subject_definitions:
<Subject 1> is Female Lead 2, the adult Chinese woman shown in <Picture 1>, with exact facial identity reinforced by <Picture 2> and full charcoal blue-grey wardrobe context from <Picture 3>. Preserve her longer angular oval face, defined cheekbones, narrow intense dark eyes, straight brows, small beauty mark beneath her left eye, black hair pulled back beneath a dark antique-gold headpiece, long blue-red bead strands, and layered charcoal blue-grey embroidered wraps. Her fixed voice is an adult Chinese female low mezzo with firm chest resonance, a cool dense tone, a faint restrained metallic edge without sharpness, shorter controlled breath units, clean consonant onsets, and downward-resolving phrase endings; composed, hard-edged, never high-pitched, sweet, girlish, breathy, shrill, or shouted. She is (S1) in this clip.
<Picture 1> is the storyboard composition for [Shot 1], defining <Subject 1>'s close facial scale, screen-left eyeline, small left-eye beauty mark, dark embroidered costume, snowy woodland background, and shallow depth of field.

summary:
[reference generation] Generate one continuous 6.0-second reverse close-up of <Subject 1>. <Picture 1> controls the composition while <Picture 2> locks her face and <Picture 3> reinforces her full dark costume and the shared snowy setting only. She delivers the complete second dialogue block in three measured semantic units, with two silent pauses and a closed-mouth reaction tail. Female Lead 1 never enters the frame. Only <Subject 1>'s clean dialogue is audible.

retention_analysis:
<Subject 1> (appears throughout [Shot 1]): fully_preserved - retain her sharper facial structure, left-eye beauty mark, narrow eyes, headpiece, blue-red beads, charcoal blue-grey costume, adult proportions, low cool voice, and screen-left attention.
<Picture 1> ([Shot 1] composition): fully_preserved - retain the close framing, face scale, leftward eyeline, winter light, snowy soft-focus background, and controlled expression while allowing precise speech animation.

detailed_description:
Preserve the exact live-action texture of <Picture 1>—cool winter shadows, natural skin detail, fine hair, and believable dark blue-grey embroidery—without redesigning the image or borrowing any visual treatment from another project.

[Shot 1] Hold a stable eye-level reverse close-up matching <Picture 1>. <Subject 1> fills the right-middle of the landscape frame and looks toward screen-left, maintaining the same 180-degree axis established by the previous Female Lead 1 close-up. Female Lead 1 remains off-screen and never appears as a shoulder, reflection, duplicate face, or background figure. The camera is static during the first phrase, then begins a barely perceptible small push-in after the first pause. Keep the face, beauty mark, costume color, bead arrangement, lighting and background stable. Snowflakes and fine hair may move visually without producing sound. No hand or weapon action enters the crop.

Use <Picture 3> only to reinforce the same dark costume and shared snowfield lighting. Do not copy its second woman, wider framing, body position, or two-person composition into this solo close-up.

From 0.00 to 2.06 seconds, <Subject 1> (S1) speaks in her fixed low mezzo voice, with firm chest resonance, clear consonant onsets, and no sharp high frequencies: <d>[Chinese] 因为刀的真意。</d> The delivery is calm and factual rather than defensive. She closes her lips fully at 2.06 seconds.

From 2.06 to 2.58 seconds, she holds a silent 0.52-second pause, eyes fixed screen-left. From 2.58 to 3.82 seconds, <Subject 1> (S1) lowers her chin by only a few millimetres and says, without increasing volume: <d>[Chinese] 不在杀。</d> Make “杀” short, hard, and downward-resolving. Her lips close again.

From 3.82 to 4.60 seconds, hold a silent 0.78-second pause. Her eyes become slightly heavier while her brows and head remain controlled. From 4.60 to 5.80 seconds, <Subject 1> (S1) delivers the final unit in the same voice: <d>[Chinese] 在藏。</d> Place the full weight on “藏,” lower the ending, and complete the syllable without whispering, shouting, or stretching it beyond the window.

From 5.80 to 6.00 seconds, her lips remain closed for a 0.20-second silent reaction tail while the screen-left eyeline stays fixed. Do not add a smile, aggressive glare, repeated line, extra word, weapon movement, subtitle, caption, or cut.

overall_soundscape:
Only the clean, dry Mandarin dialogue of <Subject 1> (S1) is audible. Every pause is absolute silence. No wind, snow, breathing, fabric, beads, footsteps, weapon sound, room tone, reverberant ambience, or other sound effect is present.

non_diegetic_music:
N/A
