MiniMax H3 / Hailuo 3.0 · T2VA
谢尔顿与佩妮的纯提示词生成
以谢尔顿和佩妮作为文字指令中的角色，用 MiniMax H3 直接完成角色场景。
- 生成模式
- T2VA
- 模型
- MiniMax H3 (creator-attributed)
- 输出
- 12s · 1376×768 · landscape
- 标签
- 纯提示词 · 双角色 · 人物生成
- 提示词来源
- 创作者原文
原始 Prompt
以下为来源公开的原文 Prompt，未翻译、改写或补全。
Create a 12-second live-action sitcom-style scene in 16:9 aspect ratio featuring Sheldon Cooper and Penny from The Big Bang Theory. The scene is a very silly, funny, character-accurate conversation about artificial intelligence that ends with both characters realizing in horror that they are not real people, but prompts inside an AI-generated video. The tone should feel exactly like a classic Big Bang Theory hallway scene: fast, witty, natural, awkward, and playful, with precise comic timing and highly exaggerated, animated facial expressions. The pacing must feel natural and unforced, like an actual exchange from the original sitcom series. IMPORTANT DIALOGUE RULE: Only the dialogue written in double quotes below may be spoken. There must be no extra dialogue, no improvisation, no narration, no voice-over, and no additional spoken words. Each line must be spoken exactly once, by the specific character assigned to it. Sheldon Cooper says: “AI language models predict the next most statistically probable word.” Penny says: “So what am I about to say?” Sheldon Cooper says: “Something intellectually disappointing.” Penny says: “Wait… somebody already wrote this sentence.” Sheldon Cooper says: “We’re scripted?” Penny says: “Worse. Prompted.” No other spoken words are allowed. LOCATION AND SCENE SETTING: Set the entire scene in the iconic apartment hallway and elevator lobby area from The Big Bang Theory, outside Penny’s apartment 4B. The elevator is visibly broken as usual, with the dark empty elevator shaft area and caution-taped or damaged elevator look recognizable in the softly blurred background. Penny’s apartment door is already open at the start of the scene. Penny stands immediately inside the doorway of apartment 4B, facing outward into the hallway. Sheldon Cooper stands immediately outside the doorway, positioned very close to Penny so Sheldon Cooper and Penny can both fit naturally inside an extremely tight two-person close-up without requiring a wide composition. The visual style, set layout, hallway walls, warm apartment lighting, and sitcom staging should feel extremely faithful to the original series. CRITICAL CAMERA FRAMING RESTRICTION: Use ONLY extremely tight close-ups, tight two-person close-ups, and individual close-ups. Absolutely no medium-wide shots, wide shots, waist-up two-shots, full-body shots, distant framing, establishing wide shots, or camera pullbacks at any moment. Sheldon Cooper’s face and Penny’s face must remain large, sharp, high-definition, and fully recognizable throughout the entire 12 seconds. Whenever Sheldon Cooper and Penny are visible together, use an EXTREMELY TIGHT TWO-PERSON CLOSE-UP: frame both characters approximately from the upper chest or shoulders upward, with both complete faces occupying most of the 16:9 frame. Keep Sheldon Cooper and Penny close together near the apartment doorway and on nearly the same focal plane. Both heads, both complete faces, eyes, mouths, and facial expressions must remain fully inside the frame and sharply resolved. Allow only a small amount of hallway and doorway background around them. Do not zoom out or pull back to accommodate both characters. Instead, physically compose Sheldon Cooper and Penny closer together inside the doorway area so both faces remain large. This tight two-person close-up requirement applies especially to the OPENING SHOT and the FINAL SHOT. The final second must NOT become wider under any circumstances. CAMERA AND FRAMING: Begin immediately with an extremely tight two-person close-up showing Sheldon Cooper and Penny together from approximately the upper chest or shoulders upward. Sheldon Cooper stands immediately outside apartment 4B while Penny stands immediately inside the already-open doorway. Both faces occupy a large portion of the frame and remain simultaneously sharp. After the opening, use clean sitcom-style cuts into individual close-ups and tight medium close-ups that never frame below the upper chest. Keep faces extremely large and detailed. Camera movement should remain minimal and controlled, similar to polished sitcom coverage, with subtle slow push-ins of very small amplitude for comedic emphasis. Never create additional distance from the subjects. For the ending, return to an extremely tight two-person close-up with both Sheldon Cooper and Penny’s complete faces filling most of the frame. Do NOT pull out, zoom out, widen the lens, reveal their full bodies, or show additional hallway space during the final reaction. LIGHTING AND VISUAL STYLE: Use warm, bright, natural sitcom interior lighting. Penny’s apartment interior should glow warmly behind Penny, while the hallway is evenly lit in the familiar Big Bang Theory style. Both faces must be extremely well-lit, clearly visible, and flattering, with natural skin tones, crisp facial detail, clear eyes, and strong separation from the background. The overall visual style should feel like a professionally shot network sitcom scene, not cinematic noir, not over-stylized, and not dark. Use shallow-to-moderate depth of field so the environment remains recognizable while Sheldon Cooper and Penny’s faces stay exceptionally sharp. CHARACTER APPEARANCE AND MANNERISMS: Sheldon Cooper must look and behave exactly like Sheldon in the series: upright posture, slightly stiff body language, precise diction, intellectual smugness, mildly condescending confidence, quick blinking, intense focus, tiny head movements, tightly controlled hand gestures, and an overly serious tone even when saying ridiculous things. Sheldon Cooper should wear a typical Sheldon outfit, such as a bright geeky science-themed T-shirt layered over a long-sleeve shirt. Penny must look and behave exactly like Penny in the series: relaxed posture, highly expressive face, grounded and casual energy, playful sarcasm, fast reaction timing, expressive eyebrows, animated eye movements, slight head tilts, knowing half-smiles, and a naturally conversational speaking style. Penny should wear a casual everyday apartment outfit appropriate to the series, such as a fitted top and jeans or a soft casual blouse. EXAGGERATED FACIAL PERFORMANCE: Throughout the entire 12-second clip, both characters should use more animated, heightened sitcom expressions than before while remaining believable and human. Every comedic beat should register clearly through the eyes, eyebrows, mouth, cheeks, head position, and micro-reactions. Sheldon Cooper’s smugness should be visibly exaggerated: raised chin, narrowed eyes, tiny superior smile, lifted eyebrow, precise mouth movements, and self-satisfied pauses. When Sheldon Cooper realizes something is wrong, Sheldon Cooper’s transformation must be dramatic and instantly readable: smug smile disappears, eyebrows lift sharply, eyes widen, mouth slightly opens, head freezes, then Sheldon Cooper slowly turns toward the camera with unmistakable existential dread. Penny’s sarcasm should be equally animated: larger eyebrow raises, stronger skeptical looks, playful eye rolls, amused half-smiles, slight head tilts, and quick expressive changes. When Penny realizes that Penny’s sentence has already been written, Penny’s expression must visibly progress from casual confidence to confusion, then suspicion, then sudden wide-eyed realization. Penny’s final “Worse. Prompted.” reaction should combine alarm, disbelief, and deadpan comedic certainty. PERFORMANCE AND COMEDIC RHYTHM: Sheldon Cooper begins the scene in full lecture mode, speaking as though giving a factual correction to Penny. Penny listens with exaggerated amused skepticism. Sheldon Cooper delivers the first line with smug certainty and clipped precision. Penny responds with casual, direct, slightly teasing confidence. Sheldon Cooper replies with a dry, superior, mildly insulting punchline. Penny’s next line is the turning point: Penny suddenly becomes aware of the artificiality of the scene, pauses, looks toward the camera, and says the line with rapidly growing confusion. Sheldon Cooper’s expression visibly collapses from smug confidence into dawning existential panic. Sheldon Cooper then slowly looks toward the camera and says “We’re scripted?” with exaggerated worried disbelief. Penny delivers “Worse. Prompted.” with a horrified but matter-of-fact realization. Both Sheldon Cooper and Penny end by staring directly toward the camera in exaggerated shared existential horror. TIMING, SHOTS, AND EXACT DIALOGUE ASSIGNMENT: 0.0 to 2.2 seconds: Begin on an EXTREMELY TIGHT TWO-PERSON CLOSE-UP outside apartment 4B. Sheldon Cooper and Penny are framed together only from approximately the upper chest or shoulders upward. Both complete faces are very large in frame, sharply detailed, and positioned close together on nearly the same focal plane. Penny remains just inside the already-open doorway while Sheldon Cooper stands immediately outside. Use no establishing wide shot. The apartment doorway and a very small portion of the familiar hallway provide enough environmental context behind their large faces. Sheldon Cooper faces Penny with an exaggeratedly serious, self-important expression and says exactly once: “AI language models predict the next most statistically probable word.” Penny remains silent throughout Sheldon Cooper’s line. Penny reacts with animated skeptical eyebrows, a small amused smile, and a slight head tilt. 2.2 to 3.4 seconds: Cut to an extremely tight close-up of Penny from approximately shoulders upward. Penny’s face occupies most of the frame. Penny tilts Penny’s head, raises Penny’s eyebrows with exaggerated playful skepticism, gives Sheldon Cooper an amused look, and says exactly once: “So what am I about to say?” Sheldon Cooper remains silent during Penny’s line. 3.4 to 5.3 seconds: Cut to an extremely tight close-up of Sheldon Cooper. Sheldon Cooper’s face fills most of the frame. Sheldon Cooper lifts one eyebrow, produces an exaggerated tiny smug smile, slightly raises Sheldon Cooper’s chin, and delivers the line with precise intellectual superiority. Sheldon Cooper says exactly once: “Something intellectually disappointing.” Immediately after Sheldon Cooper finishes the line, add a brief classic sitcom audience laugh. Sheldon Cooper maintains the exaggerated smug expression for a fraction of a second to let the punchline land. 5.3 to 7.8 seconds: Cut back to an extremely tight close-up of Penny. Penny begins with a sarcastic amused expression, then Penny’s expression changes dramatically and visibly in stages. Penny’s smile fades. Penny’s eyebrows draw together. Penny’s eyes shift as if thinking. Penny suddenly stops moving for a beat, then Penny’s eyes widen. Penny slowly looks directly into the camera lens as if Penny has suddenly become aware of the artificial nature of the scene. Penny says exactly once: “Wait… somebody already wrote this sentence.” Penny’s delivery begins puzzled and ends genuinely disturbed. Add a smaller sitcom audience laugh that fades quickly into increasingly awkward silence. 7.8 to 9.7 seconds: Cut to an extremely tight close-up of Sheldon Cooper. Sheldon Cooper’s entire face remains sharply resolved and large in frame. Sheldon Cooper’s smug expression disappears completely. Sheldon Cooper’s eyebrows rise dramatically, Sheldon Cooper’s eyes widen, Sheldon Cooper’s lips part slightly, and Sheldon Cooper freezes for a beat. Sheldon Cooper slowly shifts Sheldon Cooper’s gaze directly toward the camera with exaggerated dawning existential dread. Sheldon Cooper says exactly once: “We’re scripted?” Sheldon Cooper’s tone should be anxious, confused, and deeply disturbed. 9.7 to 11.0 seconds: Cut to an extremely tight close-up of Penny. Penny keeps looking directly toward the camera. Penny’s eyes are wide, Penny’s eyebrows remain raised, and Penny’s face combines horrified realization with Penny’s characteristic dry comedic delivery. Penny says exactly once: “Worse. Prompted.” Penny delivers “Worse” with a short deadpan pause before “Prompted,” allowing the final word to land clearly. 11.0 to 12.0 seconds: Cut to an EXTREMELY TIGHT TWO-PERSON CLOSE-UP. Absolutely DO NOT widen the framing for this ending. Sheldon Cooper and Penny are positioned very close together near the doorway, framed only from approximately the shoulders upward. Both complete faces occupy most of the 16:9 frame simultaneously and remain exceptionally sharp, detailed, undistorted, and clearly recognizable. Sheldon Cooper and Penny both stare directly toward the camera lens in exaggerated shared existential horror. Sheldon Cooper’s eyes are extremely wide, eyebrows raised, jaw slightly dropped. Penny’s eyes are wide, Penny’s mouth slightly opens, and Penny exchanges one extremely quick horrified side-glance toward Sheldon Cooper before looking back into the camera. Neither Sheldon Cooper nor Penny says anything further. Add one final short sitcom audience laugh that initially reacts to “Worse. Prompted.” and then abruptly fades into awkward silence while Sheldon Cooper and Penny remain frozen in the extremely tight two-person close-up. IMPORTANT ENDING CAMERA RULE: During 11.0–12.0 seconds, maintain the exact tight facial scale. No camera pullback. No zoom out. No transition to a medium shot. No waist-up framing. No reveal of additional hallway space. No wider composition whatsoever. End directly on the large, sharply rendered faces of Sheldon Cooper and Penny. AUDIO DESIGN: No background music. No narration. No voice-over. No on-screen text. No watermark. No AI parody label. Use only the exact dialogue lines listed above, plus natural sitcom set ambience and classic audience laugh cues. Include subtle hallway room tone, faint apartment ambience, light clothing rustle, tiny foot shifts, and clean studio-style dialogue recording. The audience laughter should feel like a classic Big Bang Theory studio-audience reaction: brief, well-timed, and placed after the strongest joke beats, especially after “Something intellectually disappointing,” after “somebody already wrote this sentence,” and a final shorter laugh sting after “Worse. Prompted.” before the awkward silence. All dialogue must remain crystal clear and perfectly lip-synchronized. Sheldon Cooper’s voice must only come from Sheldon Cooper. Penny’s voice must only come from Penny. Never blend, swap, duplicate, overlap, or garble the voices. FINAL RESTRICTIONS: No wide shots. No medium-wide shots. No waist-up two-shots. No full-body framing. No distant faces. No opening establishing wide shot. No wider ending shot. No camera pullback. No zoom-out. No distorted faces. No extra dialogue. No repeated dialogue. No ad-libs. No narration. No subtitles. No captions. No on-screen text of any kind. No watermark. No AI parody disclosure. No music. No other characters entering the frame. No door knocking. The door is already open at the beginning. Maintain extremely tight facial framing, high-definition facial clarity, accurate speaker assignment, exaggerated animated expressions, natural sitcom pacing, faithful Sheldon Cooper and Penny mannerisms, and believable Big Bang Theory comedic rhythm throughout the entire 12-second clip.