夜街上的背负与轻声告白
A Piggyback Walk and a Quiet Request
夜色中,男子背着短发女子缓缓前行,两人的面部始终同框,她靠近他的肩头轻声说话。 原生视频规格为 15 秒 · 1944×1080。原帖完整公开了 Prompt,本页按原文保留。
プロンプト全文
r34l1sm subject_definitions: <Subject 1> is the adult man from @Image1, preserving his facial identity, tousled short black hair, natural skin texture, white shirt, dark plaid waistcoat, and suspenders. <Subject 2> is the adult elf woman from @Image2, preserving her facial identity, short dark bob with bangs, pointed ears, and oversized gray hoodie. Her expression is relaxed, affectionate, and quietly happy. @Image3 provides the intimate piggyback pose and teal-and-amber nighttime lighting. Use it for pose and atmosphere, giving the woman a warmer, more contented expression. summary: [reference generation] A fifteen-second affectionate love story in one continuous 16:9 two-person facial close-up. <Subject 1> slowly carries <Subject 2> through a quiet nighttime street. She softly asks to stay like this a little longer. He responds silently with a small smile and slower footsteps. Both faces remain simultaneously sharp. Her Japanese dialogue sounds naturally recorded within the street, with audible distance and environmental space. retention_analysis: <Subject 1> (appears in [Shot 1]): fully_preserved — retain his facial identity, hairstyle, and clothing. <Subject 2> (appears in [Shot 1]): fully_preserved — retain her facial identity, hairstyle, pointed ears, and hoodie. Her performance conveys comfort and affection. @Image3 ([Shot 1] pose and atmosphere): partially_preserved — retain the piggyback relationship and nighttime palette; adapt the framing into a horizontal facial close-up with greater depth of field and a quietly happy female expression. detailed_description: Live-action cinematic realism, 16:9 landscape, fifteen seconds. A 35mm cinema lens, fine film grain, soft highlight halation, warm natural skin tones, and cool teal nighttime surroundings. Use moderately deep depth of field with enough soft, motivated street lighting to render both faces clearly. [Shot 1] One uninterrupted take. A tight two-person facial close-up from slightly ahead and to the couple's left. <Subject 1>'s complete face occupies the lower-left center; <Subject 2>'s complete face appears above his right shoulder in the upper-right center. Include small portions of his collar and her gray sleeves. Her pointed ear remains visible through her short hair. Maintain a comfortable camera distance and natural facial proportions. Both faces remain simultaneously in sharp focus throughout the entire shot. Clearly resolve both characters' eyes, eyelashes, lips, and skin texture. Use sufficient depth of field to cover the distance between their faces and their gentle walking movements. The woman is as sharply focused as the man, including when she rests against his shoulder. Prioritize equal facial clarity over strong background blur. The street is gently softened, with recognizable shapes. No selective focus on the man, rack focus, or focus hunting. He carries her slowly along the street. The camera retreats at their pace, maintaining a nearly constant distance and the same facial framing. Its movement has a small, smooth breathing sway and gentle vertical drift motivated by careful footsteps. Fixed focal length, continuous perspective, no cuts, orbit, zoom, or abrupt reframing. During the opening seconds, he watches the path ahead with a relaxed expression. She rests comfortably against him, her brow smooth and her lips carrying a faint, natural smile. She briefly looks toward his cheek, then lowers her gaze toward his collar. Her expression conveys affection and ease, with no sorrow, tearful eyes, trembling lips, or worried brows. Neither looks into the camera. Near the middle, <Subject 2> (S1) turns her mouth slightly toward his ear and speaks naturally in a warm, gently smiling Japanese voice: <d>[Japanese] もう少し、このままで。</d> Hear her sentence from the camera's position, approximately one to two meters away. Her voice is modest in volume, slightly softened by distance, and blended with the surrounding street. Turning toward his ear makes it subtly less direct to the listener. Keep the words intelligible without making the voice unnaturally prominent. Avoid exaggerated whispering, amplified mouth sounds, and studio-isolated vocal clarity. He remains silent. After a brief listening pause, his cheek lifts into a small smile and he slows his next steps. She notices, her smile warms slightly, and she gently rests her cheek against his shoulder. Her lips remain closed after the line. Keep both faces sharp as a warm streetlight passes softly across them. Hold their comfortable closeness through the remaining seconds, with natural blinking and gentle walking motion. End while they are still moving together. No subtitles, narration, or additional dialogue. overall_soundscape: A continuous, spatially coherent nighttime street recording heard from the camera's perspective. Footsteps and subtle clothing rustles occupy the nearby space. A low storefront ventilation hum comes from one side. Distant traffic sits farther away; a passing car gradually approaches, moves across the stereo field, and recedes without obscuring the dialogue. The woman's voice belongs within this same acoustic environment. Keep street ambience continuously audible beneath her sentence, without muting it or dramatically lowering it when she speaks. Include faint, short reflections from nearby shopfront glass and building surfaces while retaining an open-air sound. No long reverberation, obvious echo, or indoor acoustic character. Breathing remains subtle at the camera's distance. Preserve one consistent listening perspective for dialogue, footsteps, clothing, and environmental sounds. Use natural vocal dynamics without heavy compression or artificial vocal enhancement. Maintain uninterrupted ambience before, during, and after the sentence. The man remains silent. non_diegetic_music: N/A. Let the location sound, pauses, and quiet footsteps carry the tenderness.
関連する事例
テキストから動画 デモ12 秒16:9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。
始点と終点のフレーム デモ15 秒16:9追逐午夜龙的碎月少年
少年 Kael 目睹银月裂开、午夜龙衔走碎片,随即穿过湖畔与森林追向禁忌山谷。 原生视频规格为 15 秒 · 2560×1440。原帖完整公开了 Prompt,本页按原文保留。
始点と終点のフレーム デモ15 秒16:9温室里的害羞雨云
小园丁 Luma 在旧水壶中发现害羞雨云 Nimbus,一滴雨让枯萎向日葵重新抬头。 原生视频规格为 15 秒 · 2560×1440。原帖完整公开了 Prompt,本页按原文保留。
テキストから動画 デモ15 秒629:360黄昏加油站旁的家庭录像
女子站在旧车与加油机旁对拍摄者说话,手持镜头留下倾斜和晃动。
テキストから動画 デモ15 秒16:9公寓门前的韩剧式对峙
两人在公寓内对峙,双人构图切换为各自近景。
テキストから動画 デモ15 秒16:9晨起女子与海景露台的一日开场
长发女子先坐在明亮卧室的床上整理头发,随后画面切到摆有木桌椅的海景露台,最后以落日海面和“Good vibes, Good life”文字收尾。 成片规格为 15 秒 · 1088×720。原帖完整公开了 Prompt,本页按原文保留。