prompt.txt

<Subject 1> is the young woman in <Picture 1>, with light bangs and a high ponytail, wearing a gradient sleeveless leotard that fades from pale beige at the top to soft light blue at the bottom, clean sporty gymnastics look, bare arms and legs, standing in a dance studio.
<Picture 1> is the front-view identity anchor for <Subject 1>; use it for the face, hair, and leotard.
<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 the leotard gradient and ponytail details.
<Picture 3> is the exact last frame of the previous clip — the opening frame of this clip MUST be identical to <Picture 3> in face angle, expression, mouth state, head tilt and lighting.
<Audio 1> is the reference song, exactly the part to lip-sync from 00:00.000 to 00:11.000.

summary:
[video continuation + audio reuse] Continue seamlessly from <Picture 3> (the last frame of the previous clip): the same vertical half-body shot of <Subject 1> in a modern dance studio: a large wall mirror on one side, practice barre, warm wooden floor, big windows with soft daylight, a few training props, clean bright atmosphere; her lips move in perfect syllable-by-syllable sync with the singing in <Audio 1>, every mouth shape matching the song for the entire clip from 00:00.000 to the end, with no drift and no gap. The first frame is exactly <Picture 3>, then the clip flows on with the same face, framing, her eyes are moist and luminous with a soft glistening sheen, the gaze glides slowly and naturally — eyes subtly roving with a wistful, longing melancholy, alive and expressive, never vacant or staring; the corners of her lips are soft and relaxed, gently downturned in a quiet sorrowful curve, no smile, no stiff set, and soft swaying motion, camera unchanged, no cut, no jump.

retention_analysis:
<Subject 1> (appears in [Shot 1]): fully_preserved - face, ponytail, and gradient leotard stay identical to <Picture 1>.
<Picture 1>: fully_preserved - identity only.
<Picture 2>: partially_preserved - leotard gradient and ponytail details.
<Picture 3> ([Shot 1] first frame): fully_preserved - the opening frame must reproduce this exact frame: same angle, same mouth, same expression, same light.
<Audio 1>: fully_copy - the reference audio drives the lip-sync for the entire 11-second clip; every syllable is matched from 00:00.000 to 00:11.000, no drift.

detailed_description:
[Shot 1] Photorealistic. Vertical 9:16. Half-body medium shot. This clip begins EXACTLY on <Picture 3> — the last frame of the previous clip — with the identical framing, face angle, mouth position, expression and lighting, then continues as one seamless shot with no cut, no jump, no reset.
00:00-00:11.000: The camera holds the same static half-body shot of <Subject 1> from the waist up in the a modern dance studio: a large wall mirror on one side, practice barre, warm wooden floor, big windows with soft daylight, a few training props, clean bright atmosphere. From the opening frame on, her lips move in perfect syllable-by-syllable sync with the singing in <Audio 1>, every mouth shape matching the song for the entire clip from 00:00.000 to the end, with no drift and no gap. She keeps singing along with the music, moving her lips to every word, continuing seamlessly from where the previous clip stopped. her eyes are moist and luminous with a soft glistening sheen, the gaze glides slowly and naturally — eyes subtly roving with a wistful, longing melancholy, alive and expressive, never vacant or staring; the corners of her lips are soft and relaxed, gently downturned in a quiet sorrowful curve, no smile, no stiff set. Her head tilts slowly and her shoulders and upper body sway subtly with the music rhythm, flowing continuously without any new start. Her ponytail sways gently with the movement. The same soft daylight shapes her face, warm and natural. Only <Subject 1> appears, no text, no logos, no other people.

overall_soundscape:
Soft fabric rustle; gentle breathing between sung phrases; a faint quiet hall ambience with light reverb.

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