# E23-03 · H3 提示词

首帧 = **f1.png**。图片按 Picture 序号挂载：

- Picture 1 = f1.png（**首帧**，钉 0.00 秒）
- Picture 2 = f2.png（钉 5.80 秒）
- Picture 3 = f3.png（钉 10.50 秒）

---

How the reference pictures align with the target video — Picture 1 (from Shot 1) aligns with the 0.00-second mark of the target video; Picture 2 (from Shot 2) aligns with the 5.80-second mark of the target video; Picture 3 (from Shot 3) aligns with the 10.50-second mark of the target video.
integrated_multimodal_description:
[Shot 1] Wide shot at night: two figures half-hidden by the cold-blue formation glow, a slim scarred youth pressing toward the stone steps while the red-robed woman pulls his hand, thin fog between pillars, cinematic film still (camera: static shot) <d>[Chinese] 叛徒。我拿令闯山门，进去报信。</d>
[Shot 2] At 00:05.800, the camera cuts to <Picture 2>: close-up of Close shot slowly pushing in: her clear-cold eyes steady, the youth's wounded hand trembling, moonlight carving the deep shadow of the mountain gate behind them, his breath short, cinematic film still (camera: push in) <d>[Chinese] 你进去，血衣客正好一网打尽。</d> <d>[Chinese] 师父还在里面。</d>
[Shot 3] At 00:10.500, the camera cuts to <Picture 3>: extreme close-up of Extreme-close arc shot: her gaze falls on the worn bronze token in his fist, pale light washing over his coarse bleeding palm, the formation hum trembling through stone, cinematic film still
overall_soundscape: Cold wind through pillars, low rumbling hum of the protective formation, rustle of sleeves and a quiet gasp.
non_diegetic_music: A sparse guqin line with low dizi drones, growing tense but unresolved.