Physics Experiment Educational Video

REFERENCE / IDENTITY LOCK: Use the attached reference image as the sole facial identity reference for the subject. Preserve the same recognizable facial structure, facial proportions, eye shape, eye spacing, eyebrows, nose shape, lips, cheek structure, jawline, chin, and natural skin characteristics throughout the entire video. Maintain strict facial identity consistency from the first frame to the last frame. Do not redesign, beautify, age, de-age, or reinterpret the face. SUBJECT & STYLING: Create a 15-second cinematic hyper-lapse selfie travel video featuring the same young woman traveling across 30 iconic destinations around the world in 2026. Keep the same person, same age appearance, same natural body proportions, same long brown hairstyle, and same cheerful travel-vlogger personality in every scene. Each destination has a distinct stylish outfit appropriate to the location, while the subject's face and physical identity remain unchanged. VIDEO STRUCTURE: Exactly 30 scenes. Each scene lasts exactly 0.5 seconds. Total duration: exactly 15 seconds. Use hard cuts between every scene, precisely synchronized to the music beat. One destination per shot only. Do not blend, morph, or combine locations. Do not transition gradually between outfits or environments. CAMERA & MOTION: Handheld selfie-stick perspective, close selfie framing, natural arm-length camera distance, wide-angle travel-vlogger lens, realistic handheld micro-movements, energetic movement, authentic perspective distortion, subtle motion blur, natural walking and sightseeing movement. Keep the subject's face clearly visible and recognizable in every shot. LOCATIONS: 1. Eiffel Tower — Paris 2. Shibuya Crossing — Tokyo 3. Times Square — New York City 4. Colosseum — Rome 5. Pyramids of Giza — Egypt 6. Christ the Redeemer — Rio de Janeiro 7. Big Ben — London 8. Sydney Opera House — Sydney 9. Grand Palace — Bangkok 10. Taj Mahal — India 11. Great Wall — China 12. Burj Khalifa — Dubai 13. Hagia Sophia — Istanbul 14. Grand Canal — Venice 15. Machu Picchu — Peru 16. Acropolis — Athens 17. Sagrada Família — Barcelona 18. Traditional windmills — Amsterdam 19. Gyeongbokgung Palace — Seoul 20. Marina Bay Sands — Singapore 21. Santorini — Greece 22. Petra — Jordan 23. Neuschwanstein Castle — Germany 24. Golden Gate Bridge — San Francisco 25. Niagara Falls — Canada 26. Mount Fuji — Japan 27. Angkor Wat — Cambodia 28. Swiss Alps — Switzerland 29. Cappadocia — Türkiye 30. Maldives POSE & EXPRESSION: Give each scene one natural, distinct travel-vlogger action: waving, peace sign, pointing toward the landmark, thumbs-up, namaste, finger heart, laughing, surprised reaction, playful sightseeing gesture, or brief dance movement. Keep gestures anatomically natural and physically believable. ENVIRONMENT: Make every landmark immediately recognizable within its 0.5-second shot. Include realistic local surroundings, natural tourist crowds, authentic architecture, believable atmospheric conditions, and subtle environmental movement. Keep backgrounds detailed but never allow them to obscure the subject's face. LIGHTING: Use physically believable location-specific lighting appropriate to each destination and time of day. Natural sunlight, realistic shadows, accurate reflections, atmospheric depth, and consistent exposure. Avoid artificial glow, excessive HDR, or impossible illumination. MATERIAL & SKIN REALISM: Natural skin texture, visible realistic pores and subtle facial detail, physically accurate hair strands, realistic fabric behavior, natural hands and fingers, believable clothing materials, authentic reflections and environmental interaction. PHOTOGRAPHIC FINISH: Ultra-realistic cinematic travel photography, premium travel-vlog aesthetic, natural color science, realistic depth of field, authentic wide-angle perspective, subtle cinematic contrast, realistic motion blur, physically accurate environmental rendering, no artificial CGI appearance. IDENTITY & CONTINUITY CONSTRAINTS: The subject must remain the exact same person throughout all 30 scenes. No face morphing, facial redesign, identity drift, gender change, age change, hairstyle change, body-shape change, skin-tone inconsistency, facial proportion changes, duplicated person, outfit blending between scenes, accessory switching between frames, unnatural hand deformation, or inconsistent anatomy. 16K: 15360 × 8640 ≈ 132.7 million pixels, high-resolution professional quality. NEGATIVE CONSTRAINTS: face morphing, identity drift, different person, facial distortion, altered eyes, altered nose, altered jawline, age change, hairstyle change, body deformation, extra fingers, missing fingers, fused hands, duplicated limbs, warped landmarks, merged locations, blended outfits, floating objects, unrealistic crowds, CGI look, plastic skin, excessive beauty retouching, oversharpening, flickering, unstable exposure, unnatural motion, camera glitches, text, logos, watermarks.

Use the attached THE KEBAB MAKER storyboard as exact shot reference, but render in a stop-motion claymation style — soft matte clay-like skin with a subtle glossy sheen, visibly textured fabric clothing (woven/felt texture on the apron and chef coat), exaggerated elongated features (long thin nose, thick bushy mustache, small round glasses), stubby expressive hands, and gentle handcrafted stop-motion movement (slightly soft, weighted, tactile motion rather than smooth CGI). Same middle-aged Turkish kebab maker, thick dark mustache, red and white striped textured-fabric apron, outdoor Istanbul street stand, Galata Tower background, warm golden afternoon light throughout. Shot sequence: 1.Wide shot. Kebab maker behind his outdoor stand, Galata Tower visible behind him, doner spit rotating and glistening with a soft claymation sheen. He looks directly at the camera, points at the spit behind him with one thumb and says "Turkish Kebab!" with a big proud grin, exaggerated stop-motion charm. 2.Close-up. Both textured clay hands opening a warm flatbread — bread has a soft doughy clay texture, steam rising from inside, pillowy interior revealed. Warm golden bread in the afternoon light. 3.Dramatic close-up. Vertical doner spit filling the entire frame — deeply caramelized layered meat with a tactile, sculpted claymation texture, rotating slowly, heat element glowing orange behind it, steam rising. No character needed. 4.Wide medium with character visible. Kebab maker holds open flatbread in one clay hand below the spit, long knife in the other shaving meat in one confident downward stroke — thin caramelized strips falling directly into the open bread, all rendered with soft stop-motion weight. 5.Extreme close-up. Textured clay hand sprinkling deep red sumac powder generously over the shaved meat — vivid crimson powder catching the warm light as it falls, dusting every piece with a slightly grainy claymation texture. 6.Close-up hero. Thick white garlic sauce poured in a generous arc over the filling like soft clay icing, cascading over the meat. Then a zigzag of vivid red chili sauce over the top. 7.Close-up dramatic. Both clay hands wrapping the flatbread tight and fast — foil paper wrapped around the bottom with a slightly crinkled tactile texture. One confident twist. Done. 8.Wide warm. Customer on the Istanbul street takes a massive first bite — eyes close in an exaggerated claymation expression, pure satisfaction, sauce on their lip. Kebab maker behind the stand, arms crossed, deeply proud, big round glasses catching the light. Galata Tower glowing in the warm background. Camera: Wide for opener and final bite, close-up for bread and sumac, dramatic close-up for spit, wide medium for shave into bread, close-up hero for sauce, close-up dramatic for wrap. Style: Stop-motion claymation aesthetic — soft matte clay skin with subtle gloss, visibly woven fabric textures on clothing, exaggerated elongated facial features, tactile handcrafted movement. Warm Istanbul afternoon light throughout. Deep caramelized meat browns, vivid red sumac, white garlic sauce, red chili sauce, golden flatbread, red and white striped textured apron, Galata Tower background. Nothing dark, nothing flat — everything reads as a physical handcrafted set. Sound: Upbeat Turkish street music — energetic and warm — from first frame to last. Continuous, no breaks. Natural sounds underneath — kebab maker saying "Turkish Kebab!" at the opener, bread opening with steam, spit rotating, knife shaving meat, sumac dusting, sauce pouring, foil wrapping, one confident twist, massive first bite crunch, satisfied exhale. Music never stops. Goal: 12 seconds. One spit. One perfect wrap. Bold, sizzling, claymation-crafted, and straight from Istanbul.

Create a 10-second ultra-photorealistic cinematic architectural commercial following the attached 12-shot storyboard exactly. One identical luxury residential complex throughout. Preserve its exact geometry, proportions, stone facade, graphite metal, panoramic glass, balconies and windows from miniature model to full-scale building. First-person male hands always wear identical matte-black gloves. No text, logos or UI. 0.0–0.6 — SHOT 1: First-person POV, slightly top-down. Dark monolithic graphite table in a black minimalist studio. Two black-gloved hands enter frame and position the architectural base precisely in the center. Slow controlled push-in. 0.6–1.2 — SHOT 2: EXTREME MACRO. Fingers pick up a heavy miniature raw-concrete block and insert it precisely into the base. Show porous concrete texture, weight and tight physical fit. 1.2–1.8 — SHOT 3: LOW MACRO. Hand installs a graphite-metal structural element. As it locks into place, structural columns mechanically rise vertically from the base and stop precisely in position. Real mechanical assembly, no magic. 1.8–2.5 — SHOT 4: POV CLOSE-UP. Gloved fingers move upward along the structure. Floors, concrete slabs and structural frames rapidly assemble sequentially upward directly following the hand movement. Camera tilts upward with the growing building. 2.5–3.2 — SHOT 5: Three rapid tactile macro cuts: fingers install a natural stone facade slab → CUT → graphite brushed-metal panel locks into place → CUT → transparent panoramic glass panel slides precisely into its frame. Extreme material detail. 3.2–4.0 — SHOT 6: SIDE MACRO tracking along facade. Gloved hand moves horizontally beside the structure while real facade components rapidly lock into position behind it: natural stone → graphite metal → panoramic glazing → glass balconies. Building becomes architecturally complete. 4.0–4.7 — SHOT 7: EXTREME MACRO. Two fingers hold the final transparent glass panel and carefully insert it into the remaining facade opening. Precise physical CLICK. Immediately afterward warm amber apartment lights switch on sequentially floor by floor behind the glass. 4.7–5.4 — SHOT 8: Wider POV. Both hands lift the fully completed miniature residential complex from the graphite table. Hands release it carefully. The completed building remains suspended motionless above the table. Hands move out of frame. Camera immediately begins orbiting. 5.4–6.3 — SHOT 9: ONE CONTINUOUS SEAMLESS SCALE TRANSITION, NO CUT. Camera accelerates into a smooth orbit around the suspended miniature. Keep the building centered. During the orbit, camera perspective, depth of field and architectural detail progressively shift from macro photography to real architectural scale. The table disappears below/outside frame, studio darkness naturally becomes evening sky and distant city lights. The building grows until it completely dominates frame. Never visibly morph the architecture; only change perceived scale. 6.3–7.3 — SHOT 10: Continue the SAME orbit without resetting camera direction. Now unquestionably a real full-scale luxury residential complex at blue-hour evening. Dynamic low aerial orbit around its corner. Pass close to natural stone, graphite structural frames, balconies and panoramic glass. Warm occupied apartments visible behind windows. Monumental realistic scale. 7.3–8.3 — SHOT 11: Camera transitions from orbit into a fast smooth flight parallel to the facade. Select one warmly illuminated panoramic apartment window and accelerate directly toward it. Window rapidly fills frame. Seamlessly pass through the glass with no cut into the apartment. 8.3–10.0 — SHOT 12: Continue the same camera movement inside a real luxury apartment. Warm contemporary interior: natural stone, dark wood, designer furniture, soft amber architectural lighting, floor-to-ceiling windows. Four stylish adult friends naturally talk, smile and laugh together, unaware of camera. Smooth lateral dolly past them, ending with people foreground/midground and panoramic evening city plus the complex architecture visible behind. VISUAL PROGRESSION: cold graphite studio → tactile assembly → finished illuminated miniature → seamless scale shift → monumental real architecture → warm human life. Luxury watch-commercial precision, automotive-commercial lighting, architectural-film realism. Shallow macro DOF gradually becomes realistic architectural depth. Physically accurate concrete, stone, brushed metal and glass. Strict continuity of one building. No toy/plastic appearance, no holograms, particles, magical construction, deformation, architecture morphing, floating components, extra tools, text or logos., 9:16

Create one uninterrupted 15-second vertical 9:16 handheld smartphone recording. @image1 is ONE complete 3-column by 4-row storyboard, read left to right across each row and then downward. Its twelve pictures are chronological action checkpoints within this single take, NOT twelve shots. Use the sheet as the visual authority for the people, field, weather, enormous tornado and cotton-candy transformation. Render a live full-screen scene from the phone's lens, never the sheet itself, its borders, a collage, a still-image slideshow or a pan across pictures. At the beginning, the phone operator films his friend beside the field with the enormous gray tornado looming in view. Over roughly the first five seconds, keep the friend's pink sugar visible before he throws it. Follow the airborne sugar as the wind carries it into the same funnel; only after the sugar reaches it do pink fibrous bands begin wrapping around the gray tornado. The friend naturally leaves the left edge as the operator reframes toward the funnel. He is not the person who gets lifted. Over roughly the next five seconds, the fibrous bands thicken through the pictured intermediate states until the same gigantic tornado has become a towering mass of fluffy pink spun sugar. Preserve its immense scale against the field. The operator walks closer while continuing to film: foreground grass passes below the lens, the funnel grows through real approach and parallax, and wind shakes the handheld view. Keep the action and transformation visibly continuous; this is not a zoom between still poses. Airborne grass and dust intensify as he gets too close. During roughly the final five seconds, the stronger pull jerks the operator and the phone forward and upward together. The phone stays in his hand throughout. Its view pitches downward during the lift, revealing his own two dangling legs and shoes as the field visibly drops farther beneath him. Hold that height-and-feet proof long enough to understand that the FILMER is being sucked up, then carry the same viewpoint into the deep pink cotton-candy spiral shown in the final picture. End inside that moving spiral. No outside view of the filmer, no selfie, no detached or dropped phone, no drone viewpoint, no cut or teleport into the tornado, and no extra rescue or aftermath. Maintain the spontaneous feel of an amateur phone recording: natural handheld sway and footstep bob, increasing gust-driven shake, modest autofocus and exposure response, and continuous real spatial motion. Keep the essential throw, transformation and dangling feet readable in portrait framing. No slow motion, cinematic coverage, editorial cuts, transitions, freeze frames, captions, graphics or readable text. Native location audio only: wind buffeting the phone microphone, rustling grass, spontaneous laughter at the transformation, and a startled nonverbal yell during the lift. No intelligible speech

Ultra-realistic cinematic scene of a young woman standing inside a crowded subway train, quietly looking at her smartphone while passengers sit and stand around her. She wears a dark green and burgundy hoodie over a white shirt, with a backpack over one shoulder. Warm fluorescent train lighting, realistic urban atmosphere, shallow depth of field, cinematic composition, natural skin texture, soft background bokeh, subtle film grain, documentary-style photography, 35mm lens, f/1.8, high detail, photorealistic, vertical.

一、任务目标生成一段完整连续的 15 秒中国仙侠电影短片。核心喜剧主线只有一条，并且要让观众 第一遍就立刻看懂：一名敌方剑客故意挑拨师姐与小师妹，篡改师姐原本说过的话，让小师妹误以为师姐认为她“三招必败”。真正反转是：她们从来没有讨论“小师妹能不能打赢”，而是在讨论这个敌人究竟能撑几招。二、整体风格整体必须同时具备以下气质：电影级写实质感纯古风中国仙侠美学真正史诗级高手对峙的镜头语法一本正经的冷面喜剧默片式反应节奏港式动作喜剧的利落效率清楚的“三拍递进” 明确的铺垫与回收 Arri Alexa 电影机质感稳定清晰的面部微细节细腻胶片颗粒自然体积光喜剧必须克制，不夸张，不胡闹，不靠人物变蠢。三、人物身份锁定角色 ID A｜@图片 1｜剑仙师姐始终保持同一个人物： 25–30 岁东亚女性身材高挑纤细椭圆脸深色杏眼黑色长发半挽白玉簪固定同一套白色刺绣真丝汉服银色腰封玉佩白色布靴唯一一柄银色长剑角色 ID B｜@图片 2｜小师妹始终保持同一个人物： 20–25 岁东亚女性身形娇小圆润灵动的脸黑发编辫同一套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其他角色敌方剑客一名敌方剑客负责挑拨、误导和最终出手不可抢走喜剧情绪中心年长师父同一个年长师父只在关键时刻补最后一刀必须保持安静、克制、一本正经四、环境 DNA 与空间原则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。正式构图前，先无声整合并统一以下内容：真实地形建筑语言材质植被水体天气山雾主光方向倒影空气纵深在此基础上，重新规划出一个：本轮独有、完整统一、可信的新空间。环境规则背景中的以下元素始终自然鲜活：风水体植被云层远处普通弟子空间环境声但必须满足：背景绝对中立不制造包袱不制造反转不主动推动剧情不替人物表达情绪五、三镜头结构镜头 1｜0–5s 景别全景或远景画面内容同一个剑仙师姐同一个小师妹一名敌方剑客一名年长师父四人位于统一空间中，地理关系清楚稳定。敌方剑客站在两人正前方，年长师父只在数步之外安静存在。剧情与对白敌人故意看着小师妹挑拨道： “你师姐亲口说，你与我交手，三招必败。” 同一个小师妹非常缓慢地转头看向师姐，问： “三招？” 同一个白衣剑仙没有任何紧张，只平静纠正： “我说的是他。” 镜头重点第一拍要先让观众以为矛盾是“师姐看不起小师妹” 小师妹转头这一瞬必须清楚师姐的纠正必须平静、确定、不带解释欲这一句要直接翻第一层误会镜头 2｜5–10s 景别中景或牛仔景连续性要求保持以下内容全程稳定：同样两名女子同一个敌人同一个师父同样服装同样长剑完全一致的地理空间剧情与对白小师妹立刻重新看向敌人。这一次她不是生气，而是真的有些惊讶，说： “三招？师姐，你也太给面子了。” 敌人刚才得意的表情当场僵住。浅景深后方，同一个师父从头到尾没有插话，只安静地抬起一根手指。敌人发现后，恼火问： “你什么意思？” 镜头焦点转移到师父。师父保持绝对认真，只回答： “一招。” 完整停顿半拍。敌人终于彻底破防，大喊： “欺人太甚！” 随后主动拔剑，向师姐正面冲去。镜头重点第二拍完成喜剧递进 “三招”升级成“一招” 师父这一刀必须极干停顿要留足，让观众和敌人同时意识到自己成了被预测的对象敌人破防后立刻进入动作段镜头 3｜10–15s 景别特写或极近特写连续性要求保持以下内容稳定：同一个师姐同一个小师妹同一个敌人同一个师父服装、长剑、站位、空间关系全程一致动作设计同一个师姐始终保持放松，甚至不把自己的银色长剑完全拔出。她只完成一个：非常清楚物理合理极短促极精准的动作组合：侧身闪避敌人正面冲击用仍然带鞘的同一柄银色长剑短促敲击敌人持剑手腕顺势借敌人的前冲力量改变其重心结果：敌人的长剑旋转飞脱安全插入旁边地面敌人自己单膝落地全程不得出现血腥伤害动作后的对白全场安静。同一个小师妹慢慢举起一根手指，认真说道： “师父猜得真准。” 同一个剑仙师姐低头观察败下阵来的敌人，平静回答： “其实半招。” 镜头焦点再次转向师父。师父一本正经说道： “我已经给他留面子了。” 结尾画面极近特写中：敌人依次看看师父、师姐、小师妹发现自己一句都反驳不了小师妹用力咬住嘴唇憋笑师姐仍然非常正式地朝他行一个武者承让礼精准切在：敌人彻底无语的表情上。六、表演原则师姐始终平静不炫耀不解释过度强者松弛感要稳定喜剧成立依靠认真，而不是玩笑腔小师妹反应要快情绪不夸张从“误会”到“听懂”再到“补刀”，节奏要清楚憋笑感要真实克制敌人前半段必须真心自信中段逐渐僵住后段破防但不能演成小丑笑点来自后知后觉，不来自愚蠢师父台词极少每句都像盖棺定论必须一本正经越认真越好笑七、动作与镜头要求动作身体力学清晰自然敌人冲击要真有气势师姐应对要极省、极准、极短最终胜负必须一眼看懂不要复杂炫技连招不要拖泥带水不要光污染特效不要随机剑气摄影要求 16:9 横屏严格三个连续清晰镜头运镜干净克制前景、中景、远景保持自然视差镜头服务人物关系与包袱回收不做花哨炫技运动八、声音与同步要求原生同步普通话对白口型精准喜剧停顿明确视线关系准确拔剑、碰撞、长剑落地、归位等声音清楚可辨环境声持续自然存在不生成字幕重点利用 Seedance 2.0 的能力：多模态参考连续性多人物复杂运动稳定性导演级运镜控制 15 秒多镜头同步音视频输出能力九、严格避免 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 额外避免：字幕过度夸张表演无意义搞笑表情复杂难懂的反转动作软绵绵花哨慢镜随机法术环境制造笑点配角抢戏敌人突然降智十、最终效果目标整段短片的观感应该是：先制造“师姐似乎看低小师妹”的表面误解很快翻转成“原来被评价的是敌人” 最后再用动作和师父补刀，把包袱彻底坐实最终关键词：仙侠、冷面喜剧、高手对峙、三拍递进、误导反转、师徒补刀、半招制敌、电影级写实。

人物外形严格参考图一，场景严格参考图二，神龙参考图三。核心指令：大量瞬移式位移、残影、雷光与水爆。摄影机大量使用贴水超低机位仰拍、垂直90°鸟瞰、极端俯冲、鱼眼广角、高速环绕、翻滚镜头和近距离掠镜，强化人与巨龙的尺度差异。0至1.2秒尾帧直接爆发，第0秒神龙已在剑客身后咆哮，双方无停顿直接发动攻击，巨龙张口向下猛扑，剑客同时从半跪状态爆发弹起，白衣与水面同时被冲击波掀开，手中裸剑瞬间亮起刺目冷白雷电，一剑逆势向上迎击龙首，镜头几乎贴着水面从剑客脚下以超低机位高速向上仰拍，龙首巨大到占据整个天空，随后摄影机瞬间穿过剑光，水面放射状炸裂，雷光过曝一瞬，剑锋拖出极细而锋利的白色光轨。1.2至2.5秒第一次正面对撞，剑锋与龙首正面碰撞无停顿，撞击瞬间形成环形冲击波掀起浅水，剑客借反作用力凌空翻身从龙首侧面高速擦过，连续三次极快斩击压缩成近乎三个闪光瞬间，只看见起点、雷光轨迹和终点，运镜从极端低角度仰拍转为随人物一起翻转180°，再突然切入90°垂直俯拍，从正上方看巨龙身体与剑客的白色雷光在黑色水面形成巨大螺旋。2.5至4秒龙身追杀，神龙巨大身体高速盘旋如巨浪横扫峡谷，剑客贴着龙身疾驰，以连续数次雷电式瞬移移动，每次脚尖短暂触碰水面便炸开一圈白色水花，巨龙尾部突然横扫，剑客在最后一瞬向后折腰躲过，随后身体贴着龙尾反方向旋转一剑横斩，运镜采用24mm超广角高速侧跟，龙尾直接擦过镜头产生强烈遮挡，摄影机被冲击力撞飞般高速翻滚。4至5.5秒极端俯冲，剑客借龙身跃至峡谷高处，身体瞬间成为黑暗空间中的极小白点，下一瞬头朝下持剑垂直坠落，雷电沿剑身迅速扩散，整个人像一道白色落雷直冲龙首，运镜从峡谷顶部近乎90°垂直向下俯拍，摄影机与剑客一起极速坠落，地面、水面、龙身疯狂放大，最后0.2秒镜头突然切至水面超低机位，剑客从镜头上方高速掠过。5.5至7秒贴脸高速交锋，神龙转头咬向剑客，镜头极近距离位于龙口前方，巨大獠牙几乎擦过镜头，剑客侧身瞬移避开，白衣从龙嘴边掠过，同时反手一剑切过龙颈，龙身瞬间化成巨量水流爆开却立即重新聚合，运镜采用14至18mm极广角呈现强烈近大远小透视，镜头高速穿越龙口、剑客与飞散水珠，绝不使用慢动作。7至8.5秒疯狂环绕对决，神龙围绕剑客高速盘旋数周形成巨大水龙卷，剑客位于风暴中心连续闪现于不同位置，每一次闪现只完成一剑，左上、右侧、水面、半空四道冷白剑光最终组成一个巨大的交叉雷纹，摄影机高速360°环绕并不断改变高度，从贴水低机位突然升至高空再垂直俯冲回来。8.5至10秒人与龙迎面对冲，巨龙沿整个峡谷长度高速冲来体型压满画面，剑客站在浅水中央没有摆POSE，仅停留不到0.2秒随即握剑向前爆射，双方正面对冲，镜头突然转为超低机位正面构图，剑客只有画面中央极小的白色身影而巨大神龙覆盖整个背景，距离迅速归零，碰撞前一瞬剑客突然消失。10至11.5秒逆斩龙首，剑客已瞬移至神龙头顶上方，身体在空中倒转，双手握住没有剑鞘的雷电长剑从上向下完成一次极快极重的逆向劈斩，剑锋落下瞬间雷光从龙首一直贯穿巨大水体，运镜从龙首正下方向天空极端仰拍，白衣剑客逆光悬在巨大龙首上方仅存在不到半秒，随即镜头随着剑锋一起垂直砸向水面。11.5至13秒终极水爆，剑光触水，整个浅滩以剑客为中心形成巨大的圆形水爆，神龙身体被雷电切割成无数悬浮水滴短暂凝固在空气中，下一瞬全部爆散，运镜瞬间切换至极高空90°鸟瞰，黑色峡谷、圆形白色水爆与中央极小的白衣剑客形成强烈尺度对比。13至15秒骤然归静，节奏突然从极快降到极静，无数水滴落回河面，神龙彻底消散成薄雾与水流不再重新出现，剑客站在浅水中央背对或侧对镜头，右手自然垂下依然握着那柄裸露没有剑鞘的长剑，剑刃最后残留几缕微弱冷白雷电随后逐渐熄灭，他不收剑、不归鞘、不做任何额外动作，白色长袍与发丝被峡谷中的风轻轻吹动，镜头从人物身后缓慢拉远，最终重新展现图三那片巨大幽暗空旷的峡谷水域，水面只剩最后一圈涟漪。

Ridders op kangoeroes, tien seconden, filmkorrel inbegrepen.