝固。焦点转移到远处同一个年长师父;师父竟然认真思考一下,然后一本正经说道: “连续两次……也算本事。” 极近特写:敌人整个信念开始崩塌,缓慢问道: “那我刚才到底输给谁了?” 同一个小师妹面无表情回答: “你自己的脑补。” 同一个白衣剑仙终于彻底听懂,转头看看这一群人,小声吐槽: “你们对高手是不是有什么误会?” 最终浅景深后方,同一个师父立刻把脸转向另一侧,故意藏住已经快压不住的笑意,精准切黑。镜头要求 “前辈,我服了”必须当真真相揭晓后必须留出两次完整安静停顿师父补刀要像认真仲裁,不像捧哏 “你自己的脑补”是最后逻辑收束结尾切黑前,师父藏笑是最后余味七、表演节奏要求师姐全程自然,不自知地制造误会不是故意装高手最后的困惑必须真实吐槽时也要轻声、克制小师妹负责观众视角负责逐渐意识到大家脑补过头负责后段轻微憋笑负责最后精准点破敌人开始自负中段动摇后段郑重认输最终精神系统崩塌师父每次开口都像正式判断自己越认真,越好笑八、摄影与空间要求 16:9 横屏三个连续清晰镜头 Arri Alexa 电影机观感清晰稳定的面部微细节自然体积光细腻胶片颗粒喜剧停顿必须清楚可读两次剑击轨迹必须清晰且安全配角站位稳定连续前景、中景、远景保持自然视差视线关系必须准确九、声音要求原生同步普通话对白口型精准喜剧停顿明确真实空间环境声背景风、水体、山雾、植被、远景脚步与空间氛围持续自然存在不生成字幕十、技术规格总时长:15 秒画幅:16:9 横屏三个连续清晰镜头原生同步普通话对白精准口型与喜剧停顿人物身份、服装、长剑、玉佩、配角站位与地理空间全程稳定两次剑击轨迹清楚且保持安全距离视线连续真丝衣料和头发物理效果真实背景拥有持续自然视差和真实空间环境声不生成字幕十一、Negative blurry, bad quality, low quality, low resolution, noisy, jpeg artifacts, watermark, text, error; deformed, mutated, bad anatomy, poorly drawn hands, bad composition, out of frame, disfigured; inconsistent character, changing clothes, face morphing, background shift, glitching cuts, disappearing props
SCENE CONTEXT A young woman with deep red hair sits alone at a small dark red table while a dense crowd of standing people rushes around her in every direction. She slowly turns a small red speckled object in her fingers and holds her gaze off screen-left, past the frame edge. ACTIVE REFERENCES [IMAGE REFERENCE] is the exact first frame. It controls her face, red hair, black leather jacket, the dark red tabletop, the white cup, the folded sunglasses, the small red speckled object, the crowd density and the low-key grade. She is in her late twenties, calm, withdrawn, hands together over the object, lips closed. 100% matches the reference. FIRST FRAME AND SPATIAL BLOCKING The first visible frame is identical to [IMAGE REFERENCE] — no empty establishing frame, no reveal. Screen x-axis 0% left to 100% right, y-axis 0% top to 100% bottom. The woman sits mid-ground, x 50%, y 46%, torso angled screen-left, face in near-profile to camera, gaze locked off screen-left past the frame edge. The dark red tabletop fills the foreground lower center, x 46%, y 74%. Her forearms rest on the table, both hands cupped around the small red speckled object at x 45%, y 66%. Folded sunglasses lie flat at x 37%, y 68%; the small white cup stands at x 34%, y 66%. Standing crowd figures occupy every remaining zone, foreground through background, pressing in from all four edges. Exactly one woman is seated at the table and nobody joins her. FORMAT MODE Single continuous take. Real-time motion. No cuts, no dissolves, no transitions. OPTICS 29° diagonal field of view, short telephoto portrait lens character, camera 5 meters from her at slightly elevated height looking mildly down. Close framing achieved through lens reach, not proximity. She stays razor-sharp, the background compresses closer behind her, the surrounding bodies dissolve into soft smeared bokeh. CAMERA Camera locked on a heavy head with only faint operator breath — a slow, almost imperceptible creep forward of a few centimeters across the whole take, holding her head and the red tabletop in frame. Focus stays pinned on her face and hands the entire time; no rack focus, no pans, no whip moves, no orbit. ACTION TIMING The shot opens already in motion: the crowd around her is streaking while she is still. Her thumb rolls the small red speckled object a quarter turn against her palm. A brief beat later she blinks slowly, chest lifting with one shallow breath, a few strands of red hair settling against her jacket collar. Her eyes drift a fraction further screen-left, then hold. Near the end she lowers her chin barely a centimeter and stays there. She never speaks and never looks at camera. PHYSICS The crowd moves at walking-to-hurried speed with heavy shutter drag: every passing body smears into long elongated motion trails of muted gray, brown and dull blue that swirl and overlap continuously, arms and heads dissolving into streaks while shoulders occasionally cross the foreground as huge dark blurred masses. The trails never freeze and never fully clear. Her body carries real weight — forearms pressed into the tabletop, leather jacket creasing at the elbow with cloth delay, hair strands lagging a fraction behind her small head movement. The object has small solid mass and stays in contact with her fingers. The table does not move. LIGHTING Single dim warm overhead practical source above and slightly screen-left, motivated from the ceiling. Exposure is set for her face and the dark red tabletop; the rushing crowd falls two stops darker and reads mostly as shadowed smear. Soft shadow rolloff on her cheek with a catchlight in her eye so the micro-expression stays readable. No flat frontal fill, no lens flare. AUDIO Dense muffled indoor crowd ambience: overlapping footsteps, fabric rustle, chair scrape, a low unintelligible murmur of many voices with no distinguishable words, slightly hollow room reverb. The murmur swells softly as bodies pass close to camera. No dialogue, no narration, no music. POSITIVE LOCKS She remains perfectly sharp, still and in focus for the entire take while all motion belongs to the crowd. Cinematic photorealistic footage shot on ARRI Alexa 35, low-key naturalistic grade, natural film grain, the red tabletop and her red hair the only saturated notes. No text, no labels, no watermarks, no logos, no UI elements.
Style & Mood: AAA game cutscene aesthetic, live-action realism - real human actors, visible skin pores, realistic hair strands, real cloth and armor physics, fabric and scarves reacting naturally to extreme wind, no animation or cartoon stylization; hyper-saturated cinematic color grading, dramatic lighting, photorealistic 35mm film quality, sharp focus, fine film grain, blazing desert afternoon light cutting through a massive sandstorm, HDR bloom on golden dust, glowing energy cores, and polished metal surfaces. Narrative Summary: Above a forgotten desert megacity suspended between colossal sandstone cliffs, a lone sky raider infiltrates a flying fortress during a massive aerial battle. She crosses collapsing airships, hijacks an enemy energy core, and escapes as a giant ancient sky creature awakens beneath the storm clouds, tearing the entire battlefield apart. Dynamic Description - three layers of action in every beat: 0–3s: FOREGROUND - a damaged airship wing tears across the frame as sand and sparks blast toward camera. A metal platform collapses beneath the heroine's feet while cables snap around her. MIDDLEGROUND - the heroine runs across the exterior hull of a massive flying warship, using a magnetic grappling blade to pull herself between two airborne vessels. Enemy soldiers fire energy weapons as she slides beneath incoming attacks. BACKGROUND - dozens of airships battle inside a gigantic desert storm vortex, with a floating ancient city visible between towering cliffs, its golden towers glowing through the dust. 3–6s: FOREGROUND - an enemy fighter lunges toward camera with an energy spear; the heroine redirects the attack with her armored gauntlet and throws him into the storm below. MIDDLEGROUND - she jumps onto the enemy flagship deck, fighting through armored guards in close combat. She disables a massive engine mechanism while the ship tilts violently. BACKGROUND - another airship crashes through the storm behind her, breaking apart into burning metal fragments while lightning illuminates the distant floating fortress. 6–9s: FOREGROUND - broken mechanical parts and glowing energy shards fly past the camera as the flagship's core begins to overload. MIDDLEGROUND - the heroine tears out the glowing energy core, attaches it to her equipment, and sprints across the collapsing deck while the entire ship falls apart beneath her. BACKGROUND - beneath the clouds, a COLOSSAL ANCIENT SKY LEVIATHAN emerges from the sandstorm, its enormous crystalline wings spreading across the sky and dwarfing the entire fleet. 9–12s: FOREGROUND - the heroine leaps from the collapsing flagship, suspended in the air as debris and sand swirl around her. MIDDLEGROUND - she swings on a broken cable between two falling airships, landing on a smaller escape craft while enemy ships explode behind her. BACKGROUND - the awakened sky leviathan crashes through the battlefield, its wings generating massive shockwaves that scatter the remaining airships across the storm. 12–15s: FOREGROUND - the heroine stands on the edge of the escape craft, coat and hair whipping violently in the wind, holding the stolen glowing energy core. MIDDLEGROUND - the escape craft dives through the storm clouds as fragments of the destroyed fleet rain around it. BACKGROUND - the floating desert city is revealed beneath the clearing storm: gigantic sandstone towers, ancient technology, suspended bridges, and glowing energy channels stretching across the canyon. Camera slowly pulls back to reveal the impossible scale of the desert civilization. Static Description: A vast desert civilization floating above a deep canyon - enormous sandstone towers connected by suspended bridges, ancient futuristic architecture covered with glowing energy circuits, giant airship docks carved into cliffs, endless dunes below, and a violent golden sandstorm surrounding the city. Heroine: photorealistic young adult woman, athletic build, pale olive skin, sharp gray eyes, short silver hair with a long side fringe, wearing a dark indigo desert combat coat with layered armor plates, lightweight tactical harness, reinforced boots, metallic gloves, and a glowing energy device attached to her waist. Her design combines futuristic explorer equipment with ancient desert warrior elements. Enemies: armored sky raiders wearing dark bronze mechanical armor, carrying energy spears and magnetic weapons. Sky Leviathan: a colossal ancient flying creature with obsidian scales, translucent crystal wings, glowing blue energy patterns beneath its armor-like body, original creature design. Original characters and creature design, not resembling any famous actress, no franchise resemblance, no logos, no readable text. Audio: Powerful desert wind roaring through the canyon, airship engines vibrating, metal structures cracking, cables snapping under tension, energy weapons firing, armor impacts, explosions echoing between cliffs, the deep atmospheric roar of the sky leviathan, massive wing beats creating shockwaves, intense orchestral score building toward the creature's awakening and final reveal. Quality: Photorealistic live-action, AAA cinematic game trailer quality, 35mm film look, HDR, sharp deep focus holding all three depth planes, realistic dust and particle physics, natural motion blur on flying debris and fast movement, consistent character identity throughout, no morphing, no warped anatomy, no duplicated characters, no text or watermarks.
Single continuous shot, one take no cuts, cinematic oner, AAA game cutscene quality, render aesthetic, photorealistic, 35mm film quality, professional color grading, sharp focus, high detail texture, film grain, depth of field mastery, FPV camera fluidity, 30 fps A vast ruined colosseum at night, its cracked stone floor split by glowing ember fissures, braziers burning along collapsed tiers and drifting sparks filling the smoky air. A red-haired gladiatrix with a long braid, silver ornate cuirass, yellow tattered cape, crimson leather skirt and a longsword faces a towering pale ogre wearing a horned bull skull, its stitched flesh wrapped in chains, swinging a massive double-headed axe. She fights with videogame agility - dodge rolls, slides, lunging strikes - while the camera behaves like an aggressive in-engine action cam, whipping, orbiting and diving through the fight in one unbroken take. Diegetic sound only: axe wind, boots on stone, chain rattle, roaring embers. Single continuous shot 15s: The camera starts low behind the gladiatrix's shoulder in the framing of the source image as she explodes forward off her back foot, sprinting at the ogre with her longsword trailing sparks along the cracked stone. The ogre's axe crashes down where she was - the camera whips right and drops to ground level, tracking her videogame-style dodge roll under the blade as shattered stone and embers spray across the lens, the frame jolting with the impact. She springs up mid-roll and the camera swings into a tight orbiting arc around her body as she slashes twice across the ogre's stitched flank, each hit answered by a directional camera shudder and a spray of sparks, the yellow cape snapping with her pivots. The ogre swings its chain flail in a wide horizontal arc - the camera dives beneath it with her as she slides on her knees between the monster's legs, brazier light strobing across her armor. Rising sharply behind her, the camera pushes into a medium close-up as she plants her boot on a fallen stone block and launches upward, sword raised in both hands toward the ogre's bull-skull face, jaw clenched, embers streaming past her - the frame racing up with her and cutting to black at the apex of the leap, blade a heartbeat from contact.
@图1 为角色一致性参考:红金黑战衣、红色兜帽围巾、金色闪电纹面罩、胸口金色圆环纹章、黑色装甲猴尾配金环、指尖与尾尖可射出可操控的黄金丝线,完全以图1为准。【渲染风格】电影级3D渲染动画,表面覆盖手绘油画质感贴图(Painterly 3D Render):真实三维体积与透视,材质表面呈现清晰手绘笔触与颜料堆叠纹理,电影级体积光与大气透视,浅景深自然虚化,高端3D绘画动画剧集质感(Arcane风格)。绝非扁平2D作画,绝非塑料感CG。晴朗正午,湛蓝天空洁白积云,强烈明媚阳光。16:9。【核心节奏·务必执行】全片15秒无任何停顿、无铺垫、无对峙——从第0帧即为全速运动,每2秒切换全新机位与新技巧。所有搞笑桥段必须在高速运动中完成,绝不为了搞笑而停下动作,速度与幽默同时进行。【音效核心设定】黄金丝线绷紧与弹射时发出竖琴般的清脆音符,随速度密集连奏。 [0-2秒] 零帧启动·被拽入画:开场第一帧即为极特写——黑色战斗爪手套的指尖爆射出四道黄金丝线,镜头随丝线高速前冲划过整条街道空中,丝线"锵"地锚定远处塔尖绷紧——镜头猛然回甩,英雄被丝线拽着从画面下方以极速斜向上冲入画面,红色兜帽围巾在身后炸开,气浪掀翻街边彩旗。 [2-4秒] 主观穿索·鸽子糊脸:切至第一人称主观视角,他贴着密集交错的缆车钢索高速荡行,钢索与条幅贴着镜头呼啸掠过——一只受惊的鸽子迎面"啪"地糊在他面罩上完全挡住视野,画面被灰白羽毛填满,他单手把鸽子从脸上揭下来托在掌心,两者在高速飞行中对视半秒,他歉意地把鸽子往上一抛放飞,鸽子愤怒地扑棱着飞走,他继续换线荡行,全程没有减速。 [4-6秒] 垂直墙面冲刺·猫爪突袭:他落上米白色大理石立面四肢着壁极速斜向奔跑,掌心金色纹章亮起、金爪抠出石屑拖痕——冲过一扇窗台时,一只趴着晒太阳的橘猫猛地挥爪拍向他,他不减速地向后仰身躲过,顺手用尾尖轻轻点了一下猫鼻子,橘猫炸毛僵住——他已冲到屋檐,猴尾如撑杆狠狠一撑弹射翻越尖顶,天空瞬间灌满画面。 [6-8秒] 自由落体·顺手买个苹果:他背朝下坠入两座塔楼之间的垂直峡谷,双臂张开,指尖射出的金丝线在身后飘成数条舞动的绸带——坠落途中猴尾精准地从下方水果摊的帐篷上卷走一颗红苹果,同时指尖弹出一道金丝线把一枚金币"叮"地送进摊主的钱罐里,摊主茫然抬头四处张望,他已在坠落中翻身咬了一口苹果,镜头与他一同旋转下坠。 [8-10秒] 金线弹弓·条幅裹头:坠到底前他双手同时射线锚定两侧塔楼,身体后仰将丝线拉成一张巨大的弓——升格慢镜半拍蓄力,金线绷至极限微微震颤——"崩"地释放,他如箭般水平超速弹射出去,却一头扎穿一面垂落的深红条幅,布料整个罩住他的脑袋,他在高速飞行中手忙脚乱地扒拉了半秒,一把扯下来——布料顺势挂在肩上变成一件飘扬的红色披风,他立刻若无其事地摆了个帅气姿势,苹果还叼在嘴里。 [10-12秒] 索间穿花·空中合影:他在密集钢索之间高速穿行,手与尾交替接力荡行——经过一节满载乘客的红铜色缆车时,车内乘客正举着手机隔窗拍他,他顺势倒挂在车窗外与他们同速并行,猴尾比出一个V字手势,乘客们欢呼大笑,他把嘴里的苹果隔着敞开的车窗递给一个小孩,随即一蹬车顶弹射离开,披风猎猎。 [12-15秒] 巨像掠水·回环收尾:他贴着黄铜巨像喷泉低空掠过,尾尖点水拖出一道弧线——水花却"哗"地反弹糊了他一脸,他在飞行中甩了甩头把水甩开、像猫一样抖了抖尾巴上的水珠,随即拉升冲向高空,在最高点身体舒展翻转一圈,披风与彩虹在身侧展开——就在冲出画面的瞬间,他再度伸手朝镜头方向射出四道黄金丝线,构图与光线与开场第一帧完全一致,形成无缝循环,画面在丝线爆射的瞬间切断。运镜:指尖极特写随线前冲回甩→第一人称主观穿索→贴墙平行跟拍急上摇→旋转自由落体→慢镜蓄力后急速后拉→并行跟拍缆车→低空掠水急速仰拍循环切断。全程高速运动无静止,每2秒全新机位。手绘质感3D渲染,专业电影调色,16:9。环境音:风声呼啸贯穿全片,黄金丝线的竖琴音符密集连奏,鸽子受惊的扑棱与鸣叫,猫的炸毛嘶声,金币落罐的"叮",条幅撕裂的布料声,乘客的欢呼与笑声,水花飞溅,轻快诙谐的管弦乐与鼓点持续推进不留空隙,结尾在丝线爆射的一声重音上骤然切断。 Negative prompt:扁平2D作画,赛璐璐平涂,塑料感CG,游戏引擎质感,照片写实,与参考图不符,战衣改变,兜帽缺失,尾巴缺失,纹章改变,蛛网图案,场景风格改变,静止镜头,动作停顿,缓慢铺垫,夜晚,阴天,畸形的手,多余的手指,丝线形状混乱,动物造型混乱,运镜混乱,掉帧卡顿,构图失衡,低分辨率,模糊,过度饱和,文字,字幕
All Video Categories
Grok Imagine
Pink and Cyan Liquid Text Intro
Find Your Next AI Prompt
One search covers every model and both mediums, and copying what you find costs nothing.















