【故事与角色】  参考图锁定身份（五官/发型/服装/背景/构图）+ 情绪动机一句话
【机位与构图】  6秒，16:9，单场景单镜头，中近景，固定机位，延续原图光线与浅景深
【镜头描述】    镜头不推拉摇移，焦点锁定双眼和嘴部，口型清晰
【声线与台词】  声线质感 + 台词逐句的语气/停顿/音量/尾音处理
【正文】        0—1秒 / 1—3秒 / 3—5秒 / 5—6秒 逐段写表演动作，台词插在时间点里
【声音设计】    环境底噪 + 关键动作声，明确"全程不用背景音乐"
【禁止项】      负面清单：表情过头、身份漂移、手指畸形、口型错位、运镜切镜、字幕水印等

For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.

integrated_multimodal_description: [Shot 1] Preserve the exact identity and composition from <Picture 1>: ...
（先锁身份，再按 At 0.00–1.00 / 1.00–3.00 / 3.00–5.00 / 5.00–6.00 时间码写表演，
台词用 <d>中文台词</d> 包裹，结尾统一声明单镜头 + 保持身份 + 禁止项）

overall_soundscape: （环境声 + 动作声 + 同步台词）
non_diegetic_music: N/A

For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.

integrated_multimodal_description: [Shot 1] Preserve the exact identity and centered medium-close 16:9 composition from <Picture 1>: one young East Asian woman facing the camera in a quiet traditional courtyard, with the same oval face, dark classical updo and pale hair ornaments, translucent ivory outer hanfu, pale celadon embroidered dress, peach ribbons, soft natural daylight, blurred grey architecture, shallow depth of field, and stable background. The camera is completely locked off; keep focus on her eyes and mouth. At 0.00–1.00, she gives the camera a light pretend glare, knitting her brows while suppressing a smile and keeping relaxed shoulders. At 1.00–3.00, she leans forward slightly and says in a young playful, faintly nasal voice: <d>我恨你</d>; she makes a tiny chin dip and one corner of her mouth briefly lifts. At 3.00–5.00, she turns her head slightly aside, makes one very small shooing gesture, steals a glance back at the camera, and softly says <d>你走吧</d> with a gently rising ending. At 5.00–6.00, her raised hand pauses then slowly returns as an irrepressible small smile appears. Use exactly one continuous shot. Preserve identity, face, hairstyle, ornaments, costume, hands, background, lighting, framing, and accurate Chinese lip synchronization. No camera motion, cuts, subtitles, readable text, watermarks, logos, UI, extra people, background changes, black frames, or transitions.

overall_soundscape: Quiet courtyard room tone, a tiny nasal huff, soft fabric rustle, natural breathing, and precise synchronized dialogue.

non_diegetic_music: N/A

For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.

integrated_multimodal_description: [Shot 1] Preserve the exact identity and centered medium-close 16:9 composition from <Picture 1>: one young East Asian woman facing the camera in a quiet traditional courtyard, with the same oval face, dark classical updo and pale hair ornaments, translucent ivory outer hanfu, pale celadon embroidered dress, peach ribbons, soft natural daylight, blurred grey architecture, shallow depth of field, and stable background. The camera is completely locked off; keep focus on her eyes and mouth. At 0.00–1.00, she stares into the camera without moving, brows drawn inward, lips progressively pressed tight, neck and shoulders rigid. At 1.00–3.00, she releases her lips and says <d>我恨你</d> slowly in a low tense controlled voice; her jaw hardens and the fingers at her side gradually clench. At 3.00–5.00, she swallows once without breaking eye contact, then says <d>你走吧</d> at lower volume with a short controlled exhale. At 5.00–6.00, she remains still, hand clenched and gaze coldly fixed on the viewer. Use exactly one continuous shot. Preserve identity, face, hairstyle, ornaments, costume, hands, background, lighting, framing, and accurate Chinese lip synchronization. No camera motion, cuts, subtitles, readable text, watermarks, logos, UI, extra people, background changes, black frames, or transitions.

overall_soundscape: Faint courtyard breeze, one clear swallow, brief nasal breathing, subtle fabric movement, and synchronized Chinese dialogue.

non_diegetic_music: N/A

For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.

integrated_multimodal_description: [Shot 1] Preserve the exact identity and centered medium-close 16:9 composition from <Picture 1>: one young East Asian woman facing the camera in a quiet traditional courtyard, with the same oval face, dark classical updo and pale hair ornaments, translucent ivory outer hanfu, pale celadon embroidered dress, peach ribbons, soft natural daylight, blurred grey architecture, shallow depth of field, and stable background. The camera is completely locked off; keep focus on her eyes and mouth. At 0.00–1.00, she grits her teeth at the camera, brows low and shoulders rising as she takes one restrained breath. At 1.00–3.00, her head and upper body suddenly thrust a little forward and she sharply says <d>我恨你</d>; one hand clenches hard at her side and the sleeve sways once. At 3.00–5.00, after one quick inhalation, she gives a single firm pointing gesture off to one side and forcefully says <d>你走吧</d>. At 5.00–6.00, the hand slowly lowers while she remains slightly forward, breathing heavily and maintaining an angry gaze. Her face stays recognizable and natural, never grotesque. Use exactly one continuous shot. Preserve identity, face, hairstyle, ornaments, costume, hands, background, lighting, framing, and accurate Chinese lip synchronization. No camera motion, cuts, subtitles, readable text, watermarks, logos, UI, extra people, background changes, black frames, or transitions.

overall_soundscape: Courtyard wind, a quick inhalation, one sharp sleeve rustle, heavy controlled breathing, and clear synchronized Chinese dialogue.

non_diegetic_music: N/A

For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.

integrated_multimodal_description: [Shot 1] Preserve the exact identity and centered medium-close 16:9 composition from <Picture 1>: one young East Asian woman facing the camera in a quiet traditional courtyard, with the same oval face, dark classical updo and pale hair ornaments, translucent ivory outer hanfu, pale celadon embroidered dress, peach ribbons, soft natural daylight, blurred grey architecture, shallow depth of field, and stable background. The camera is completely locked off; keep focus on her eyes and mouth. At 0.00–1.00, her gaze drops diagonally downward, lips tuck inward, shoulders slowly close in, and fingers grip the edge of her sleeve. At 1.00–3.00, she quickly looks back up and says <d>我恨你</d> in a low strained voice as her eyes redden slightly and the fabric creases under her tightening fingers. At 3.00–5.00, she lowers her gaze again, swallows, and quietly says <d>你走吧</d>; her chin gives a tiny tremor and breath catches in her throat. At 5.00–6.00, she remains withdrawn, still gripping her sleeve, with tears gathered at the lower eyelid but not falling. Use exactly one continuous shot. Preserve identity, face, hairstyle, ornaments, costume, hands, background, lighting, framing, and accurate Chinese lip synchronization. No camera motion, cuts, subtitles, readable text, watermarks, logos, UI, extra people, background changes, black frames, or transitions.

overall_soundscape: Very quiet courtyard ambience, a sleeve-grip rustle, one swallow, short nasal breath, and restrained synchronized Chinese dialogue.

non_diegetic_music: N/A

For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.

integrated_multimodal_description: [Shot 1] Preserve the exact identity and centered medium-close 16:9 composition from <Picture 1>: one young East Asian woman facing the camera in a quiet traditional courtyard, with the same oval face, dark classical updo and pale hair ornaments, translucent ivory outer hanfu, pale celadon embroidered dress, peach ribbons, soft natural daylight, blurred grey architecture, shallow depth of field, and stable background. The camera is completely locked off; keep focus on her eyes and mouth. At 0.00–1.00, she looks up from below toward the camera, moisture gathering in her eyes, chin slightly tucked and shoulders close to her body. At 1.00–3.00, with a small inhale and a soft nasal trembling voice, she says <d>我恨你</d> while her lower lip quivers and fingers twist the sleeve in front of her. At 3.00–5.00, her gaze briefly meets the camera then slides away as she softly says <d>你走吧</d>, turning her body a little aside without truly leaving. At 5.00–6.00, she begins to look back but stops halfway; lips press together and tears remain pooled but do not fall. Use exactly one continuous shot. Preserve identity, face, hairstyle, ornaments, costume, hands, background, lighting, framing, and accurate Chinese lip synchronization. No camera motion, cuts, subtitles, readable text, watermarks, logos, UI, extra people, background changes, black frames, or transitions.

overall_soundscape: Soft courtyard breeze, one brief inhale, delicate sleeve friction, quiet breathing, and lip-synchronized Chinese dialogue.

non_diegetic_music: N/A

For the target video, at 0.00 seconds into the target video, <Picture 1> (from [Shot 1]) is fully referenced.

integrated_multimodal_description: [Shot 1] Preserve the exact identity and centered medium-close 16:9 composition from <Picture 1>: one young East Asian woman facing the camera in a quiet traditional courtyard, with the same oval face, dark classical updo and pale hair ornaments, translucent ivory outer hanfu, pale celadon embroidered dress, peach ribbons, soft natural daylight, blurred grey architecture, shallow depth of field, and stable background. The camera is completely locked off; keep focus on her eyes and mouth. At 0.00–1.00, she looks quietly into the camera, her focus gradually softening as moisture forms under her eyes; the corners of her mouth fall and shoulders draw slightly inward. At 1.00–3.00, she refocuses briefly and says <d>我恨你</d> in a low controlled voice, lips delicately trembling while the tears remain at the eyelid. At 3.00–5.00, she blinks once; exactly one natural tear slowly travels down one cheek as she quietly says <d>你走吧</d> and her gaze lowers diagonally. At 5.00–6.00, she does not wipe the tear, closes her lips, and remains bowed with the tear track visible. Use exactly one continuous shot. Preserve identity, face, hairstyle, ornaments, costume, hands, background, lighting, framing, and accurate Chinese lip synchronization. No camera motion, cuts, subtitles, readable text, watermarks, logos, UI, extra people, background changes, black frames, or transitions.

overall_soundscape: Still courtyard ambience, a controlled inhale, very light nasal breathing, and soft lip-synchronized Chinese dialogue; no sobbing.

non_diegetic_music: N/A