I'm preparing some prompts for the imminent release of Seedance 2.5 And I'm stil…

Made in aesthetic: continuous vertical tracking shot following one woman through a crowded farmers market on a drizzly morning, 9:16, gimbal-smooth but handheld-textured, gray daylight, umbrellas and puddles, one take, no cuts. audio: market chatter, rain on canvas awnings, her footsteps through puddles, a busker's guitar passing in and out, coins clinking, no music. timeline: - 0-5s: The camera follows her from behind through a dense oncoming crowd: fifty distinct faces pass, each unique, no duplicates, no melted background people. She weaves, shoulder-checks politely, and high-fives a passing friend without looking: both hands must connect cleanly, ten fingers each. - 5-9s: She stops at a mirror stall: the camera catches her AND the crowd behind reflected accurately in three angled mirrors, each reflection consistent and correctly reversed, moving in sync with reality. She fixes her hair in the reflection; the reflection matches. - 9-13s: She buys cherries: exact coin count into an open palm (three coins, visible), takes the paper bag, and walks on eating one cherry, spitting the pit into her fist. Rain intensifies: drops bead and run correctly on her waxed jacket, puddle reflections ripple with each of her steps. - 13-15s: She turns to the camera, holds up the cherry bag, and smiles with realistic teeth. Behind her, the entire crowd keeps moving naturally. Hold. A pigeon lands, unbothered, at her feet. quality: single continuous take, zero duplicated or deformed faces in a dense crowd, anatomically correct hands in contact interactions, physically accurate triple mirror reflections, true water and coin physics, natural teeth, no artifacts anywhere.