Seedance 2.0
Two xianxia masters turn a plain exchange of glances into an epic duel with quiet pauses
Prompt
Seedance Prompt | Turn a staring contest into a master-level duel I. Project Positioning Cinematic realistic texture, pure ancient Chinese Xianxia aesthetics. Overall style requirements: Cinematic grammar of an epic master duel. Restrained visual comedy. Arri Alexa cinematic texture. Clear and stable facial micro-details. Delicate film grain. Natural volumetric lighting. Abundant controlled quiet pauses. Precise reaction rhythm. Core Plot: Two true masters turn the mundane act of "staring at each other" into a battle of dignity. Hard Constraints: No failed spells. No secrets revealed. No environmental accidents. No character humiliation. All humor must stem from the two characters' pride, competitive spirit, micro-expressions, breathing, and rhythm. II. Narrative Principles Plot progression must come solely from the two characters': Intentions. Eye contact. Breathing. Micro-expressions. Pauses. Rhythm. Dialogue. Minimal physical reactions. The environment must always be a vivid, continuously running, real world, but remain narratively neutral. Environment elements that must persist: Flowing water where applicable. Mountain mist moving according to real terrain. Vegetation at different depths responding to natural wind at different amplitudes. Clouds and natural light changing slowly. Reflective surfaces changing naturally with camera angles. Background maintaining natural parallax and spatial ambient sound. Environmental Constraints: Must never trigger the plot. Must never solve problems. Must never interrupt characters. Must never create burdens for characters. Must never decide the winner for characters. III. Reference Control 1. Character Anchors @Image 1 strictly as Character ID A: Sword Immortal Senior Sister. @Image 2 strictly as Character ID B: Junior Sister. 2. Environmental Reference Treat all newly uploaded background and location reference images as a single set of environmental DNA. Before formal composition, silently execute an environmental planning phase, deducing: Real terrain. Architectural language. Spatial scale. Materials. Vegetation ecology. Water systems. Weather. Cloud layers. Mountain mist density. Main light direction. Reflective surfaces. Overall color palette. Aerial perspective/depth. Realistically feasible spatial movement. Reorganize into: A new, complete, and unified space unique to this round that has never appeared before. Requirements: Do not mechanically copy a single reference image. Retain the identity and logic of the reference world. Three shots share the same geographical space and environmental system. IV. Character Settings Character ID A | Sword Immortal Senior Sister Same as @Image 1. 25–30 years old, East Asian female, oval face, fair natural skin, dark almond eyes, long black hair half-up with a white jade hairpin, tall and slender. Always wearing: White embroidered silk Hanfu. Translucent layered wide sleeves. Silver waist belt. Jade pendant. White cloth boots. Carries one silver longsword, kept sheathed throughout. Character ID B | Junior Sister Same as @Image 2. 20–25 years old, East Asian female, round and lively face, braided black hair, petite figure. Always wearing: Cyan-green linen Hanfu. Dark belt. Wooden hairpin. Black cloth shoes. Carries one dark steel sword, kept sheathed throughout. V. Core Performance Logic The scene looks like a master duel on the surface, but the core is: Not about the sword. About composure. About who breaks first. About who can suppress basic human reactions. Must be filmed like: A real life-or-death duel. A real master's contest. A real battle of martial dignity. Humor comes from the contrast: The matter is trivial. The ritual is grand. Both are extremely serious. All pauses are filmed as critical moments deciding victory or defeat. VI. Storyboard 0–5s | Wide or Long Shot In the main quiet space planned according to the real geographical logic of the reference images, the two women stand or sit reasonably, exactly three steps apart, both swords sheathed. They stare at each other with the seriousness of facing a life-or-death enemy. The white-clad Sword Immortal says calmly: "Today we don't compete with swords, but with composure." The cyan-clad Junior Sister raises one eyebrow slightly: "How?" Senior Sister replies: "Whoever looks away first, loses." Junior Sister immediately replies: "Bring it on." Requirements: Establish the space clearly. Clear distance between the two. Swords clearly sheathed. Serious expressions like a real duel. Camera stable, restrained, slightly ritualistic. Background runs naturally, not intervening in the plot. 5–10s | Medium or Cowboy Shot Same two women, same costumes, same sheathed swords, same geographical space. Maintain a clear, restrained, symmetrical two-person composition, then the camera pushes in extremely slowly. Neither person's head moves; eyes locked. Junior Sister squints slightly. Senior Sister raises one eyebrow slightly in response. Breathing remains stable. After a moment, one pair of eyes begins to show slight physiological moisture, followed by the other. No exaggerated facial contortions; must be filmed like a top-tier swordsman duel. Junior Sister says in a low voice without moving her lips: "Senior Sister, your eyes are red." Senior Sister replies calmly: "You too." Requirements: No head movement. Eyes must not wander. Micro-expressions must be tiny, precise, and readable. Camera push-in should feel like closing in on the core of the duel. Background continues natural movement, zero causal participation. 10–15s | Close-up or Extreme Close-up Switch to extreme close-ups of both pairs of eyes. Requirements: Clear natural trembling of eyelashes. Lower eyelids gradually gathering tiny moisture. Skin pores. Subtle facial muscle tension. At the exact same moment, both blink once. Absolute silence for half a beat. Junior Sister says immediately: "You blinked first." Senior Sister replies without changing expression: "You spoke first." Another full half-beat pause. Junior Sister asks: "Draw?" Senior Sister replies: "Again." Final extreme close-up: Both simultaneously straighten their posture, lock eyes again, and lean forward by the exact same tiny amount, as if a second epic battle is about to begin. Cut to black precisely before anyone blinks again. Requirements: First blink must be synchronized. Synchronization is the core of the humor. The argument must remain solemn and serious. The ending must leave a strong aftertaste. Cut to black must be clean and sharp. VII. Performance and Action Requirements Character State Both are smart, powerful, and dignified. No exaggerated faces. No funny faces. No deliberately comical actions. Performance restrained to the point of being strict. Micro-expression Control Focus on: Squinting. Raising eyebrows. Stable breathing. Slight eye moisture. Eyelash trembling. Lower eyelid moisture. Tiny nervous tension. Body Control Head must not move. Swords sheathed throughout. Posture upright and stable. Body center of gravity must be realistic. Final synchronized lean is small but must be identical. VIII. Photography and Camera Requirements 16:9 landscape. Three continuous clear shots. Arri Alexa cinematic look. Camera movement restrained and stable. From wide shot establishing space. To medium shot pushing in for pressure. To extreme close-up magnifying physiological details. Have the courage to let the screen pause; allow silence to persist. No flashy camera angles. No digital-feeling zooms. No exaggerated editing rhythm. IX. Sound Requirements Native synchronized Mandarin dialogue. Precise lip-sync. Restrained soundtrack. Real spatial ambient sound. Wind, water, and background natural sounds persist. Silence itself must be part of the tension. The pause before and after the blink should make the scene feel almost empty. No exaggerated comedy sound effects. X. Environmental Requirements Environment is always vivid but always neutral. Real physical changes that must persist: Water flow. Mountain mist movement. Vegetation responding to wind. Slow cloud migration. Natural light changes. Reflective surfaces changing with camera angles. Natural parallax in the distance. Spatial ambient sound persists. Prohibited: Environment triggering humor. Environment helping decide the winner. Wind suddenly increasing to match dialogue. Light suddenly changing to match a punchline. Any background accidents. Any "divine intervention" style reactions. XI. Technical Specifications Total duration: 15 seconds. Aspect ratio: 16:9 landscape. Three continuous clear shots. Seedance 2.0 Fast enhanced for multi-character identity consistency. Character faces, hairstyles, costumes, swords, and geographical relationship stable throughout. Native synchronized Mandarin dialogue. Precise lip-sync. Facial micro-performances clear and readable. Eye moisture, eyelashes, skin, silk fabric, and hair physics are realistic. Background is vivid but narratively neutral. Foreground, midground, and background maintain natural parallax and real spatial ambient sound. No subtitles. Only two visible characters at all times. XII. Negative
Seedance 提示词|把对视演成高手决斗 一、项目定位 电影级写实质感,纯古风中国仙侠美学。 整体风格要求: 史诗级高手对决的镜头语法 克制的视觉喜剧 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 大量有控制的安静停顿 精准反应节奏 核心剧情: 两个真正的高手,把一件普通到不能再普通的“互相盯着看”,硬生生上升成一场关乎武者尊严的较量。 硬性限制: 不允许法术失败 不允许秘密揭晓 不允许环境事故 不允许人物出丑 所有笑点只能来自两个人的骄傲、胜负欲、微表情、呼吸和节奏 二、叙事原则 剧情推进只能来自两名人物自己的: 意图 眼神 呼吸 微表情 停顿 节奏 对白 极小的身体反应 环境必须始终是一个鲜活、持续运行、真实存在的活世界,但叙事上保持绝对中立。 环境持续存在的内容包括: 有水的位置,水体持续流动 山雾按照真实地形游走 不同纵深植被以不同幅度响应自然风 云层与自然光缓慢变化 反光表面随摄影机角度自然改变 远景保持自然视差和空间环境声 环境限制: 绝不能触发剧情 绝不能解决问题 绝不能打断人物 绝不能替人物制造包袱 绝不能替人物决定输赢 三、参考控制 1. 人物锚点 @图片 1 严格作为角色 ID A 剑仙师姐身份锚点 @图片 2 严格作为角色 ID B 小师妹身份锚点 2. 环境参考 将本轮新上传的所有背景与地点参考图共同视为同一套环境 DNA。 正式构图前,先无声执行一轮环境规划,综合推演: 真实地形 建筑语言 空间尺度 材质 植被生态 水系 天气 云层 山雾密度 主光方向 反光表面 整体彩色 空气纵深 真实可行的空间动线 再重新组织成: 一个本轮独有、完整统一、此前从未出现过的新空间。 要求: 不机械复制单张参考图 保留参考世界的身份 and 逻辑 三个镜头共享同一地理空间与环境系统 四、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,白皙自然肤色,深色杏眼,黑色长发半挽,以白玉簪固定,身材高挑纤细。 始终穿同一套: 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 持 唯一一柄银色长剑,全程入鞘状态。 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。 始终穿同一套: 青绿色亚麻汉服 深色腰带 木簪 黑色布鞋 持 唯一一柄深色钢剑,全程入鞘状态。 五、核心表演逻辑 这段戏表面像真正的宗师决斗,内核却是: 比的不是剑 比的是定力 比的是谁先失守 比的是谁更能忍住最基本的人体反应 必须拍得像: 真正的生死决斗 真正的高手较量 真正的武者尊严之争 笑点来自反差: 事情极小 仪式感极大 两人极其认真 所有停顿都被拍得像决定胜负的关键瞬间 六、分镜脚本 0–5s|全景或远景 在根据本轮参考图真实地理逻辑自然规划出的主要安静空间中,同样两名女子根据场地合理选择站立或端坐,相隔严格三步,两柄长剑均保持入鞘状态。 她们却以面对真正生死大敌一般的严肃神情互相凝视。 同一个白衣剑仙平静说道: “今日不比剑,比定力。” 同一个青衣小师妹轻轻挑起一侧眉毛: “怎么比?” 师姐回答: “谁先移开视线,谁输。” 小师妹立即回答: “来。” 要求: 空间先建立清楚 两人距离明确 长剑明确处于入鞘状态 神情严肃得像真正决战 摄影机稳定、克制、略带仪式感 背景自然运行,但不介入剧情 5–10s|中景或牛仔景 保持同样两名女子、同样服装、同样入鞘长剑和完全相同的地理空间。 先保持一个清楚、克制、略带仪式感的对称双人构图,随后摄影机极其缓慢地向两个人推进。 两个人的头都绝不能移动,视线始终牢牢锁定对方。 同一个小师妹极轻地眯了一下眼睛。 同一个剑仙师姐只用极小幅度抬起一侧眉毛回应。 两个人呼吸始终稳定。 片刻之后,其中一双眼睛开始产生极轻的生理性湿润,另一双眼睛也逐渐出现同样变化。 绝不能夸张挤眉弄眼,必须把这一切拍得像真正的顶级剑客决斗。 小师妹几乎不动嘴唇地低声说道: “师姐,你眼红了。” 师姐同样平静地回答: “你也一样。” 要求: 头部不能乱动 视线不能飘 微表情必须极小、极准、极可读 摄影机推进要像在逼近决斗核心 背景前中远景继续自然运动,但零因果参与 10–15s|特写或极近特写 依次切换同样两双眼睛的极近特写。 要求清楚呈现: 睫毛极轻的自然颤动 下眼睑逐渐聚集微小水光 皮肤毛孔 面部细微肌肉紧张 就在完全相同的一瞬间,两个人终于同时眨了一次眼。 全场绝对安静半拍。 同一个小师妹立刻说道: “你先眨的。” 同一个剑仙师姐面不改色回答: “你先开口的。” 再次完整停顿半拍。 小师妹问: “平手?” 师姐回答: “再来。” 最终极近特写: 两个人竟然同时重新挺直身体,再次牢牢锁定彼此视线,并且以完全相同的极小幅度同时向前倾了一点,仿佛第二场传世大战即将开始。 在任何人再次眨眼之前精准切黑。 要求: 第一次眨眼必须同步 同步是笑点核心 争执必须仍然庄重认真 收尾必须留出强烈余味 切黑点要干净利落 七、表演与动作要求 人物状态 两人始终聪明、强大、有尊严 不做夸张鬼脸 不做挤眉弄眼式搞笑 不做刻意滑稽动作 表演克制到近乎严苛 微表情控制 重点表现: 眯眼 挑眉 稳定呼吸 眼部轻微湿润 睫毛微颤 下眼睑水光 极小的神经性紧张 身体控制 头绝不能随意移动 长剑全程入鞘 姿态端正稳定 身体重心要真实 最后的同步前倾幅度很小,但必须完全一致 八、摄影与镜头要求 16:9 横屏 三个连续清晰镜头 Arri Alexa 电影观感 摄影机运动克制稳定 从全景建立空间 到中景推进压迫感 再到极近特写放大生理细节 画面停顿要有勇气,允许安静持续存在 不用花哨机位 不用数字感强的推拉 不用夸张剪辑节奏 九、声音要求 原生同步普通话对白 精准口型 配乐克制 真实空间环境声 风声、水声、远景自然声持续存在 安静本身要成为张力的一部分 同步眨眼前后的停顿要让现场几乎空掉 不使用夸张喜剧音效 十、环境要求 环境始终鲜活,但始终中立。 需要持续存在的真实物理变化: 水体流动 山雾游走 植被响应风 云层缓慢迁移 自然光轻微变化 反光表面随摄影机角度改变 远景自然视差 空间环境声连续存在 禁止: 环境触发笑点 环境帮助分胜负 风突然变大配合对白 光突然变化配合包袱 任何背景事故 任何“天意配合”式反应 十一、技术规格 总时长:15 秒 画幅:16:9 横屏 三个连续清晰镜头 针对 Seedance 2.0 Fast 强化多参考人物身份一致性 人物面部、发型、服装、长剑和地理关系全程稳定 原生同步普通话对白 精准口型 面部微表演清楚易读 眼部水光、睫毛、皮肤、真丝衣料和头发物理效果真实 背景始终鲜活但在叙事上完全中立 前景、中景、远景保持自然视差和真实空间环境声 不生成字幕 始终只出现两名可见角色 十二、Negative
- Needs your photo: attach an image to the request — the prompt works with it.
The rights to this prompt belong to its author @Soranlan. You are responsible for any commercial use — promoting AI tools, aggregators or Telegram bots: mozgi.io accepts no liability. Details →











