Seedance 2.0
Urban xianxia "Shadow Play": a giant battle on a wall turns out to be two women showing a child upstairs a shadow theater
Prompt
PROMPT: Seedance 2.0|15-second short film 《影戏》 I. Generation Objective Generate a complete, continuous 15-second cinematic Chinese urban xianxia short film titled: 《影戏》 Overall style: Cinematic realism 融合 Chinese traditional shadow puppetry aesthetics Combined with silent-film-style spatial comedy Restrained, deadpan performances Emphasize the color-temperature contrast between warm yellow sodium-vapor streetlights and cool blue night Clear foreground, midground, and background depth Complete a twist in the final 5 seconds, making the audience reinterpret the true purpose of the entire “battle” II. Core Concept This is not a real battle to the death. For the first 10 seconds, the audience should believe that two women are engaged in a grand, epic shadow battle with a strong Eastern fantasy atmosphere. Only in the final 5 seconds is it revealed: They are actually using the relationship between real human movements and projected shadows to perform a “shadow play” for a child upstairs who cannot be seen on screen. The twist must work as follows: All the exaggerated grandeur in the first part comes from the shadows on the wall The real-world movements of both characters remain small, precise, and restrained throughout A single off-screen child’s voice later redefines the true purpose of all the preceding movements The comedy must not come from either person embarrassing themselves The comedy comes from the contrast between “epic-scale grandeur” and “the needs of a children’s game” III. Workflow and Technical Direction This segment is well suited to being structured according to Seedance 2.0’s multimodal reference approach: Support images, audio, and video as references Emphasize director-level control over character performance, lighting, shadows, and camera movement This type of two-person performance, which relies on the relationships between light and shadow, movement rhythm, and spatial visual reversal, is especially suitable Key elements to reinforce: Stable identities of both characters Stable shadow geometry Continuity of the sword and bicycle Stable logic of the wall projection Mandarin lip-sync Clear causal relationship between real-world movements and shadow movements IV. Character Identity Anchors Character ID A|@image1|Sword Immortal Sister Always maintain the same character identity and appearance: East Asian woman, 25–30 years old Oval face Naturally fair skin Sharp dark brown eyes Long, straight black hair Tall, slender figure White cloth boots Flowing white embroidered silk hanfu Translucent wide sleeves Silver waist ornament Jade hairpin The same silver longsword Character traits: Calm Restrained Serious to the point of being almost completely straight-faced Even when performing a “shadow play,” she behaves as if handling something genuinely important Character ID B|@image2|Bicycle Sister Always maintain the same character identity and appearance: East Asian woman, 25–30 years old Expressive face Brown ponytail Fixed body proportions Yellow jacket Blue jeans White sneakers Fixed accessories The same bicycle Character traits: Highly coordinated Equally serious Restrained reactions Her quiet muttered complaint at the end creates the deadpan punchline V. Environment Scene A narrow courtyard alley in an old urban district after rain. The following spatial elements must be clearly established: Rough white stucco walls Wet stone pavement Warm yellow streetlights Overhead electrical wires Shadows of bamboo leaves Drifting thin mist Reflections in puddles One bicycle Lighting Relationships Overall: Warm yellow sodium-vapor streetlight as the primary projection source Cool blue night as the ambient color base The shadows on the wall must be clear, huge, and readable The real figures remain normal in scale, while their shadows on the wall are exaggerated in scale The projection geometry must be physically believable and must not drift randomly or become unmotivated visual effects VI. Environmental Motion Principles The background must remain continuously lively while staying narratively neutral. Elements that may exist and continue moving: Thin mist flowing naturally Overhead electrical wires remaining mostly static or vibrating slightly Shadows of bamboo leaves swaying gently Reflections in puddles and on the wet stone pavement continuously changing Clothing, hair, and bicycle details responding naturally to the characters’ movements The following must be prohibited: Sudden background changes The white wall deforming Streetlight position jumping The bicycle disappearing or being replaced Shadows breaking the logic of the light source Shadows falling out of sync with the characters’ movements The wall projection suddenly becoming pure special-effects animation Unmotivated modern subtitles or UI VII. Shot Structure 0–5s|Shot One|Wide Shot, Locked-Off Camera + Extremely Slow Dolly In Visual Task Establish the complete spatial relationship and the misleading atmosphere of a “final battle.” Visual Content In the narrow courtyard alley of an old urban district after rain: Rough white stucco walls Wet stone pavement Warm yellow streetlights Overhead electrical wires Shadows of bamboo leaves Drifting thin mist Reflections in puddles One bicycle Character A and Character B appear in the same space. Their enormous shadows cast by the streetlight onto the white wall look like two ancient masters about to engage in a decisive battle. Cinematography Requirements Wide shot Locked-off camera Extremely slow dolly in Stable image, 24fps cinematic motion Establish the space first, then establish the shadow relationship Performance Requirements The real-world movements of the two women must not be exaggerated; they must remain restrained, steady, and precise. However, the enormous shadows on the wall must convey the powerful aura of two masters facing off from the very beginning. 5–10s|Shot Two|Cowboy Shot / Medium Shot|Smooth Lateral Tracking Visual Task Place the “small movements in reality” alongside the “epic battle on the wall.” Visual Content The same Sword Immortal Sister wearing a white embroidered silk hanfu swings the same silver longsword. The same Bicycle Sister spins the front wheel of the same bicycle and rapidly turns the handlebars. Key Requirements The real-world movements of both women must remain: Small Precise Restrained They must not look like they are truly engaged in a violent fight However, the shadows projected onto the white wall gradually transform into: A gigantic silver dragon A phoenix-like colossal beast made of circular wheels The two creatures engage in an astonishingly powerful battle across the wall. Background Lock Throughout the entire shot, the following must remain unchanged: The same white wall The same streetlight The same bicycle The same thin mist The same electrical wires The same puddles Sound Requirements The following sounds must be precisely synchronized and clearly layered: Sword ringing Sound of bicycle spokes spinning Rustle of clothing Deep drumbeats Sound of shadows colliding 10–15s|Shot Three|Close-Up Progressing to an Extreme Close-Up Visual Task Complete the twist and redefine the purpose of the entire “battle.” Visual Content The epic shadow battle suddenly stops. A child on the second floor, not visible in the frame, calls out from off-screen through the window: “姐姐,来个小兔子!” The same two women instantly freeze and exchange an extremely serious look. Then: Character A silently raises two fingers behind Character B’s head Character B folds both hands into the shape of tiny rabbit ears Focus Shift The camera focus shifts to the white wall. The giant dragon and phoenix that had just covered the entire wall have disappeared, leaving only a ridiculously small rabbit shadow hopping across the wall. The child laughs happily off-screen. Closing Dialogue Extreme close-up of the Sword Immortal Sister suppressing a smile. The same Bicycle Sister says quietly: “这个比龙难。” VIII. Performance Principles General Principle All performances must be restrained No exaggeration No mugging or deliberate ugliness Do not rely on falling down or embarrassing mishaps for the comedy The comedy comes from the contrast between serious performances and the reversal of purpose Character A Always serious Has a deadpan sense of humor Suppresses a smile at the end Character B Maintains the same tacit coordination Quietly complains at the end in a natural, understated, funny way Child’s Off-Screen Voice Never appears on screen The entire reversal is completed through sound alone Natural, cheerful, innocent tone IX. Cinematography and Visual Control Cinematography Style 16:9 landscape 24fps Arri Alexa cinematic camera quality Sharp, clear image Fine film grain Natural volumetric light Clearly defined spatial depth Camera Principles Shot One: Wide shot, locked-off camera + extremely slow dolly in Shot Two: Cowboy shot / medium shot, smooth lateral tracking Shot Three: Close-up progressing to an extreme close-up Prohibited: Rapid aerial shots Showy orbiting camera moves Uncontrolled drifting Game-like camera movement AI morph transitions Unmotivated camera jumps X. Continuity Requirements The following must remain consistently stable throughout: Character A’s identity, face, clothing, and longsword Character B’s identity, face, clothing, and bicycle The same streetlight The same white wall The same alley structure The same puddles and wet stone pavement The same nighttime color-temperature structure The direction and geometric relationship of the projected shadows Mandarin lip-sync The continuous physical relationships among the longsword, bicycle, clothing, hair, mist, and reflections XI. Final Video Requirements Strict total duration: 15 seconds 16:9 landscape Three clean shots Native synchronized ambient sound and action sound effects Accurate Mandarin lip-sync No subtitles generated on screen No watermark No random text XII. Negative blurry, bad quality, low quality, low resolution, noisy, jpeg artifacts, watermark, text, subtitles, error; deformed, mutated, bad anatomy, poorly drawn hands, bad composition, out of frame, disfigured; inconsistent character, changing clothes, face morphing, hairstyle change, background shift, glitching cuts, disappearing props, unstable shadow geometry, fake shadow projection, shadow detached from light source, bicycle disappearing, sword morphing, wall deformation, exaggerated acting, slapstick overacting, random fantasy effects, random CGI creatures, modern interface elements
Seedance 2.0|15秒短片《影戏》 一、生成目标 生成一段完整连续的 15 秒电影级中国都市仙侠短片,片名为:《影戏》 整体为: 电影级写实质感 融合 中国传统皮影美学 结合 默片式空间喜剧 表演克制、冷面 强调 暖黄色钠灯 与 冷蓝夜色 的色温对比 具有明确的前景、中景、背景纵深 最后 5 秒完成反转,让观众重新理解前面整场“大战”的真实用途 二、核心创意 这不是一场真正的生死决斗。 前 10 秒,观众会以为两名女子在进行一场极具史诗感、带有东方奇幻气质的巨大影子对决。 最后 5 秒才揭示: 她们其实是在用真实人物动作与投影关系,给楼上看不见的小孩表演一场“影戏”。 反转成立方式必须是: 前面所有夸张感都来自 墙面影子 两位角色现实中的动作始终小幅、精准、克制 后面通过一句画外童声重新定义整场动作的真实用途 喜剧不来自任何一方出丑 喜剧来自“史诗级气势”与“儿童游戏需求”之间的落差 三、工作流与技术倾向 本段适合按 Seedance 2.0 的多模态参考思路组织: 支持图片、音频、视频作为参考 强调对 人物表演、灯光、阴影、摄影机运动 的导演级控制 这类依赖 光影关系、动作节奏、空间影像反转 的双人表演尤其适合 重点强化: 双人身份稳定 阴影几何关系稳定 长剑与自行车连续性 墙面投影逻辑稳定 普通话口型同步 现实动作与影子动作的清晰因果关系 四、人物身份锚点 角色 ID A|@图片1|剑仙姐 始终保持同一个人物身份与外观: 25–30 岁东亚女性 椭圆脸 自然白皙皮肤 锐利深棕色眼睛 黑色长直发 高挑纤细身材 白色布靴 飘动的白色刺绣真丝汉服 半透明宽袖 银色腰饰 玉簪 同一柄银色长剑 人物气质: 冷静 克制 认真到近乎一本正经 即使在做“影戏”,也像在处理真正重要的事 角色 ID B|@图片2|单车姐 始终保持同一个人物身份与外观: 25–30 岁东亚女性 富有表情的脸 棕色马尾 固定身材比例 黄色外套 蓝色牛仔裤 白色运动鞋 固定饰品 同一辆自行车 人物气质: 配合默契 同样认真 反应克制 结尾低声吐槽形成冷面笑点 五、环境设定 场景 雨后的老城区狭窄院落巷道。 必须清楚建立以下空间元素: 粗糙白色灰泥墙 湿润石板路 暖黄色街灯 头顶电线 竹叶阴影 流动薄雾 积水倒影 一辆自行车 光线关系 整体采用: 暖黄色钠灯作为主要投影光源 冷蓝夜色作为环境底色 墙面影子必须清楚、巨大、可读 现实人物体量正常,墙上影子体量夸张 投影几何要真实可信,不能乱飘或变成无来由特效 六、环境运动原则 背景必须持续鲜活,但在叙事上保持中立。 允许存在并持续运动的元素: 薄雾自然流动 头顶电线轻微静态存在或轻微振感 竹叶影轻微晃动 地面积水与湿润石板倒影持续变化 人物衣料、头发、自行车细节随动作自然响应 必须禁止: 背景突然切换 白墙变形 街灯位置跳变 自行车消失或替换 影子脱离光源逻辑 影子与人物动作不同步 墙面投影忽然变成纯特效动画 无理由新增现代字幕或 UI 七、镜头结构 0–5s|镜头一|全景固定机位 + 极缓慢推轨 画面任务 建立完整空间关系与误导性的“决战氛围”。 画面内容 雨后的老城区狭窄院落巷道中: 粗糙白色灰泥墙 湿润石板路 暖黄色街灯 头顶电线 竹叶阴影 流动薄雾 积水倒影 一辆自行车 角色 A 与角色 B 出现在同一空间中。 她们被街灯投射在白墙上的 巨大影子,看起来像两名即将决战的古代高手。 摄影要求 全景 固定机位 极缓慢推轨 画面稳定,24fps 电影运动质感 先建立空间,再建立影子关系 表演要求 现实中的两人动作不能夸张,必须克制、稳、精确。 但墙面上的巨大影子要从一开始就带有强烈“高手对峙”的气势。 5–10s|镜头二|牛仔景 / 中景|平稳横向跟拍 画面任务 把“现实中的小动作”与“墙上的史诗大战”并置。 画面内容 同一个穿白色刺绣真丝汉服的剑仙姐,挥动同一柄银色长剑。 同一个单车姐,旋转同一辆自行车的前轮,并快速转动车把。 关键要求 两个人现实中的动作始终: 小幅 精准 克制 不要像真正大打出手 但投射到白墙上的影子却逐渐变化成: 一条巨大的银色长龙 一只由圆形车轮构成的凤凰状巨兽 二者在墙面上展开气势惊人的搏斗。 背景锁定 整个镜头中必须始终保持: 同一面白墙 同一盏街灯 同一辆自行车 同样的薄雾 同样的电线 同样的积水 音效要求 以下声音必须精确同步并层次清楚: 剑鸣 自行车辐条旋转声 衣料抽动声 低沉鼓点 影子碰撞声 10–15s|镜头三|特写进入极近特写 画面任务 完成反转,重新定义整场“大战”的用途。 画面内容 史诗般的影子大战突然停止。 二楼没有出现在画面中的小孩,用窗口外画音喊道: “姐姐,来个小兔子!” 同样两位女性瞬间僵住,极其严肃地互相看一眼。 随后: 角色 A 默默把两根手指举到角色 B 头顶后方 角色 B 则将双手折成小小的兔耳形状 焦点转移 镜头焦点转移到白墙。 刚才遮满整面墙的巨龙与凤凰已经消失,只剩下一只 小得荒谬的兔子影子 在墙面蹦动。 小孩在画外开心大笑。 收尾对白 极近特写剑仙姐强忍笑意。 同一个单车姐低声说道: “这个比龙难。” 八、表演原则 总原则 全员表演必须克制 不浮夸 不扮丑 不靠跌倒、出糗制造笑点 喜剧来自严肃表演与用途反转之间的落差 角色 A 始终认真 有冷面幽默感 结尾强忍笑意 角色 B 同样保持配合默契 结尾低声吐槽,要自然、低调、好笑 小孩画外音 不出现画面 只通过声音完成整段反转 语气自然、开心、天真 九、摄影与视觉控制 摄影风格 16:9 横屏 24fps Arri Alexa 电影机质感 画面锐利清晰 细腻胶片颗粒 自然体积光 空间纵深明确 摄影机原则 镜头一:全景固定 + 极缓慢推轨 镜头二:牛仔景 / 中景,平稳横向跟拍 镜头三:特写进入极近特写 禁止: 快速航拍 花哨环绕 失控漂移 游戏感镜头 AI morph 转场 无来由镜头跳变 十、连续性要求 必须始终稳定保持: 角色 A 身份、脸、服装、长剑 角色 B 身份、脸、服装、自行车 同一街灯 同一白墙 同一巷道结构 同一地面积水与湿润石板 同一夜间色温结构 阴影投射方向与几何关系 普通话口型同步 长剑、自行车、衣料、头发、薄雾与倒影的连续物理关系 十一、成片要求 严格总时长:15 秒 16:9 横屏 三个干净镜头 原生同步环境声与动作音效 普通话口型准确 不生成画面字幕 不出现水印 不出现随机文字 十二、Negative blurry, bad quality, low quality, low resolution, noisy, jpeg artifacts, watermark, text, subtitles, error; deformed, mutated, bad anatomy, poorly drawn hands, bad composition, out of frame, disfigured; inconsistent character, changing clothes, face morphing, hairstyle change, background shift, glitching cuts, disappearing props, unstable shadow geometry, fake shadow projection, shadow detached from light source, bicycle disappearing, sword morphing, wall deformation, exaggerated acting, slapstick overacting, random fantasy effects, random CGI creatures, modern interface elements
- The prompt is in Chinese: paste it as is — the model understands 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 →











