OpenZhenZhen AI 提示词画廊 · SEEDANCE2.5 · video
Seedance 2.5 at 1080p, which is now at 50% OFF on Pollo AI!
公开作者的 SEEDANCE2.5 视频案例,包含可复用的完整提示词。

提示词 / Prompt
English prompt
Create a 30-second, 1080p ultra-realistic documentary home-video of an ordinary late morning in a quiet Korean residential neighborhood. The entire video should feel like genuine footage casually recorded by a friend on an early-2000s consumer DV camcorder. MAIN SUBJECT The same clearly adult Korean woman in her early 20s, 18+, throughout the entire video. Natural everyday appearance, realistic skin texture, minimal makeup, warm and approachable personality. She wears a faded charcoal-grey sleeveless crop top, loose high-waisted light-wash jeans, black canvas sneakers, and a simple black cord necklace. Her black wavy hair is tied in a messy side ponytail with wispy bangs and loose strands. Keep her face, identity, body proportions, hairstyle, clothing, necklace, and overall appearance consistent from beginning to end. No outfit or hairstyle changes. LOCATION A quiet, authentic Korean residential neighborhood during late morning: narrow concrete alleys, low-rise homes, small terraces, potted plants, laundry lines, utility poles, overhead wires, mature trees, concrete walls and steps, and a few parked bicycles. No stores, cafés, advertisements, crowds, tourists, or commercial activity. CAMERA / VISUAL STYLE Early-2000s consumer DV camcorder footage recorded by a friend. Heavy but believable handheld shake, imperfect framing, occasional autofocus hunting, lens breathing, exposure changes between sun and shade, slight motion blur, subtle rolling shutter, mild digital compression, faded colors, soft contrast and slight sensor noise. No stabilization, gimbal movement, drone shots, cinematic camera moves, modern color grading, slow motion, dramatic lighting, or polished commercial cinematography. The camera operator sometimes reacts slightly late and imperfectly frames the subject. These imperfections are intentional. The footage should feel accidentally captured, not staged. 00:00–00:05 — HOUSE ENTRANCE Outside a small residential house, she sits casually on a low concrete wall and fixes her messy side ponytail with both hands. A light breeze moves loose strands of hair. She briefly notices the camera, gives a small natural smile, then looks away. The camera struggles to focus for a moment before finding her face. Keep the movement simple and natural. 00:05–00:10 — STRAY CAT She walks slowly into the narrow alley. A stray cat approaches from the side. She notices it, crouches down and gently pets it with one hand. She offers a tiny piece of food with her fingers and the cat naturally takes it. Her hand stays outside the cat’s mouth. The cat behaves like a real cat. The camera briefly shifts focus between her and the animal. 00:10–00:14 — LAUNDRY Beside the house, she hangs a couple of lightweight clothes on a simple clothesline. One piece slips slightly from her hand. She catches it and gives a small spontaneous laugh before securing it properly. The clothes move gently in the breeze. The clothesline stays fixed in place. 00:14–00:19 — TERRACE / COFFEE She sits on a small terrace with a ceramic coffee cup beside her. She picks up the cup with one hand and takes one small natural sip. The cup remains a separate physical object and stays outside her mouth and face. She lowers the cup and holds it naturally near her lap. With her free hand, she brushes a loose strand of hair behind her ear. She looks toward the neighborhood rather than posing for the camera. 00:19–00:23 — OFF-CAMERA MOMENT Someone casually calls to her from off-camera. She turns toward the voice and responds naturally. She stands up while holding the coffee cup securely. The camera reacts slightly late and briefly frames her imperfectly. No sudden object changes or teleportation. 00:23–00:26 — “ANNYEONG” The person behind the camera says something casually. She turns toward the camera, gives a small friendly wave and naturally says: “Annyeong.” She smiles briefly, then looks away. Her mouth movement should match the single word naturally. No repeated dialogue and no prolonged posing. 00:26–00:30 — WALKING AWAY She slowly walks down the quiet tree-lined residential lane while carrying the ceramic cup. The camera follows from several steps behind and slightly to the side. Her walking is natural, with realistic foot contact against the concrete. She notices the camera once more, gives a tiny genuine smile, then looks away and continues walking. At approximately 00:29, the recording abruptly cuts to black mid-motion, as if the old camcorder was switched off. No fade-out. PHYSICAL REALISM Maintain believable real-world physics throughout. Hands must correctly hold and release objects. No extra fingers, fused hands, duplicated limbs, distorted anatomy, floating objects, teleporting objects, disappearing objects, or sudden object transformations. The ceramic cup must remain separate from her face and mouth when drinking. The cat, food, clothing and other objects must interact naturally. Parked bicycles remain completely stationary unless a person physically touches or moves one. They must never move, ride, rotate or change position by themselves. Her feet remain connected to the ground while walking; no sliding or unnatural movement. Keep all background objects stable and consistent. AUDIO Natural location audio only: birds, light wind, leaves, distant motorcycles, faint neighborhood
主题与风格
探索更多图片和视频提示词 · 生成图片 · 生成视频 · 在 AI 画布中创作
相关作品
- They Picked The Wrong Girl
Create ONE ultra-realistic cinematic live-action action sequence at night on the famous San Antonio River Walk, Texas. Keep the recognizable River Walk visible throughout: illuminated bridges, stone architecture, trees, river reflections, and warm lights. Exactly ONE young woman, THREE male attacke
- Three Layer Travel Vlog Montage
Create a realistic 15-second vertical travel vlog using the reference image. Keep all three horizontal layers visible and continuously moving for the entire video. Never freeze, remove, or stop any layer. Change the location and shot in ALL THREE layers together every 4 seconds: 0–4 sec: Scene 1
- Underground Worm Chase Escape
生成一支30秒的写实地下怪兽追逐短片。 采用好莱坞商业生物惊悚片的动作调度:开场立即遇险,中段短暂躲过,随后危险突然升级,最后一次奋力逃脱带来短暂释放,再以撞击收尾。全片是同一场连续事件,多镜头剪辑,时间不跳跃。 角色与怪兽 成年女主角严格参考 #2  图片1 ,发型与服装严格参考 #3  图片_20260911221703_12_537图片_2026091122… ,保持面容设计。右手握带腕绳的手电。她不是始终对着镜头展示惊恐,而是根据眼前发生的事做出不同反应:逃跑时咬牙控制呼吸;听见身后扑击时骤然缩肩、抽气;看见逃生门时眼神集中、身体加速;摔倒后先疼得闭眼,再挣扎爬起;门关上后才出现压
- New York Vertical Travel Film
Create a ultra-realistic cinematic vertical travel film in strict 9 16, showcasing the iconic beauty of New York City, USA America, using the generated image as the visual reference for composition, skyline, lighting, and landmarks. — EAST RIVER OPENING Begin with a cinematic view over the East Riv
- Miniature Luxury Treehouse Timelapse Build
在树上建造一座小型豪华住宅 🏡生成提示词 MOTION PROMPT 1 Step1:在 Google flow 生成一张初始图片,选择 Nano Banana 2 生图模式,复制下面的提示词,生成图片—— A hyper-realistic miniature luxury treehouse construction site at the very beginning of construction, built high among the thick branches of a large mature tree, set in a natural forest environm
- Bedroom Goth Makeup Transformation Video
Duration:16秒 Aspect ratio:9:16竖屏 Overall style: 真人美妆变身短视频,固定手机机位,近距离人像构图。 前半段是日常卧室里的自然状态,光线明亮中性;后半段变为暗黑华丽的哥特妆造,脸部有柔和正面光,黑色羽毛与珍珠反光清晰。 保留真实皮肤纹理、细小亮片、薄纱和羽毛质感,不做动漫脸或塑料磨皮。 SUBJECT / 角色绑定: @图片1 命名为【女主】,全片唯一人物。保持角色卡的面部身份与骨相;前后造型可以变化,不能变成另一张脸。 ENVIRONMENT / 场景: 前半段是普通卧室:浅灰色墙面,后方可见床靠和床品,画面右上角有一幅抽象挂画。固定手机正