prompt.txt

subject_definitions:
<Subject 1> is the young woman in <Picture 1>, with fair skin, long straight black hair with light bangs, and a high bun decorated with white flower hairpins and long beaded tassels falling beside her ears, wearing a pale mint-green hanfu with sheer layered sleeves and a white inner skirt.
<Picture 1> is the front-view identity anchor for <Subject 1>; use it for the face, hair, and costume.
<Picture 2> is a character turnaround sheet showing the SAME single person from three angles (front, side, back) — the three figures are ONE person, not three different people; use it only for hair ornament and sleeve details.
<Audio 1> is the reference song used for lip-sync timing.

summary:
[reference generation + audio reuse] A 12-second vertical extreme close-up of <Subject 1> singing along with the reference song, her expression a gentle sorrowful melancholy, head and upper body swaying softly with the music, camera locked on her face for the whole shot.

retention_analysis:
<Subject 1> (appears in [Shot 1]): fully_preserved - face, hair, and mint hanfu stay identical to <Picture 1>.
<Picture 1>: fully_preserved - identity only.
<Picture 2>: partially_preserved - hairpin and sleeve details.
<Audio 1>: fully_copy - the song and voice follow the reference audio exactly, lips in sync.

detailed_description:
[Shot 1] Photorealistic. Vertical 9:16. Extreme close-up, the face fills the frame from the very first frame. One continuous take, no cut, a locked static camera.
00:00-00:12.000: The camera holds a static extreme close-up on <Subject 1>'s face and upper body, her face filling most of the frame. She sings along with the music in exact sync, moving her lips to the song. Her expression is a beautiful, restrained melancholy: soft longing eyes, slightly parted lips, a wistful gentle sadness; her brows stay relaxed, her gaze occasionally drifts down and back up. Her head tilts slowly and her shoulders and upper body sway subtly with the music rhythm, like a flower swaying in a gentle breeze. Her hairpin tassels and sheer sleeves drift slightly with the movement. The background is a soft, clean, softly-lit studio backdrop with shallow depth of field. Only <Subject 1> appears, no text, no logos, no other people.

overall_soundscape:
Soft fabric rustle as the sleeves sway; gentle breathing between sung phrases; quiet warm room ambience.

non_diegetic_music:
N/A (the song is the diegetic reference audio she sings along to).
下载此文件