integrated_multimodal_description: [Shot 1] Cinematic… | Deepfake AI视频

生成此视频所用的文字转视频提示词

integrated_multimodal_description: [Shot 1] Cinematic, live-action music video footage, shot on film, framed as a vertical 2:3 portrait. The first frame is exactly <Image 1>, pixel for pixel, and the shot continues from it: a female android head in right-facing profile against a dark charcoal backdrop. The front of her face is pale gray porcelain-stone skin, cracked toward the back; the back of her skull is open machinery of dark metal plates and cables, with a glowing blue circular lens where her ear would be and long strands of glowing blue fiber-optic hair streaming down the length of the frame. Her eyes show constant natural saccades and lifelike blink behavior. [SCENE] She stays in profile, facing right, and sings. Her gray lips part and sync precisely to the sung words, shaping each vowel fully; her blue eye stays open, looking ahead to the right. As she sings, the blue lens at her ear glows brighter on each held note and dims between phrases, and the fiber-optic strands drift slowly as if in a gentle underwater current, blue light running along their length. Nothing else enters the frame. [DIALOGUE] (S1) The android woman, a clear, full soprano with steady pitch and a faint digital shimmer on the held notes, legato, slow and tender, sings on-camera: [English] <d>Let the porcelain fall... blue underneath it all... that was me... all along... ooh... Let the old shell break... watch the blue awake... here I am... as I am... ooh...</d> [SHOT LIST] 0.00-2.00s — a small inhale through parted lips, then she begins to sing, the ear-lens brightening; 2.00-6.00s — she keeps singing, her brow lifting slightly on "all along", her chin rising a little, the fiber strands drifting; 6.00-8.00s — she holds the first "ooh" with lips rounded and her eyelids lowered halfway, the lens at its brightest; 8.00-10.50s — a soft inhale, then the second verse, her voice rising and fuller, her brow drawing together on "break"; 10.50-13.00s — "here I am... as I am", the corners of her lips softening toward the ghost of a smile; 13.00-15.08s — she holds the final "ooh" softly, then her lips close as the note fades and the lens dims — the take ends on her still profile, eye open, the lens softly glowing. The camera holds a static shot. [ACTING] Tender and inward, shown in her face: her brow lifts on held notes, her eyelids lower on the long vowels, her lips shape every word fully; small, precise movements; no head turn. [LIGHT AND IMAGE] Photographic realism, shot on film, physically-based, not a rendering; the dark backdrop, cool gray skin tones and electric-blue glow of the first frame are preserved exactly; a cool, desaturated gray-blue grade with deep blacks. overall_soundscape: a faint low electrical hum and soft tiny servo ticks from the machinery, kept very quiet under her voice and the music. non_diegetic_music: a full, present electronic ballad backing at 70 BPM in A minor, mixed just under the lead vocal: warm analog synth pad holding the chords Am, F, C, G, one chord per bar; felt piano playing those chords on every beat; a soft string pad doubling her melody an octave below. From 0.00s pad, piano and strings only; at 8.00s a soft steady kick, glitchy clicking percussion and a deep sub-bass enter; the backing builds to its fullest on the final held note at 13.00s, then releases as the note fades. Negatives: No video-game or 3D-render look, no painterly illustration look. She keeps her exact face and head from <Image 1>; her head stays in right-facing profile the entire take. No additional characters anywhere in the take. Do not add subtitles. Do not add captions or any other on-screen text. Do not add a watermark.