日本寺庙与山道的旅行音乐视频

旅行风景 · Seedance 2.5 · Anissa · 2026-09-02

一条信息量很大的 30 秒提示词：服装时间表（7 秒和 16 秒换三套衣服）、约 20 个精确到小数秒的镜头，以及指定的标题文字叠加。作为带时间码的混剪模板，比随手想法更有参考价值。

提示词

创建一支 30 秒、16:9、写实电影感的日本旅行时尚音乐视频，始终由同一位年轻女性贯穿全片。保留她完全一致的身份、面部结构、长直黑发、妆容、耳环、身材比例、美甲和白色有线耳机。不要说话。24fps，180° 快门，真实的手持电影机运动，自然的皮肤/布料/头发物理效果，轻微胶片颗粒，具有氛围感的景深，真实人群，不要 CGI 感、失真、美颜磨皮、扭曲的手/脸、logo、商标或随机文字。服装：0–7s 深樱桃红短款尼龙夹克，黑色背心和长裤，白色运动鞋。7–16s 松绿色短款运动上衣，浅薄荷色工装长裤，白色单肩包，银/金首饰，白色运动鞋。16–30s 哑光黑轻薄夹克，白色滚边，黑色上衣/长裤和银色铆钉腰带。仅在 7s 和 16s 更换服装。分镜： 0–2.4s：日本城市公园，树后可见酒店建筑；她使用玫瑰金手机。2.4–4.2s：手和手机特写，带通用音乐播放器界面。4.2–7s：近景肖像，她调整耳机并微笑。 7–9.2s：寺庙祈愿区，木牌、灯笼和苔藓。9.2–11.4s：宽幅石阶，有游客；俏皮地迈步并抬膝。大号芥末黄标题：“JAPAN”。11.4–13.2s：她转身面向山景，头发自然摆动，真诚大笑。13.2–14.6s：香火特写，烟雾缭绕。14.6–16.2s：她穿过祈愿区。 16.2–18s：传统木桥，远处雾气缭绕的群山。18–19.6s：背面视角，俯瞰森林峡谷。19.6–21.4s：山景肖像。21.4–23.2s：森林小径，传统轿夫和工人。23.2–24.6s：戏剧性的日本群山穿云而出。 24.6–26.2s：她和一位女性朋友走在日本山间小路上；两人自然大笑。朋友：年轻东亚女性，长深色头发，浅薄荷色宽松 T 恤，黑色阔腿长裤，运动鞋。 26.2–27.4s：快速插入池塘/乌龟镜头，然后是坐在长满苔藓岩石上的日本猕猴。 27.4–30s：山间寺庙广场；两人即兴跳舞并大笑。28.2s 时出现芥末黄窄体标题：“NEVER SAY / NEVER”。轻微手持推进，最后定格在她举起双臂的画面，然后硬切黑场。环境：真实的日本公园、寺庙、杉树林、石板路、瓦顶、红色旗帜、灯笼、祈愿牌、山桥、缆车、雾气缭绕的悬崖和自然的日本人群。使用 35/50mm 人像镜头、24mm 广角、85–100mm 插入镜头。自然日光、温暖的寺庙灯光、冷调漫射的山地光线。音频：手机点击声、耳机咔哒声、寺庙环境音、脚步声、风铃声、香火环境音、山风、缆车机械声、桥梁吱呀声、水声和真实笑声。不要对白、旁白、演唱、歌词、字幕、说明文字，除非是 “JAPAN” 和 “NEVER SAY NEVER”。不要水印或 UI 叠层，除了短暂的手机界面。

译文来自来源仓库，原文附在下方。

原文（英文）

Create a 30-second, 16:9 photorealistic cinematic travel-fashion music video in Japan featuring the SAME young woman throughout. Preserve her exact identity, facial structure, long straight black hair, makeup, earrings, proportions, manicure, and white wired earbuds. No speaking. 24fps, 180° shutter, realistic handheld cinema-camera movement, natural skin/fabric/hair physics, subtle film grain, atmospheric depth, realistic crowds, no CGI look, distortion, beauty smoothing, warped hands/faces, logos, trademarks, or random text. WARDROBE: 0–7s deep cherry-red cropped nylon jacket, black tank and trousers, white sneakers. 7–16s pine-green cropped athletic top, pale-mint cargo trousers, white shoulder bag, silver/gold jewelry, white sneakers. 16–30s matte-black lightweight jacket with white piping, black top/trousers and silver-studded belt. Change outfits only at 7s and 16s. SEQUENCE: 0–2.4s: Japanese city park, hotel architecture behind trees; she uses her rose-gold phone. 2.4–4.2s: close-up of hands/phone with generic music-player interface. 4.2–7s: close portrait, she adjusts earbud and smiles. 7–9.2s: temple wishing area with wooden plaques, lanterns and moss. 9.2–11.4s: wide stone stairs with visitors; playful step and raised knee. Large mustard-yellow title: “JAPAN”. 11.4–13.2s: she spins toward mountain scenery, hair moving naturally, genuine laugh. 13.2–14.6s: incense close-up with smoke. 14.6–16.2s: she walks through the wishing area. 16.2–18s: traditional wooden bridge with misty mountains. 18–19.6s: rear view overlooking forest canyon. 19.6–21.4s: mountain portrait. 21.4–23.2s: forest trail with traditional passenger carriers and workers. 23.2–24.6s: dramatic Japanese mountains emerging through clouds. 24.6–26.2s: she walks with a female friend on a Japanese mountain path; both laugh naturally. Friend: young East-Asian woman, long dark hair, pale-mint oversized T-shirt, black wide-leg trousers, sneakers. 26.2–27.4s: quick pond/turtle insert, then Japanese macaque on mossy rock. 27.4–30s: mountain temple plaza; both spontaneously dance and laugh. At 28.2s, mustard-yellow condensed title: “NEVER SAY / NEVER”. Small handheld push-in, ending on her with arms raised, then hard cut to black. ENVIRONMENT: Authentic Japanese parks, temples, cedar forests, stone paths, tiled roofs, red banners, lanterns, wishing plaques, mountain bridges, cable cars, misty cliffs and natural Japanese crowds. Use 35/50mm portraits, 24mm wides, 85–100mm inserts. Natural daylight, warm temple lighting, cool diffused mountain light. AUDIO: phone taps, earbud click, temple ambience, footsteps, wind chimes, incense ambience, mountain wind, cable-car machinery, bridge creaks, water and genuine laughter. No dialogue, voiceover, singing, lyrics, subtitles, captions except “JAPAN” and “NEVER SAY NEVER”. No watermark or UI overlays except the brief phone interface.

分镜拆解

主体: 年轻女性按时间换三套衣服（樱桃红夹克、松绿上衣、黑色外套），后段有一位朋友一起出现在山路上。
动作: 她从都市公园走到寺庙、石阶、山景、缆车、桥、山路，最后与朋友一起跳舞，以硬切黑场结束。
镜头: 镜头时长 1.4-2.4 秒，并指定焦段（24mm 广角、35/50mm 人像、85-100mm 特写），结尾小幅推近。
风格: 24fps 手持电影机质感，带胶片颗粒，自然日光和寺庙暖光，两处黄色标题叠字。
声音: 手机点击声、寺庙环境声、风声、缆车机械声和笑声，没有对白和歌词。

改成你自己的

约 20 个定时镜头已经是 30 秒的上限，做 15 秒就只留 0-7 秒的第一套服装段和最后的舞蹈。
只允许两个标题字符串，保持简短（JAPAN、NEVER SAY NEVER），字形变形就直接删掉。
「只在 7 秒和 16 秒换衣服」这句防止随机换装，改时间码时要保留。

来源与许可

作者: Anissa
原帖: 在 X 上查看
收录于: renoise-ai/awesome-seedance-2-5-prompts
许可证: CC BY 4.0
改动说明: Makernice 补充了效果说明、分镜拆解、推荐参数与改写建议，并删去了原帖中含链接或 @提及的整行、纯话题标签行与推文串回复；其余提示词文字未作改动。