Criar UGC com IA: um take vira as personas da sua marca no Seedance 2.5 Edit (com os prompts)

The same take, but the man ages in real time. The ONLY thing that changes is his age. RULE 0, PERFORMANCE 1:1: the source video is the full performance capture. Second N of the output is second N of the source: nothing retimed, sped up, slowed, frozen or re-acted. Every step, hand gesture, head turn and lip movement happens at the exact same moment as in the source, mouth in sync with the original audio. Same body position in the frame, same distance to the camera, same camera framing as the source (a phone on a fixed stand; he walks close to the lens and back). He keeps holding the small black fluffy microphone in his hand exactly as in the source. THE ROOM: a tall double-height modern living room: lilac sheer curtains on the left, a gray textured wall, a glass mezzanine balcony above, white sofa, orange armchairs, a round wooden coffee table, green cushions. Unless the timeline says otherwise, the room stays exactly the same. IDENTITY: the base man is in his early 30s, black rectangular glasses, short dark hair, short dark beard, navy blue polo shirt and blue shorts. Any transformation keeps his face structure, glasses and expression recognizable unless the timeline says the person changes. PERFORMANCE TRACK: 0-8.5s he is very close to the camera, talking and gesturing with both hands, around 4-5s he raises a hand next to his head; 8.5-9.8s he steps back away from the camera. He must not step back before 8.5s. AGE TIMELINE (seconds from the start of this clip): 0-2s: he looks exactly as in the source, but the camera does a fast dramatic push-in on his face, then settles into the source framing. 2-4s: he starts aging: wrinkles appear around the eyes and forehead, gray at the temples. 4-5s: his hair turns fully silver gray, deep wrinkles, he looks 65. 5-6s: his beard turns completely white. He is now a 70-year-old man with silver hair, white beard, same glasses, same navy polo. 6-7s: holds the old man look. 7-9.8s: he becomes young again, back to exactly how he looks in the source. Heavy, realistic aging (deep wrinkles, age spots, sagging skin), not a light filter. FAILURE (none of this may appear): stepping, leaning or walking at a different moment than the source; any gesture missing, early, late, mirrored or with the other hand; hands frozen where the source moves; mouth out of sync; the microphone disappearing; a transformation happening before its second mark; cuts; camera moving differently than the source; new people; film equipment; text, letters or logos except where asked; music.

The same take, but his look becomes Gen Z. The ONLY thing that changes is his style (hair and clothes). RULE 0, PERFORMANCE 1:1: the source video is the full performance capture. Second N of the output is second N of the source: nothing retimed, sped up, slowed, frozen or re-acted. Every step, hand gesture, head turn and lip movement happens at the exact same moment as in the source, mouth in sync with the original audio. Same body position in the frame, same distance to the camera, same camera framing as the source (a phone on a fixed stand; he walks close to the lens and back). He keeps holding the small black fluffy microphone in his hand exactly as in the source. THE ROOM: a tall double-height modern living room: lilac sheer curtains on the left, a gray textured wall, a glass mezzanine balcony above, white sofa, orange armchairs, a round wooden coffee table, green cushions. Unless the timeline says otherwise, the room stays exactly the same. IDENTITY: the base man is in his early 30s, black rectangular glasses, short dark hair, short dark beard, navy blue polo shirt and blue shorts. Any transformation keeps his face structure, glasses and expression recognizable unless the timeline says the person changes. PERFORMANCE TRACK: 0-1s he walks back away from the camera; 1-3s standing full body in the room doing the "6-7" hand gesture (both palms up, alternating up and down); 3-4s he steps forward close to the camera; 4-5s he walks back; 5-7s standing full body, talking. STYLE TIMELINE (seconds from the start of this clip): 0-1s: as he walks back, his look switches. 1-7s: Gen Z look: bleached platinum blond messy curly hair, a small silver earring, silver chain necklace, an oversized faded vintage t-shirt with a plain abstract tie-dye print and absolutely no letters or words, baggy light-wash cargo jeans, chunky white sneakers. Same face, same glasses, same beard trimmed shorter. Same room. FAILURE (none of this may appear): stepping, leaning or walking at a different moment than the source; any gesture missing, early, late, mirrored or with the other hand; hands frozen where the source moves; mouth out of sync; the microphone disappearing; a transformation happening before its second mark; cuts; camera moving differently than the source; new people; film equipment; text, letters, words or logos on clothes or anywhere; music.

The same take, but the person becomes a female fashion model. The ONLY things that change are the person and the outfit. RULE 0, PERFORMANCE 1:1: the source video is the full performance capture. Second N of the output is second N of the source: nothing retimed, sped up, slowed, frozen or re-acted. Every step, hand gesture, head turn and lip movement happens at the exact same moment as in the source, mouth in sync with the original audio. Same body position in the frame, same distance to the camera, same camera framing as the source (a phone on a fixed stand; he walks close to the lens and back). He keeps holding the small black fluffy microphone in his hand exactly as in the source. THE ROOM: a tall double-height modern living room: lilac sheer curtains on the left, a gray textured wall, a glass mezzanine balcony above, white sofa, orange armchairs, a round wooden coffee table, green cushions. Unless the timeline says otherwise, the room stays exactly the same. IDENTITY: the base man is in his early 30s, black rectangular glasses, short dark hair, short dark beard, navy blue polo shirt and blue shorts. Any transformation keeps his face structure, glasses and expression recognizable unless the timeline says the person changes. PERFORMANCE TRACK: the whole clip he is standing full body in the room, talking and gesturing with his hands, exactly as in the source. TIMELINE (seconds from the start of this clip): 0-0.5s: the base man, as in the source. 0.5-2.5s: he transforms into a beautiful Brazilian female fashion model in her late 20s: long wavy dark brown hair, natural makeup, slim, standing in the exact same pose, same gestures, same lip movement, wearing a plain white t-shirt and light jeans. She holds the same small black microphone. 2.5-6.2s: her outfit changes into an elegant emerald green silk midi dress with thin straps and gold earrings, like a fashion campaign. Same woman, same pose, same gestures. The woman moves her hands and mouth exactly like the man in the source, second by second. FAILURE (none of this may appear): stepping, leaning or walking at a different moment than the source; any gesture missing, early, late, mirrored or with the other hand; hands frozen where the source moves; mouth out of sync; the microphone disappearing; a transformation happening before its second mark; cuts; camera moving differently than the source; new people; film equipment; text, letters or logos except where asked; music.

The same take, but the object in his hand changes. The ONLY thing that changes is the object he holds. RULE 0, PERFORMANCE 1:1: the source video is the full performance capture. Second N of the output is second N of the source: nothing retimed, sped up, slowed, frozen or re-acted. Every step, hand gesture, head turn and lip movement happens at the exact same moment as in the source, mouth in sync with the original audio. Same body position in the frame, same distance to the camera, same camera framing as the source (a phone on a fixed stand; he walks close to the lens and back). He keeps holding the small black fluffy microphone in his hand exactly as in the source. THE ROOM: a tall double-height modern living room: lilac sheer curtains on the left, a gray textured wall, a glass mezzanine balcony above, white sofa, orange armchairs, a round wooden coffee table, green cushions. Unless the timeline says otherwise, the room stays exactly the same. IDENTITY: the base man is in his early 30s, black rectangular glasses, short dark hair, short dark beard, navy blue polo shirt and blue shorts. Any transformation keeps his face structure, glasses and expression recognizable unless the timeline says the person changes. PERFORMANCE TRACK: the whole clip he is close to the camera, holding a soda can in one hand and the black microphone in the other, tilting the can around 3-4s, exactly as in the source. OBJECT TIMELINE (seconds from the start of this clip): 0-0.5s: as in the source, but the can has no brand at all. 0.5-3s: the can becomes an elegant frosted glass skincare serum bottle with a gold dropper cap and a clean white label that reads "LUMÉ" in black serif letters. He holds and tilts it exactly as he held the can. 3-6s: the serum bottle becomes a single white leather sneaker with a gum sole, no logo, held in his hand by the heel, moving exactly as the can moved. The object is about the same size in his hand, his fingers wrap around it naturally. FAILURE (none of this may appear): stepping, leaning or walking at a different moment than the source; any gesture missing, early, late, mirrored or with the other hand; hands frozen where the source moves; mouth out of sync; the microphone disappearing; a transformation happening before its second mark; cuts; camera moving differently than the source; new people; film equipment; text, letters or logos except where asked; music.

The same take, but the place changes and then HE HIMSELF turns into a woman. There is ONLY ONE PERSON in the frame for the whole clip: him. Never add a second person. RULE 0, PERFORMANCE 1:1: the source video is the full performance capture. Second N of the output is second N of the source: nothing retimed, sped up, slowed, frozen or re-acted. Every step, hand gesture, head turn and lip movement happens at the exact same moment as in the source, mouth in sync with the original audio. Same body position in the frame, same distance to the camera, same camera framing as the source (a phone on a fixed stand). He keeps holding the small black fluffy microphone in one hand and the drink can in the other exactly as in the source. THE ROOM: a tall double-height modern living room: lilac sheer curtains on the left, a gray textured wall, a glass mezzanine balcony above, white sofa, orange armchairs, a round wooden coffee table. IDENTITY: the man is in his early 30s, black rectangular glasses, short dark hair, short dark beard, navy blue polo shirt and blue shorts. PERFORMANCE TRACK: 0-1s he is very close to the camera and turns his head away; 1-1.8s he steps back; 1.8-6.2s he stands full body in the middle of the frame holding the can and the microphone, talking and gesturing, exactly as in the source. TIMELINE (seconds from the start of this clip): 0-0.7s: exactly the source room. 0.7-1.8s: the room becomes UGLY: dingy cramped room, stained walls with peeling paint, a broken old sofa, cardboard boxes, a flickering fluorescent tube, dull greenish light. 1.8-3.1s: as he steps back, the ugly room is swept away into a breathtaking place. 3.1-4.3s: BEAUTIFUL LOCATION, real and three-dimensional, not a flat backdrop: he stands on a luxurious Mediterranean villa terrace at golden hour, white stone floor under his feet, a turquoise infinity pool and the sea behind him, warm sunlight and real shadows on him. Same polo, can and microphone. 4.3-5.2s: on the same terrace, HIS OWN BODY TRANSFORMS into a beautiful Brazilian female fashion model in her late 20s: long wavy dark brown hair, natural makeup, an emerald green silk midi dress. She stands exactly where he stood, in his exact pose, holding the same can and microphone, with the same gestures and lip movement. The man is gone while she is there; she replaces him, she does not appear next to him. 5.2-6.3s: she turns back into him, exactly as in the source, in the original living room. FAILURE (none of this may appear): a second person in the frame at any moment; a woman standing next to him or behind him; the terrace looking like a flat printed backdrop or a green screen; stepping or walking at a different moment than the source; any gesture missing, early, late or mirrored; mouth out of sync; the microphone or can disappearing; cuts; camera moving differently than the source; film equipment; text, letters or logos; music.

The same take, but he becomes ready for an advertising shoot. The ONLY things that change are his clothes, the object in his hand and, at the end, the room. RULE 0, PERFORMANCE 1:1: the source video is the full performance capture. Second N of the output is second N of the source: nothing retimed, sped up, slowed, frozen or re-acted. Every step, hand gesture, head turn and lip movement happens at the exact same moment as in the source, mouth in sync with the original audio. Same body position in the frame, same distance to the camera, same camera framing as the source (a phone on a fixed stand; he walks close to the lens and back). He keeps holding the small black fluffy microphone in his hand exactly as in the source. THE ROOM: a tall double-height modern living room: lilac sheer curtains on the left, a gray textured wall, a glass mezzanine balcony above, white sofa, orange armchairs, a round wooden coffee table, green cushions. Unless the timeline says otherwise, the room stays exactly the same. IDENTITY: the base man is in his early 30s, black rectangular glasses, short dark hair, short dark beard, navy blue polo shirt and blue shorts. Any transformation keeps his face structure, glasses and expression recognizable unless the timeline says the person changes. PERFORMANCE TRACK: 0-5s standing full body in the room holding a can in one hand; around 2.5-4.5s one hand goes up to his ear like a phone call; 5-6s he walks forward toward the camera; 6-7.4s close to the camera. He must not walk forward before 5s. TIMELINE (seconds from the start of this clip): 0-1s: his polo and shorts become a perfectly tailored slim navy blue suit, crisp white shirt, no tie, brown leather shoes. 1-2s: holds the suit look. 2-3s: the can in his hand becomes the frosted glass skincare serum bottle with a gold dropper cap and white label "LUMÉ". 3-5s: holds suit and serum bottle. 5-7.4s: as he walks to the camera, the living room transforms into a professional advertising studio: seamless light gray backdrop, large softboxes and studio light glowing around him (lights out of frame, only their glow), polished commercial look. He keeps the suit and the bottle. FAILURE (none of this may appear): stepping, leaning or walking at a different moment than the source; any gesture missing, early, late, mirrored or with the other hand; hands frozen where the source moves; mouth out of sync; the microphone disappearing; a transformation happening before its second mark; cuts; camera moving differently than the source; new people; film equipment; text, letters or logos except where asked; music.

Keep this video exactly as it is: same man, same navy suit, same serum bottle, same room turning into the advertising studio, same camera, same steps and gestures at the same moments. The ONLY thing that changes is what he says and his mouth. AUDIO: use the reference audio as the soundtrack and the dialogue of this video, word for word, at its exact timing. 0-4.6s: he speaks Portuguese exactly as in the source. From 4.6s: he speaks the Spanish line from the reference audio, "Y ahora sí, estoy listo para el comercial.", with his lips, jaw and mouth shapes perfectly lip-synced to the Spanish words, as if he is really speaking Spanish. FAILURE: the original Portuguese audio after 4.6s; mouth movement that does not match the Spanish words; any change to his face, glasses, beard, suit, the bottle, the studio, the camera or the timing of his walk and gestures; cuts; music; text.