Total duration: 15 seconds. Aspect ratio: 9:16 vertical video. Number of shots: 5 shots. Format: sma

Total duration: 15 seconds. Aspect ratio: 9:16 vertical video. Number of shots: 5 shots. Format: smartphone vertical video for TikTok / Reels / Shorts. Visual style: natural live-action cinematography, contemporary Tokyo gallery opening party, small-scale comedy with restrained cinematic realism. Genre: modern comedic drama. Tone: witty, socially tense, awkwardly funny, elegant but uncomfortable. Main emotion: social politeness collapsing into brutally honest criticism. Target viewer reaction: immediately understand the social situation, then laugh at the awkward honesty and the silent shock. Reference image roles: @Image 1 is the primary Vertical Manga-Cinematic Shot Sheet. Use it for shot rhythm, panel importance, camera direction, subject movement, emotional beats, and manga-to-cinema timing. @Image 2 is the Character Lock Sheet. Use it to preserve the identity, face, hairstyle, body type, outfit, silhouette, age impression, and expression traits of all three characters. @Image 3 is the Vertical Look / Background Sheet. Use it for the modern Tokyo gallery space, lighting, color palette, atmosphere, background depth, and material texture. @Image 4 is the Vertical Camera & Motion Map. Use it for camera movement, push-in / pull-back, eye-line direction, hand movement, object movement, subject path, and timing. The reference images are directing documents, not visual overlays. Do not render any storyboard panels, manga frames, arrows, labels, handwritten notes, speech bubbles, focus lines, speed lines, or SFX text in the final video. Translate those visual marks into cinematic timing, motion, rhythm, emotion, and camera direction only. Story setting: A small contemporary art gallery opening party in Tokyo at night. Only three main characters are present in the scene: 1. The famous guest / celebrity poet, confident and slightly self-important, holding a phone or small printed card with their own poem. 2. The protagonist, socially awkward but principled, refusing to flatter anyone. 3. The mediator / friend, anxious and trying to keep the atmosphere polite. Do not add extra speaking characters. Background guests, if absolutely necessary, must remain vague, blurred, and non-essential. The focus must stay on the three main characters. Dialogue: Use Japanese dialogue with natural timing. Keep subtitles optional only if the system requires them, but do not render large visible text unless requested. Celebrity poet: 「では、いっぺんだけ。」 Celebrity poet: 「とかいの夜に、こどくはほうせきとなる……」 Mediator: 「すてきですね、ね？」 Protagonist: 「いや、しょうじきに言うと、かなりたいくつです。」 Celebrity poet: 「……たいくつ？」 Protagonist: 「ことばが全部、かざりだけです。」 Shot 1 — 0.0s to 2.5s Shot size: vertical establishing medium-wide shot. Vertical composition: Top area: warm gallery lights, white ceiling track lights, soft reflections on glass or framed artwork. Center area: the three characters arranged in a triangular social formation. The celebrity poet stands slightly elevated in confidence, the mediator between them, the protagonist slightly apart. Bottom area: polished gallery floor, shoes, subtle party stillness, negative space. Subject action: the celebrity poet prepares to read their poem, lifting a phone or small card. The mediator smiles nervously. The protagonist watches without performing social enthusiasm. Eye direction: the poet looks outward, expecting admiration. The mediator looks from poet to protagonist. The protagonist looks directly at the poet. Hand / object action: poet raises the poem card or phone; mediator lightly grips a glass; protagonist keeps hands still. Camera movement: very slow push-in from a stable vertical composition. Depth movement: the background remains soft; the triangular relation becomes clearer as the camera advances. Emotional beat: elegant social setup, polite tension under the surface. Manga symbol interpretation: large opening panel becomes a clear hero composition; empty space becomes awkward social distance. Must remain unchanged: all character faces, hairstyles, outfits, body proportions, and relative identities from @Image 2. Must not appear: panel borders, manga arrows, labels, comic marks, extra prominent guests. Shot 2 — 2.5s to 5.5s Shot size: medium close-up of the celebrity poet. Vertical composition: Top area: gallery light halo and a hint of framed artwork. Center area: poet’s face and upper body, confident expression, mouth beginning to recite. Bottom area: hand holding phone or card, visible but not covering the face. Subject action: the poet performs the poem with theatrical self-importance, but not exaggerated slapstick. Eye direction: poet glances briefly down at the text, then looks up as if expecting admiration. Hand / object action: small graceful hand movement while reciting. Camera movement: slight slow push-in, stable and observational. Depth movement: background falls into shallow focus, isolating the poet’s vanity. Emotional beat: self-satisfied performance, comedic seriousness. Manga symbol interpretation: focus lines become shallow depth of field and a subtle push-in toward the poet’s face. Must remain unchanged: poet’s exact identity, clothing, hairstyle, facial structure. Must not appear: written poem text on screen, speech bubbles, visible manga effects. Shot 3 — 5.5s to 8.0s Shot size: two-shot / over-the-shoulder composition between mediator and protagonist. Vertical composition: Top area: quiet gallery wall, negative space, soft overhead light. Center area: mediator leaning slightly toward protagonist, trying to force a polite reaction; protagonist remains still and unsmiling. Bottom area: mediator’s tense hand holding a glass, protagonist’s relaxed but firm posture. Subject action: mediator smiles too brightly and says, “素敵ですね、ね?” The protagonist does not answer immediately. Eye direction: mediator looks at protagonist, begging silently. Protagonist looks toward the poet, then slowly turns eyes toward mediator. Hand / object action: mediator’s fingers tighten around the glass; protagonist’s hand barely moves. Camera movement: tiny lateral reframing or small push-in to emphasize social pressure. Depth movement: poet is visible blurred in foreground or background, still expecting praise. Emotional beat: social pressure, fake politeness, awkward pause. Manga symbol interpretation: repeated small reaction panels become tiny eye movements, breath, and a held pause. Must remain unchanged: character spacing and identity from reference images. Must not appear: comedy reaction graphics, text labels, speed lines, overacting. Shot 4 — 8.0s to 11.5s Shot size: close-up on protagonist, then slight reveal of poet’s reaction. Vertical composition: Top area: darkened negative space above protagonist’s head, showing emotional pressure. Center area: protagonist’s face, calm and brutally honest; eyes steady. Bottom area: protagonist’s mouth and subtle hand movement, minimal but decisive. Subject action: protagonist says, “いや、正直に言うと、かなり退屈です。” A beat of silence follows. Eye direction: protagonist looks directly at the poet, not at the mediator. Hand / object action: protagonist makes a small downward gesture, as if cutting through social performance. Camera movement: controlled push-in, stopping at the moment of the word “退屈.” Depth movement: background sound and movement feel muted; the gallery seems to freeze. Emotional beat: honest criticism lands like a social shock. Manga symbol interpretation: black fill or strong emphasis becomes a sudden reduction of background noise, tighter framing, and a heavy pause. Must remain unchanged: protagonist’s face, clothing, body shape, hairstyle, understated expression. Must not appear: dramatic comic shock marks, exaggerated facial distortion, slapstick effects. Shot 5 — 11.5s to 15.0s Shot size: three-character medium shot, vertical comedic tableau. Vertical composition: Top area: gallery lights and empty white wall, creating awkward silence. Center area: poet stunned, mediator frozen in panic, protagonist calm and almost innocent. Bottom area: small gestures: poet’s lowered card, mediator’s glass held mid-air, protagonist’s still hands. Subject action: poet repeats, “……退屈?” Protagonist answers, “言葉が全部、飾りだけです。” Mediator’s smile collapses. Eye direction: poet stares at protagonist. Mediator looks between them. Protagonist remains direct and sincere. Hand / object action: poet’s hand lowers slightly; mediator’s glass trembles subtly; protagonist remains composed. Camera movement: very slight pull-back, revealing the full social damage. Depth movement: background returns slightly, but everyone feels frozen. Emotional beat: comic aftermath, brutal honesty, elegant social disaster. Manga symbol interpretation: final large panel becomes a composed vertical tableau; silence and negative space become the punchline. Must remain unchanged: three-person blocking, character identities, gallery mood, modern Tokyo setting. Must not appear: extra characters reacting loudly, visible storyboard layout, manga symbols, captions, text overlays, distorted faces, shaky camera. Global visual instructions: Use realistic live-action lighting with soft gallery illumination, subtle shadows, natural skin texture, and restrained cinematic color. Preserve realistic human imperfections: slight facial asymmetry, natural skin texture, individual body shapes, believable posture, and clothing with lived-in detail. Avoid overly polished AI-model faces, plastic skin, perfect symmetry, fashion advertisement posing, or excessive beauty filters. Global vertical video rules: Keep faces, eyes, hands, and key emotional actions inside the central safe action zone. Do not place important information at the extreme top or bottom. Use close-ups and medium-close shots for smartphone clarity. Avoid large horizontal camera moves. Use slow push-ins, slight pull-back, eye-line cuts, and depth-based staging. Negative instructions: Do not render manga frames, storyboard panels, arrows, handwritten notes, labels, Japanese direction text, English labels, speech bubbles, focus lines, speed lines, SFX letters, comic overlays, or panel borders. Do not add extra main characters. Do not create distorted faces, overacting, shaky handheld chaos, random props, unrelated background events, excessive crowd activity, visible UI text, or cluttered gallery details.