subject_definitions: <Subject 1> is the woman from <Picture 1>, exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance with no identity drift, no face change, no beautify. <Subject 2> is the woman from <Picture 2>, exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance with no identity drift, no face change, no beautify. <Subject 3> is the man from <Picture 3>, exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance with no identity drift, no face change, no beautify. holding a silver tray of shot glasses filled with amber liquid. summary: [video editing + reference generation + audio reuse] The target video is a direct structural recreation of <Video 1> (14.09 s) with identical hard cuts, camera angles, framing, and timing. <Subject 1> (from <Picture 1>), <Subject 2> (from <Picture 2>), and <Subject 3> (from <Picture 3>) replace the original performers while preserving every action, pose, expression change, and spoken line. Exact dialogue, rooftop party environment, and tray props are retained. retention_analysis: <Subject 1> (appears in every shot): fully_preserved - exact face, tray-holding posture, and all actions/expressions from <Picture 1> locked with no drift. <Subject 2> (appears in [Shot 3], [Shot 5], [Shot 6], [Shot 7], [Shot 8]): fully_preserved - exact face. <Subject 3> (appears in every shot): fully_preserved - exact face, short dark hair, skin tone, body proportions, ethnicity, tray-holding posture from <Picture 3> locked with no drift. [Shot 1] Medium two-shot of <Subject 1> (S1) and <Subject 3> standing face-to-face on the wooden deck. <Subject 1> reaches with her right hand to adjust the collar of <Subject 3>’s white polo. She looks up at him with a slight smile and says, <d>[English] And who's this?</d> <Subject 3> stands holding his own tray, looking down at her. [Shot 2] At 00:02.000, hard cut to a tighter medium close-up of the same pair. <Subject 1> (S1) leans in, and kisses <Subject 3> on the lips. <Subject 3>’s eyes widen in surprise. [Shot 3] At 00:04.000, hard cut to a medium shot of <Subject 2> standing near the balloons, angry, pointing with her right index finger toward the pair off-screen and saying with an alarmed expression, <d>[English] That's my fiance!</d> [Shot 4] At 00:05.000, hard cut back to the tight two-shot of <Subject 1> and <Subject 3> still kissing; <Subject 3> remains wide-eyed and frozen, tray held steady. [Shot 5] At 00:06.500, hard cut to a wider three-shot showing <Subject 2> on the left, <Subject 1> in the center, and <Subject 3> on the right. <Subject 1> (S1) pulls back from the kiss, turns slightly toward <Subject 2>, and says, <d>[English] Guess we gotta make it even.</d> while squeezing her breast. [Shot 6] At 00:08.500, hard cut to a medium two-shot of <Subject 1> and <Subject 2>. <Subject 1> (S1) raises both hands, places them on either side of <Subject 2>’s face, and pulls her in for a kiss on the lips. <Subject 2>’s eyes widen in shock. [Shot 7] At 00:10.000, continues the kiss in the same framing; <Subject 1> holds <Subject 2>’s face firmly while <Subject 2> remains startled. [Shot 8] At 00:11.500, hard cut to a wider three-shot. <Subject 1> releases the kiss, lowers her hands, and looks slightly downward with a small satisfied expression. <Subject 2> stands beside her looking surprised. <Subject 3> stands on the right holding his tray, mouth open in continued shock. overall_soundscape: Rooftop evening ambience with distant city hum and soft string-light atmosphere continues throughout, matching <Video 1>. Physical sounds of tray handling and clothing movement are retained. non_diegetic_music: N/A

For the target video, at 0.00 seconds into the target video, <Picture 1> is fully referenced. subject_definitions: <Subject 1> = a woman from <Picture 1>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance from reference image. No identity drift, no face change, no beautify. <Subject 2> = a woman from <Picture 2>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance from reference image. No identity drift, no face change, no beautify. <Subject 3> = a man from <Picture 3>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance from reference image. No identity drift, no face change, no beautify. HARD COUNT: exactly 3 humans. No clones. No extra person. No fused bodies. summary: [reference generation] laundromat comedy skit with double-entendre dialogue. Exact Picture 1 (young woman), Picture 2 (mature woman), Picture 3 (man). Cinematic. Photoreal 4K. retention_analysis: <Subject 1> (entire clip): fully_preserved — Picture 1 face, hair, body, skin, proportions every frame. <Subject 2> (from 10s): fully_preserved — Picture 2 face, hair, body, skin, proportions every frame she appears. <Subject 3> (entire clip): fully_preserved — Picture 3 face, hair, body, skin, proportions every frame. ANATOMY: - 3 separate complete bodies. Clear outlines. No merge, no extra limbs. - Count: 3 heads, correct arms/legs. CAMERA: Match source shot-by-shot with hard cuts only. No pan, no orbit, no extra cuts, no slow motion. Faces required in frame as in source. Action (hard-cut multi-shot, total ~16s): 0:00–2:00 — [Shot 1 medium] <Subject 1> bent over open white washing machine loading clothes, ass toward camera, <Subject 3> standing left holding turquoise laundry basket looking at her and ask d[English] You got room for one more load?/d. Background: rows of teal washing machines. 2:00–3:50 — [Shot 2] <Subject 1> stands up, turns to face <Subject 3>, smiling. 3:50–6:00 — [Shot 3 close-up] tight on <Subject 1> face and upper chest, smiling and speaking to <Subject 3> (partially off-frame left) d[English] That depends how big it is./d.. 6:00–9:50 — [Shot 4 two-shot] <Subject 3> and <Subject 1> face each other, he holds basket and talk d[English] Medium?/d, both react with raised eyebrows / sly smile, <Subject 1> talk d[English] Might be a tight fit./d. 9:50–12:80 — [Shot 5 close-up] <Subject 2> standing behind folding table with stacked white towels, stern then wide-eyed expression, speaking d[English] You two are still talking about laundry, right?/d. 12:80–15:00 — [Shot 6 wide] <Subject 3> center holding basket with shocked wide-eyed open-mouth expression looking at camera while speak d[English] I was./d, <Subject 1> right leaning on machine with sly smile and speaking d[English] Mostly./d, <Subject 2> left background still at folding table. Audio: Diegetic laundromat ambient only. No music.

For the target video, at 0.00 seconds into the target video, <Picture 1> is fully referenced. subject_definitions: <Subject 1> = a woman from <Picture 1>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance. No identity drift. No face change. No beautify. <Subject 2> = a man from <Picture 2>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance. No identity drift. No face change. No beautify. HARD COUNT: exactly 2 humans. No clones. No extra person. No fused bodies. summary: [reference generation] auto repair shop garage, maroon car trunk. Exact Picture 1 + Picture 2. Mechanic writes on clipboard then talks to customer who turns and replies. Photoreal 4K. retention_analysis: <Subject 1> (entire clip): fully_preserved — Picture 1 face every frame. <Subject 2> (entire clip): fully_preserved — Picture 2 face every frame. ANATOMY: - 2 separate complete bodies. Clear outlines. No merge, no extra limbs. - Count: 2 heads, correct arms/legs. CAMERA: medium wide then tight close-ups matching original hard cuts, still. No pan, no orbit, no extra cuts, no slow motion. Faces required in frame: both when together, then each alone on cut. Action (one continuous take with same hard cuts): 0:00–1:00 — wide: <Subject 2> in dirty blue coveralls stands left of maroon car trunk writing on clipboard; <Subject 1> in black tank top + light blue denim shorts + black platform boots leans right hand on trunk, looking at him. 1:00–3:00 — hard cut close-up <Subject 2> looking down writing then up speaking d[Your rear end needs a lot of work.]/d. 3:00–4:00 — hard cut medium <Subject 1> facing camera/left, hand on trunk and ask d[Which one?]/d. 4:00–5:00 — hard cut rear ¾ of <Subject 1> looking back over left shoulder. 5:00–6:00 — hard cut close-up <Subject 2> neutral face looking down then up. 6:00–8:00 — hard cut medium <Subject 1> both hands on trunk speaking d[Because this one is still under warranty?]/d. then surprised expression. 8:00–end — hard cut close-up <Subject 2> looking down writing then up speaking d[Fair enough.]/d. END LOCK: mid-action, exact faces, count still 2. Audio: diegetic garage ambience only. No music.

subject_definitions: <Subject 1> = A naked woman from <Picture 1>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance from reference image. No identity drift, no face change, no beautify. in the scene she is bent forward over the counter with both hands braced on it. <Subject 2> = A woman from <Picture 2>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance from reference image. No identity drift, no face change, no beautify. Who walks in through the front door later in the scene. <Subject 3> = A naked man from <Picture 3>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance from reference image. No identity drift, no face change, no beautify. standing behind <Subject 1> with both hands on <Subject 1>'s buttocks. HARD COUNT: exactly 3 humans. No clones. No extra person. No fused bodies. summary: A cinematic, landscape-orientation adult comedy scene, shot in a single continuous take, that begins from fast-food-restaurant composition with <Subject 1> and <Subject 3> is a naked man. Behind the counter, <Subject 3> has vaginal sex with <Subject 1> from behind with both hands on her buttocks and keeps thrusting even after <Subject 2> enters; on an electronic door ding, <Subject 1> smiles, turns her head, and greets a customer, after which the camera whip-pans left to reveal <Subject 2> entering through the front door, who stops, looks back at <Subject 1> in confusion, slap <Subject 1>'s face, and confronts her while the two behind the counter keep going. detailed_description: The target video is a cinematic live-action comedy rendered in a bright, evenly lit fast-food-restaurant interior with clean, saturated colors and a shallow depth of field. No on-screen text appears. [Shot 1] The shot begins from the composition of fast-food-restaurant, The frame holds a locked medium shot at counter height during the opening beats, then whip-pans left with no other moves. In the foreground, a beige counter with a black POS terminal sits in front of <Subject 1>, her breasts spilling over the top as she smiles at the camera. Behind her stands <Subject 3>, a naked man, both hands pressed on <Subject 1>'s buttocks, thrusting his hips back and forth as he has vaginal sex with <Subject 1> from behind, standing doggy style. Backlit menu boards with burger and fries photos glow along the top, and a glass refrigerator case stands to the left. an electronic door ding is heard, and <Subject 1> smiles wider and turns her head to the left of the screen; with an energetic, cheerful tone, <Subject 1> says "Welcome to Omex Burger." Exactly as the line ends, her lips close and her speaking motion ceases. The camera whip-pans left with fast amplitude and speed toward the front of the restaurant, revealing <Subject 2>, she walking in through the open front door, while <Subject 3> keeps thrusting and does not stop. <Subject 2> stops and looks over toward <Subject 1> with a confused, disapproving expression; as she speaks, she reaches out with one hand and slap <Subject 1>'s face with her left hand, making <Subject 1> flinch and let out a sharp moan; with an older, indignant feminine voice, <Subject 2> says, "God dammit, you let my husband fuck you?" as <Subject 3> continues pumping his hips behind <Subject 1> without pausing. The video ends on <Subject 2>'s stare at <Subject 1> then chokes her neck, while <Subject 3> is still thrusting. Audio: The bright ambient hum of the fast-food-restaurant interior fills the clip, with the soft hiss and sizzle of food frying in the kitchen and a short electronic door ding as the front door opens when <Subject 2> walks in; the continuous muffled thumps of <Subject 3>'s hips and <Subject 1>'s soft moans run throughout. A playful, whimsical cartoon-style score: crisp, bouncy pizzicato strings driving a lighthearted skipping rhythm, accented with cheerful staccato woodwinds, kept at a low, non-diegetic dynamic so it complements the visual comedy without overpowering the scene.

For the target video, at 0.00 seconds into the target video, <Picture 1> is fully referenced. subject_definitions: <Subject 1> = a woman from <Picture 1>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance from reference image. No identity drift. No face change. No beautify. <Subject 2> = a man from <Picture 2>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance from reference image. No identity drift. No face change. No beautify. HARD COUNT: exactly 2 humans. No clones. No extra person. No fused bodies. summary: [reference generation] kitchen counter conversation. Exact Picture 1 woman + Picture 2 man. Woman struggles with pickle jar, flirty dialogue exchange, same hard-cut timing and camera angles as source. Photoreal 4K. retention_analysis: <Subject 1> (entire clip): fully_preserved — Picture 1 face, hair, body, every frame. <Subject 2> (entire clip): fully_preserved — Picture 2 face, hair, body, every frame. ANATOMY: - 2 separate complete bodies. Clear outlines. No merge, no extra limbs. - Count: 2 heads, correct arms/legs. CAMERA: Medium two-shot, still, matching source framing and hard cuts. No pan, no orbit, no extra cuts, no slow motion. Both faces required in frame throughout. Action (one continuous take, ~10s): 0:00–0:01 — Woman (<Subject 1>) at kitchen island gripping pickle jar lid with both hands, strained expression, saying “I can’t.” Man (<Subject 2>) stands left of her, looking down at jar. 0:01–0:03 — Man asks “How tight is it?” while gesturing toward jar. Woman looks up at him, still holding jar. 0:03–0:05 — Woman replies “Why don’t you find out how tight it is?” with playful expression, one hand on counter, other near jar. Slight camera hard-cut to slightly wider framing. 0:05–0:07 — Man leans in, points at jar, asks “You’re talking about the jar, right?” Woman looks down at jar then back up. 0:07–0:10 — Woman smiles knowingly at man, shifts weight, hand on hip, jar still on counter. End lock: both faces exact, same kitchen, same positions. Audio: Diegetic kitchen ambience only. No music.

[reference generation] kitchen comedy confrontation. Exact Picture 1 young woman + Picture 2 mature woman + Picture 3 muscular man. Same hard-cut sequence and dialogue as source. Photoreal 4K. retention_analysis: <Subject 1> (entire clip): fully_preserved — Picture 1 face, hair, body, skin every frame she appears. <Subject 2> (from entrance onward): fully_preserved — Picture 2 face, hair, body, skin every frame she appears. <Subject 3> (entire clip): fully_preserved — Picture 3 face, hair, body, skin every frame he appears. ANATOMY: - 3 separate complete bodies. Clear outlines. No merge, no extra limbs. - Count: 3 heads, correct arms/legs only. CAMERA: Same hard-cut sequence and angles as source video. Still shots, no pan, no orbit, no slow motion. Faces required in frame when on screen. Action (hard-cut sequence matching source timing ~10s): 0:00–1:00 — Kitchen wide: <Subject 3> leans against counter looking at phone. <Subject 1> stands close facing him, body angled toward him, one leg slightly raised, looking up at his face as if just kissed or about to. 0:01 — <Subject 1> steps slightly back, speaks: d[Morning.]/d 0:02 — Hard cut close-up <Subject 3> looking down at phone, replies: d[Morning, love.]/d 0:02–0:03 — Hard cut doorway: <Subject 2> enters holding mug, surprised expression, speaks: d[Excuse me?]/d 0:04 — Hard cut extreme close-up <Subject 3> face, eyes wide, mouth open in shock. 0:05–0:06 — Hard cut medium: <Subject 2> stands left in doorway holding mug, accusatory, speaks: d[Did you just kiss my husband?]/d <Subject 1> on right at marble counter, back partially turned, looking down. 0:07–0:10 — Hard cut close-up <Subject 1> at counter, turns head toward camera with raised eyebrow and cheeky smirk, speaks: d[You want one too?]/d END LOCK: mid-dialogue cheeky expression on <Subject 1>, exact faces, count still 3. Audio: Diegetic only. Kitchen room tone, soft footsteps, fabric rustle. No music.

integrated_multimodal_description: [Shot 1] Live-action, cinematic, a medium close-up of the scene, captured from a front-facing angle, showcasing a naked man from <Picture 3> and the naked woman from <Picture 1> positioned kneeling in front of him, actively sucking on his erect penis while gripping its shaft with both hands. Behind them, on the wall, a banner made of paper letters suspended by string reads "HAPPY 18TH BIRTHDAY," with the man's head partially obscuring part of the text. Simultaneously, a mature woman from <Picture 2> enters the room from the right, her expression shocked as she looks past the frame initially, then rapidly shifts her gaze first to the woman from <Picture 1>, then to the man from <Picture 3>, repeating the shift between them while murmuring in shock, "O... my... god...". Initially, the woman from <Picture 1> pauses sucking, withdraws the penis from her mouth, and turns to face the woman from <Picture 2> while still holding the penis in her hands. The man from <Picture 3> asks the woman from <Picture 2>, "What's wrong, honey?", his head still blocking the first digit of the wall lettering. The mature woman replies in a slightly subdued voice, "You found your birthday gift... It should to be surprise.". All three individuals turn their attention toward the direction of bottom right corner of the frame where woman from <Picture 2> was initally looking, and now the man's head no longer obscures the banner, revealing the digits "48," making the full inscription read "HAPPY 48th BIRTHDAY." The camera abruptly whips around to the right to show the opposite corner of the room, where a large, human-sized box stands. It's already unwrapped and opened by the right side but the inscription "Stepdaughter for Rent — Delight Your Husband" is prominent on its front side. After a few moments, the camera sharply whips back, returning to the original composition. The woman from <Picture 2> approaches the man from <Picture 3>, gently kisses him on the lips, and speaks in a cheerful voice to him while the woman from <Picture 1> reinserts his penis into her mouth and resumes sucking, the woman from <Picture 2> saying, "Happy Birthday, my love!". She then addresses the woman from <Picture 1>, placing a hand on her nape and saying, "No hands, darling, he likes it deeper.". The woman from <Picture 2> presses one her hand onto the woman from <Picture 1> head, forcing her to move her hands behind her back and to touch the man's groin with her nose taking the entire length of the penis inside her mouth. overall_soundscape: The sound of wet sucking and deep swallowing is audible, accompanied by the muted murmur of the mature woman's shocked utterance and the subsequent clear, spoken dialogue between the characters. non_diegetic_music: A moderately paced, warm orchestral swell begins subtly during the initial shock and builds slightly as the camera whips around to the gift box, settling into a gentle, rhythmic underscore during the final affectionate exchange.

For the target video, at 0.00 seconds into the target video, <Picture 1> is fully referenced. subject_definitions: <Subject 1> = a woman from <Picture 1>. Exact face, hairstyle, skin tone, body proportions, ethnicity, overall appearance. No identity drift. No face change. No beautify. <Subject 2> = a woman from <Picture 2>. Exact face, hairstyle, skin tone, body proportions, ethnicity, overall appearance. Black French maid corset dress with white lace trim, white apron, white maid headband, black choker, black high heels. No identity drift. No face change. No beautify. <Subject 3> = a man from <Picture 3>. Exact face, hairstyle, skin tone, body proportions, ethnicity, overall appearance. No identity drift. No face change. No beautify. HARD COUNT: exactly 3 humans. No clones. No extra person. No fused bodies. summary: [reference generation] kitchen table comedy. Exact Picture 1 + Picture 2 + Picture 3. a wife and husband eating spaghetti; sexy maid enters, sits on husband's lap and eats from his plate while wife protests. Photoreal 4K. Same hard-cut sequence and camera angles as source. retention_analysis: <Subject 1> (entire clip): fully_preserved — Picture 1 face every frame. <Subject 2> (entire clip): fully_preserved — Picture 2 face every frame. <Subject 3> (entire clip): fully_preserved — Picture 3 face every frame. ANATOMY: - 3 separate complete bodies. Clear outlines. No merge, no extra limbs. - Count: 3 heads, correct arms/legs. CAMERA: Same static wide kitchen shots + tight face close-ups with hard cuts only. No pan, no orbit, no slow motion. Faces required in frame as per original shot sequence. Action (one continuous take matching original timing and hard cuts): 0:00–2:00 — Wide shot of round kitchen table. <Subject 1> (left) and <Subject 3> (right) seated, eating spaghetti with forks. <Subject 2> stands in doorway behind them holding empty silver tray, then walks forward between them. 0:02–4:00 — Hard cut to medium close-up of <Subject 1> looking down at her plate, speaking: d[Sorry, there are no extra chairs.]/d 0:04–5:00 — Hard cut to medium close-up of <Subject 2> smiling, holding tray: d[That's fine.]/d 0:05–7:00 — Hard cut back to wide shot. <Subject 2> places tray on table, walks behind <Subject 3>, sits directly on his lap with arm around his shoulder, picks up his fork and starts eating spaghetti from his plate. 0:07–9:00 — Hard cut to tight close-up of <Subject 1> turning head sharply right, eyes wide, mouth open shouting: d[Get off from my husband!]/d 0:09–12:00 — Hard cut to medium two-shot of <Subject 3> (left, surprised expression) and <Subject 2> (right, still seated on his lap, arm around him, chewing spaghetti and speaking): d[There's nowhere else to sit.]/d then She continues eating calmly. END LOCK: mid-action, exact faces from Pictures 1/2/3, count still 3, same framing. Audio: Diegetic kitchen ambience, fork clinks, chewing, speech only. No music.

For the target video, at 0.00 seconds into the target video, <Picture 1> is fully referenced. subject_definitions: <Subject 1> = adult young woman from <Picture 1>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance. No identity drift. No face change. No beautify. <Subject 2> = adult woman from <Picture 2>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance. No identity drift. No face change. No beautify. <Subject 3> = adult man from <Picture 3>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance. No identity drift. No face change. No beautify. HARD COUNT: exactly 3 humans. No clones. No extra person. No fused bodies. summary: [reference generation] kitchen reunion hug scene. Exact Picture 1 (young woman), Picture 2 (older woman), Picture 3 (man). Photoreal 4K. retention_analysis: <Subject 1> (entire clip): fully_preserved — Picture 1 face every frame. <Subject 2> (when on screen): fully_preserved — Picture 2 face every frame. <Subject 3> (entire clip): fully_preserved — Picture 3 face every frame. ANATOMY: - Exactly 3 separate complete bodies. Clear outlines. No merge, no extra limbs. - Count: 3 heads, correct arms/legs. CAMERA: Medium shot, Hard cuts only. No pan, no orbit, no extra cuts, no slow motion. Faces required in frame: all on-screen subjects. Action (one continuous take with hard cuts matching source): 0:00–1:00 — Kitchen. <Subject 1> (holding feather duster) jumps into <Subject 3>’s arms. He lifts her. She wraps arms around his neck, legs around his waist. Both smiling. She says d[Dad you're home!]/d 1:00–4:00 — Close medium shot of the embrace. <Subject 3> holds <Subject 1> tightly, both laughing and smiling face-to-face. He says d[I miss you.]/d 4:00–5:00 — Hard cut to <Subject 2> standing at kitchen counter, yellow cloth in hand, looking toward them, saying d[Okay, that's enough.]/d 5:00–7:00 — Hard cut back to couple. <Subject 3> still holding <Subject 1>, both look surprised toward camera/off-screen. He lowers her to the floor. Feather duster falls. 7:00–8:00 — <Subject 1> crouches, picks up the feather duster from the floor. 8:00–9:00 — <Subject 1> stands facing <Subject 3>, holds duster, gestures with hand. 9:00–10:00 — Medium shot of <Subject 1> in foreground holding duster, <Subject 3> behind her. She looks forward and says d[Same time tomorrow?]/d END LOCK: mid-dialogue, exact faces of Picture 1/2/3, count still 3, same kitchen background. Audio: Diegetic only. No music. Dialogue exactly as written with d[ ]/d tags. Natural kitchen ambience.

For the target video, at 0.00 seconds into the target video, <Picture 1> is fully referenced. subject_definitions: <Subject 1> = a woman from <Picture 1>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance. No identity drift, no face change, no beautify, no CGI look. <Subject 2> = a man from <Picture 2>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance. No identity drift, no face change, no beautify, no CGI look. HARD COUNT: exactly 2 humans. No clones. No extra person. No fused bodies. summary: [reference generation] comedy confrontation scene in luxury bathroom doorway / living room. Exact Picture 1 woman + Picture 2 man. Woman confronts towel-clad man, dialogue about twin, punches him, then reveals twin brother. Photoreal 4K. retention_analysis: <Subject 1> (entire clip): fully_preserved — Picture 1 face every frame. <Subject 2> (entire clip): fully_preserved — Picture 2 face every frame. ANATOMY: - 2 separate complete bodies. Clear outlines. No merge, no extra limbs. - Count: 2 heads, correct arms/legs. - Man remains shirtless in white towel only. CAMERA: Medium front / slight side ¾, still. Both faces required in frame for dialogue and reaction shots. No pan, no orbit, no extra cuts, no slow motion. Action (one continuous take, ~16s): 0:00–0:02 — Woman (<Subject 1>) stands close to man (<Subject 2>) who is holding his white towel at waist, wet hair, shocked open-mouth expression. She glares up at him and says d[You bastard!]/d 0:02–0:06 — Man gestures with open palm, explains d[Baby, she looked just like you. You never told me you had a twin sister.]/d while woman listens with angry face. 0:06–0:08 — Woman swings right arm and punches/slaps man hard across the face. Man’s head snaps back slightly. 0:08–0:10 — Woman says d[I have a twin brother.]/d with smug upward glance and slight smile. 0:10–0:13 — Close reaction on man: wide eyes, open mouth, pure shock. He says d[Well, that explains the cock.]/d 0:13–end — Cut back to woman looking sideways with pursed lips / smug closed-mouth expression. END LOCK: mid-reaction, exact faces from Picture 1 and Picture 2, count still 2. Audio: Diegetic only. Room ambience, soft fabric movement, punch impact sound. No music. Dialogue exactly as written above with mouth sync.

subject_definitions: <Subject 1> = a young woman from <Picture 1>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance. No identity drift, no face change, no beautify, no CGI look. <Subject 2> = a middle-aged woman from <Picture 2>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance. No identity drift, no face change, no beautify, no CGI look, holding a coffee mug. <Subject 3> = a middle-aged man from <Picture 3>. Exact face, hairstyle, skin tone, body proportions, ethnicity, and overall appearance. No identity drift, no face change, no beautify, no CGI look. HARD COUNT: exactly 3 humans. No clones. No extra people. No fused bodies. summary: 12-second vertical 9:16 realistic live-action Indonesian sitcom scene in a warm family home entrance/living-room area near a staircase. Exact <Subject 1> from <Picture 1> stands beside a suitcase with <Subject 2> and <Subject 3>. Multi-camera sitcom coverage with natural dialogue, reactions, and a spontaneous coffee spit-take. Photoreal broadcast television quality. retention_analysis: <Subject 1> (entire clip): fully_preserved — Picture 1 face, hair, skin, body, and identity locked every frame, every shot. <Subject 2> (entire clip): fully_preserved — Picture 2 face, hair, skin, body, and identity locked every frame, every shot, coffee mug. <Subject 3> (entire clip): fully_preserved — Picture 3 face, hair, skin, body, and identity locked every frame, every shot. detailed_description: The target video is a realistic multi-camera sitcom in a warm American family home interior near a staircase, natural household lighting mixed with soft cinematic indoor light, vertical 9:16, professional broadcast television quality, locked-off camera with only subtle necessary moves. [Shot 1] Wide three-shot. <Subject 1> stands beside a suitcase next to <Subject 2> and <Subject 3>. <Subject 1> (S1) addresses <Subject 2> naturally and says <d>[English] Mom-Dad, i'm twenty one now.</d> <Subject 2> start sipped her coffee mug while looks at <Subject 1> with a slightly confused, concerned expression. <Subject 3> watches from beside them with a neutral sitcom expression. [Shot 2] At 00:02.500, cut to medium close-up of <Subject 1>. She continues speaking with casual, slightly impatient sitcom body language and natural facial expressions, <d>[English] It's time i spread my legs, and say goodbye!</d> [Shot 3] At 00:05.000, cut to close-up of <Subject 2>. She reacted with an exaggerated, sitcom-style look of surprise; as she sipped her coffee, he accidentally spat a bit of it back into and around the cup, her face registering spontaneous shock. Coffee droplets appear realistically around her mouth. No cartoon effects. [Shot 4] At 00:06.500, cut to medium close-up of <Subject 3>. He reacts with a dry, confused sitcom expression and says <d>[English] It's wings honey, spread your wings!</d> then pauses briefly, trying to process the situation. [Shot 5] At 00:08.500, cut back to medium close-up of <Subject 1>. She gives a deadpan, slightly awkward reaction, looking toward her parents as if realizing the moment has become ridiculous. [Shot 6] At 00:10.500, return to the wider three-shot. <Subject 2> looks down toward the suitcase then back at <Subject 1>. <Subject 3> reacts with understated disbelief. Hold the final comedic reaction beat. overall_soundscape: Quiet warm household room tone, soft natural indoor ambience, subtle clothing and mug sounds, realistic coffee spit, no added effects. non_diegetic_music: N/A