提示词 / 授权来源

Prompt 028 · @qaistudio

@qaistudio · 文本生成视频

Seedance 2.5Seedance 2.5@qaistudiovideo完整提示词text-to-video
Person: Use Image1 as the exact face/hairstyle reference. Preserve identity and facial features completely; ignore its outfit, background, pose, and text. Casual everyday streetwear.

Format/Style: 16:9 horizontal video with an old iPhone 1x look: low quality, heavy noise/compression, natural skin texture, slight handheld shake, flat standard FOV. No cinematic grading, beauty filter, CG texture, fisheye, vignette, or distortion.

Setting: Small colorful purikura booth in a convenience store/arcade, soft ring-light glow, partial curtain at entrance. Screens/templates/text remain illegible.

Timeline:

- 0–4s: Enters alone, sits/stands at touch screen, fumbles through menus, mutters: “wait which one do I press—”
- 4–8s: Randomly selects a frame, lighting/colors shift, adjusts stance and checks reflection.
- 8–11s: Countdown blinks; attempts an overly cute/awkward pose and feels silly.
- 11–14s: Flash fires; breaks pose, laughs at reflection, covers mouth: “oh my god that was so awkward.”
- 14–17s: Exits to sticker printer, waits for sheet, picks it up and looks at it. Sticker uses Image1 face reference; graphics/text illegible.
- 17–19s: Peels off one sticker, places it on the back of phone case, presses it down, then looks at camera with a satisfied grin. Freeze.

Camera/Sound: Loose 1–1.5s cuts, slight natural shake entering/exiting, mostly static in booth, pink/magenta off-white balance, brief flash overexposure. No BGM—only ambient store/arcade noise, taps, countdown beep, shutter, printer sounds, laughter, and close muffled booth-echoed voice. No skipped scenes or consolidated cuts.