subject_definitions: <Subject 1> : 2D anime-style Asuka… | Deepfake AI视频

作者头像
@xmsthc

生成此视频所用的文字转视频提示词

subject_definitions: <Subject 1> : 2D anime-style Asuka from <Picture 1>[cite: 1], featuring ginger-orange twin pigtails, red A10 nerve clips, blue anime eyes, and red plugsuit 02[cite: 1]. <Subject 2> : Photorealistic live-action cosplay Asuka from <Picture 2>[cite: 2], featuring realistic human facial features, detailed blue eyes, natural auburn twin tails, and glossy red latex plugsuit 02[cite: 2]. <Picture 1> : 2D anime character sheet of Asuka Langley Soryu (asuka-anime.jpg)[cite: 1]. <Picture 2> : Photorealistic cosplay character sheet of Asuka Langley Soryu (asuka-real.jpg)[cite: 2]. <Audio 1> : Japanese female voice line saying "あんたバカァ?", bound to <Subject 1> (S1)[cite: 4]. summary: [reference generation + audio reuse] In a seamless shot against an absolute black background, 2D anime Asuka (<Subject 1>)[cite: 1] enters the frame, glares at the camera, shouts her iconic line synchronized to <Audio 1>[cite: 4], and punches through the camera glass screen with her mouth still open mid-shout, seamlessly transforming into live-action Asuka (<Subject 2>)[cite: 2] who maintains the exact identical punching posture, open shouting mouth, and fierce tsundere glare against the pure black void as the glass shatters. retention_analysis: <Picture 1>: partially_preserved[cite: 1] <Picture 2>: partially_preserved[cite: 2] <Audio 1>: fully_copy[cite: 4] <Subject 1>: partially_preserved <Subject 2>: partially_preserved detailed_description: A medium close-up shot maintaining an absolute pitch-black void background throughout, executing an in-place visual transition from 2D animation to photorealistic live-action through slow-motion glass destruction while strictly maintaining subject posture and facial expression, especially her open mouth shape across the impact. [Shot 1] The frame opens in absolute, pitch-black darkness. From the right edge against the pure black background, <Subject 1> walks into the frame in classic 2D cel-shaded animation style from <Picture 1>[cite: 1]. She stops in a medium close-up, looking left and right with guarded curiosity[cite: 1]. Her gaze abruptly locks directly into the camera lens as she spots the viewer. Her eyebrows angle sharply downward in an angry, indignant tsundere scowl[cite: 1]. Leaning forward aggressively, <Subject 1> (S1) opens her mouth wide and shouts, synchronized to the dialogue in <Audio 1>: <d>[Japanese] あんたバカァ?</d>[cite: 4]. Right as the vocalization finishes on the elongated final syllable, with her mouth still held wide open from shouting, she throws a forceful straight right punch forward into the invisible camera glass plane. The screen immediately shatters into sharp, reflective glass fragments bursting outward toward the camera. In the exact moment of impact and glass fracture, against the continuous pure black void, <Subject 1> seamlessly transitions into <Subject 2> from <Picture 2>[cite: 2], maintaining the exact identical punching posture and facial expression: her right fist remains extended straight toward the lens, her torso leans forward at the same angle, her eyebrows stay sharply furrowed over piercing blue eyes, and her mouth remains distinctly held open in the same mid-shout vocal shape, rendered with tangible photorealistic human detail and glossy latex texture under focused front lighting[cite: 2]. overall_soundscape: Opens in total silence against the void. Light anime footsteps enter, followed by an intake of breath and the synchronized playback of <Audio 1> delivering the crisp, sharp Japanese voice line: <d>[Japanese] あんたバカァ?</d>[cite: 4]. An explosive punch impact echoes with heavy sub-bass, immediately followed by sharp crystalline cracking and glass shards clattering into darkness. non_diegetic_music: none