subject_definitions: <Subject 1> is raymono, an East Asian… | Deepfake AI视频

生成此视频所用的文字转视频提示词

subject_definitions: <Subject 1> is raymono, an East Asian woman in her mid-20s with a sleek high ponytail, cold and guarded: a charcoal-grey tailored blazer over a black turtleneck, slim trousers, a minimal silver wristwatch, tiny silver hoop earrings, and a black lanyard; her voice is a cool, flat, low-key young female voice with clipped, terse delivery. summary: [reference generation] The target video is the second unit of a four-part first-person POV vertical web-drama character introduction for a rom-com. At his new desk in an open-plan start-up office, the shy protagonist-cameraman is caught staring by <Subject 1>, who never lifts her eyes from her monitor. After his confused reply, she details his exact actions — staring for two minutes forty-three seconds, lifting his cup four times, never taking a sip — as her intro card slides in, repeats her question, and after his apology scolds him that work time is for work and she must not catch him staring again; he only ever replies and never initiates. Three POV shots in the exact same office environment are used. retention_analysis: <Subject 1> (appears in [Shot 1], [Shot 2], [Shot 3]): fully_preserved - the sleek high ponytail, charcoal-grey blazer, black turtleneck, silver wristwatch, and tiny hoop earrings are retained. detailed_description: The target video is a first-person POV vertical (9:16) short web-drama style video with bright sitcom lighting, playful handheld energy, and rom-com title-card graphics. [Shot 1] A POV vertical medium shot at eye level looks over the protagonist's own desk in an open-plan start-up office: long shared desks, monitors framed with sticky notes, and potted plants. Directly across sits <Subject 1> (S1), raymono with a sleek high ponytail, charcoal-grey blazer, and black turtleneck. She is typing rapidly, her eyes completely glued to her monitor and never lifting to meet the lens. She says in a cool, flat, clipped voice without looking up: <d>[Cantonese] 打算望到幾時？</d> The shy cameraman, heard from behind the camera in a hesitant, muffled, soft young male voice (S2), gives a confused reply: <d>[Cantonese] (emm) 啊？我、我無呀……</d> [Shot 2] At 00:05.000, the camera cuts to a POV close-up of <Subject 1> (S1) across the same shared desk. Her eyes remain locked on her screen, her face illuminated by the monitor's glow with the sticky notes and potted plants softly blurred in the background. As she begins detailing his actions, a cool grey intro card slides in from the left and lands reading "梁玥 — Visual Design · the ice wall across your desk". She counts off in the same flat tone: <d>[Cantonese] 望住我兩分四十三秒，拎杯四次，一口都無飲。打算望到幾時？</d> The shy cameraman (S2) finally realizes he has been caught and stammers an apology: <d>[Cantonese] 對、對唔住……</d> [Shot 3] At 00:10.000, the shot cuts back to a POV medium shot of the exact same open-plan desk environment, with the long shared desks, sticky-note-framed monitors, and potted plants fully visible. <Subject 1> (S1) continues her relentless typing, her gaze fixed entirely on her work, offering him absolutely no warmth or welcome. She scolds him in a clipped, authoritative tone: <d>[Cantonese] 返工時間做返工嘅事。唔好再望。</d> She continues to face her work as the shot ends on a single sharp keystroke. overall_soundscape: Keyboard clatter, rolling chair casters, and a distant phone ring fill the open office. The typing is relentless and rhythmic, pausing only for a fraction of a second before resuming, and the video ends on a single sharp keystroke. non_diegetic_music: A single deadpan plucked note held under the whole video, joined by one dry pizzicato tick at each of her counted accusations.