Men's Jacket Virtual Try-On

PROMPT 1 — MiniMax H3 (15s, 9:16, Reference to Video)   subject_definitions:     <Subject 1> "streetwear model" — from #Image1: young man, early-to-mid 20s, dark       brown loosely curly/wavy hair with volume on top and shorter tapered sides, natural       tousled texture falling across the forehead. Olive warm skin tone, strong defined       jawline, high sculpted cheekbones, straight nose with a subtle bridge, full natural       lips, thick dark eyebrows with a clean arch, light hazel-green eyes with warm       undertone. Clean-shaven with very faint stubble along the jaw. Small gold hoop       earring in the left ear. Thin brown leather cord necklace resting at the       collarbone. Athletic lean build, broad shoulders. Calm confident expression with a       natural slight smolder. His face, bone structure, eye color, hair texture, skin       tone, and earring must remain identical in every frame — identity locked from       #Image1.     <Subject 2> "all-black streetwear outfit" — from #Image2: a complete men's all-black       streetwear look consisting of the following items, each described exactly as shown:       - Black leather bomber jacket: oversized fit, smooth black leather with subtle         sheen, pointed fold-down collar, silver-tone front zip closure, elastic ribbed         knit cuffs and waistband in black, slightly puffy sleeves with natural leather         creasing       - Black ribbed tank top: slim-fit ribbed cotton vest with scoop neckline, form-         fitting torso, visible ribbed knit texture, worn tucked or semi-tucked under         the trousers       - Black leather belt: smooth matte black leather strap with a rectangular silver         metal buckle, single-prong closure, clean minimal design       - Black wide-leg pleated trousers: tailored high-waisted fit at the top with a         single front pleat on each leg, full wide-leg drape, clean pressed crease,         cuffed hem at the ankles, structured woven fabric with a slight wool-blend         appearance       - Black chunky combat boots: lace-up ankle-height boots with thick platform sole,         yellow contrast stitching along the welt (Dr. Martens style), black leather         upper with 8-eyelet lacing, pull tab at the back heel, chunky rubber tread         outsole       Every piece must appear exactly as described from #Image2 — correct materials,       correct fit, correct hardware colors.   summary: keyframe completion + reference generation. #Image1 anchors <Subject 1>'s     exact face, hair, and identity in every frame — this is a Virtual Try-On vlog-style     showcase. #Image2 provides the complete outfit reference — every garment, accessory,     and boot must match the items shown in #Image2 precisely. The model from #Image1     wears the full outfit from #Image2. The output is a 15-second vertical 9:16 men's     streetwear bedroom vlog video with natural diffused indoor light, warm beige-toned     bedroom set, and casual confident energy.   retention_analysis:     <Subject 1>: fully_preserved (face, curly hair texture, hazel-green eyes, olive skin       tone, jawline, gold hoop earring, leather cord necklace — identical in all 4 shots,       no drift, no substitution)     <Subject 2>: fully_preserved (every outfit piece from #Image2 maintained — black       leather bomber jacket sheen and collar, black ribbed tank top texture, black leather       belt with silver buckle, black wide-leg pleated trousers with cuffs, chunky combat       boots with yellow stitching — all correct in every shot where visible)   detailed_description:     [Shot 1] Static medium shot, vertical 9:16. Warm beige-toned bedroom interior —     a wooden wardrobe with slatted doors visible behind on the left, a minimalist desk     with a lamp on the right, natural diffused light from a window screen-left casting     soft warm shadows. <Subject 1> stands centered in the frame from the waist up,     wearing the black ribbed tank top from #Image2 — the ribbed texture visible on his     torso, scoop neckline showing his brown leather cord necklace and collarbone. His     gold hoop earring glints in the left ear. He holds the black leather bomber jacket     from #Image2 on a wooden hanger in his right hand, presenting it toward the camera     at shoulder height — the jacket hangs open showing its silver zip, pointed collar,     elastic ribbed cuffs, and the smooth leather sheen catching the window light. He     looks directly into the lens with a slight confident smile, eyebrows relaxed,     hazel-green eyes engaging the viewer. His curly dark hair is naturally tousled.     Face is #Image1 — identical throughout.     [English] Okay this jacket is doing all the work. (S1) — casual, confident,     vlog-conversational tone, slight head nod at "all the work." Natural room tone,     warm and intimate.     [Shot 2] At 00:04.000, quick outfit transition — a fast vertical wipe or jump cut.     <Subject 1> now wears the full look from #Image2: the black leather bomber jacket     layered open over the black ribbed tank top, the black wide-leg pleated trousers     sitting at his natural waist with the black leather belt visible — silver rectangular     buckle catching light — the trousers draping with a clean pressed crease and cuffed     at the ankles above the black chunky combat boots with yellow contrast stitching and     thick platform sole. He stands steady in the bedroom with both hands casually in the     trouser pockets, the jacket falling open naturally showing the tank top underneath.     He looks down briefly at the boots — a natural self-check — then looks up at the     camera with an approving expression. Face is #Image1 — same eyes, same hair, same     earring, identity locked. The combat boots hit the floor with a single solid step     as he shifts his weight.     [English] The boots just lock it in. (S1) — relaxed, affirming tone, slight     smile, a small nod of approval. He emphasizes "lock it in" with a subtle hand     gesture pulling from the pocket.     [Shot 3] At 00:08.000, full-body shot, slightly wider framing showing head to boots.     <Subject 1> steps forward from near the wardrobe toward the camera with a relaxed     unhurried pace. The chunky combat boots from #Image2 are clearly audible — solid     deliberate footsteps on the bedroom floor. The wide-leg pleated trousers sway     naturally with each step, the leather jacket moves with his body, the belt buckle     glints. As he approaches mid-frame, he reaches to the desk and casually picks up     a pair of dark sunglasses, slides them onto the top of his head nestled into his     curly hair. He then turns 90 degrees to his left, presenting the side silhouette to     camera — the clean structure of the leather bomber jacket is highlighted in profile:     the collar line, shoulder drop, puffy sleeve, elastic cuff, and the jacket's length     falling at the hip. The wide-leg trousers create a strong vertical line below. He     holds the pose for a beat. Face is #Image1 in profile — same jawline, same earring     visible, same curly hair. Camera remains static, he walks toward it.     [Shot 4] At 00:12.000, slow smooth push-in from medium to close-up. The framing     starts at upper body — the leather bomber jacket collar, the ribbed tank top     neckline, the brown leather cord necklace resting on his collarbone all visible.     As the camera pushes in slowly, the gold hoop earring and the necklace cord become     detailed. The frame tightens to his face — <Subject 1>'s hazel-green eyes from     #Image1, his defined jawline, the slight stubble, the tousled curly hair. He smiles     confidently — a warm genuine smile reaching his eyes — and looks directly into the     camera. Face is #Image1 — identical, identity locked, final confirmation.     [English] That's the fit. (S1) — short, definitive, satisfied, with a small     decisive nod. At 00:14.500, the frame freezes — a clean freeze-frame on his smiling     face, holding for the final half-second. Slight film-grain texture appears on the     freeze-frame for a vlog-authentic finish.   overall_soundscape: Natural warm room ambience throughout — soft indoor tone, faint     distant traffic hum, the lived-in feeling of a bedroom. At 00:00.500, subtle leather     creak as he holds the jacket on the hanger. At 00:04.500, the solid thud of the     combat boot hitting the floor — heavy, grounded, deliberate. At 00:08.000–00:10.000,     rhythmic boot footsteps — chunky and audible on hard flooring, two to three steps. At     00:08.500, faint fabric rustle of the leather jacket moving. At 00:12.000, soft     leather creak from the jacket as the push-in begins. His voice is natural, close-mic,     warm and intimate — vlog proximity.   non_diegetic_music: Low-frequency street-style background instrumental — muted boom-bap     beat with a deep sub-bass line and sparse dusty snare hits, 80–85 BPM. The beat enters     softly at 00:00.000 and stays underneath the dialogue without competing. A warm muted     piano or Rhodes sample loops minimally. At 00:12.000, the music begins fading out. By     00:14.000, the beat has fully dissolved to silence before the freeze-frame. No vocals,     no lyrics. The vibe is laid-back, masculine, understated cool.   Exclusions:     No face change — <Subject 1>'s face from #Image1 must remain identical throughout,       including hazel-green eye color, curly hair texture, gold earring, and leather cord       necklace. No identity drift, no substitution.     No outfit changes beyond what is scripted — every piece from #Image2 must be worn       correctly in every shot from Shot 2 onward. No garment swaps, no color shifts.     No harsh or studio lighting — natural diffused indoor window light only.     No watermarks, no brand logos, no text overlays at any point.     No subtitles or captions.     No additional people in frame — <Subject 1> is the sole person.     No theatrical or exaggerated acting — natural, casual vlog energy throughout.     No camera shake except subtle organic handheld feel in Shot 4 push-in.     No background clutter — the bedroom is clean, minimal, warm beige-toned.     No wrong outfit items — no substituting pieces not shown in #Image2.     No face distortion or flickering identity between cuts.     No Chinese text or garbled characters on any surface.   Settings: MiniMax H3 | Reference to Video | 9:16 | 15 seconds