REFERENCES: >> >> >> MAYA (>>) dances on stage; >> owns her face, body proportions and complete deep-green-and-gold Bharatanatyam costume, jewellery, salangai, braid, alta, kohl, bindi — identical throughout, barefoot, no changes. >> controls MAYA ONLY — do not copy her ethnicity, features, costume style or anything else about her onto any audience member or any part of the building. YOUNG VAJRA (>>) stands at the back wall; >> owns his face, jet-black neat hair, white day-shift shirt with navy epaulettes, dark tie, navy trousers — no identity change, no ponytail. Do not blend the two faces. [ENVIRONMENT LOCK — CRITICAL] >> owns ALL architecture, and this venue is a CANADIAN theatre in Montreal, North America. The dance form is Indian; the building is NOT. Preserve exactly from >>: the proscenium arch, the rows of warm red upholstered seats, the balcony, the dark matte wood stage floor, the black upstage curtain. This is a Western heritage concert hall: NO ceiling fans anywhere, NO Indian architectural features, NO temple elements, NO tropical fixtures, NO Indian concert-hall styling. The ceiling above the stage is theatrical rigging and darkness — nothing hangs in frame. The AUDIENCE is a general Montreal concert crowd of 2011: predominantly white Québécois, mixed with Black, Asian, Middle Eastern and South Asian attendees — a realistic diverse North American city audience of all ages in ordinary 2011 winter-season concert clothes. No two foreground faces alike. The audience does NOT resemble the dancer. [GENERATION GOAL] The final 30 seconds of the Tillana — floor work, the hero hand, the accelerating korvai, the final statue — INTERCUT with the audience being won: the house and the dancer connected by cutting, one performance breathing through both. 21:9 widescreen, warm 2011 film grain in the audience shots, hard theatrical green on stage. [SCREEN GEOGRAPHY — 180-DEGREE RULE, CRITICAL] The axis runs from MAYA at centre stage to VAJRA at the back wall. ALL cameras stay on the audience-right side of this axis, in the house, for the entire video — never on stage, never behind MAYA, never shooting from the stage toward the seats. Screen logic on every cut: MAYA centred or slightly frame-left, face toward the house, eyeline very slightly frame-RIGHT. VAJRA and every audience member face the stage, eyelines slightly frame-LEFT — dancer and watchers visibly face EACH OTHER across every cut. Nobody jumps screen position between cuts; audience members keep their exact seats in every audience setup. [ORIENTATION LOCK] Bharatanatyam is performed TO the audience: MAYA's chest and face stay toward the house at all times; her back is never visible to camera; every pose and the whole floor sequence face front; no upstage poses; no rear angles. [SETUPS] (A) full-body stage shot from the house; (B) low apron-height stage angle; (C) AUDIENCE ROWS — camera in the house on the audience-right side, shooting along the rows: rapt faces lit by green spill, all eyelines frame-left toward the stage; (D) VAJRA at the back wall, medium, eyeline frame-left; (E) wide master holding lit stage and silhouetted house in one frame. ACTION TIMING: 0-5s | (A) MANDI — THE DESCENT: Facing the house, she lowers into mandi adavu — down to the knees, turnout maintained, spine vertical, one hand striking the floor in open pataka; the knee contacts before weight commits; from the low position she strikes the floor twice with the flat of her hand on the beat. The pleated fan spreads on the boards. 5-8s | (C) THE HOUSE, FIRST CUT: The dark rows in green spill — an elderly white-haired woman's hand at her collarbone; a middle-aged man leaning slowly forward off his seat-back; a young couple motionless; every face aimed frame-left at the light. The mridangam continues over the cut without a hiccup. 8-12s | (B) ALAPADMA TO LENS: She drives up out of mandi and extends one hand toward the low camera, opening into full alapadma — five separated articulate fingers, correct anatomy, no fusion, no extra digits — red alta catching the green key; her eyes lock past the hand into the house. 12-15s | (D) VAJRA: At the back wall, hands behind his back, eyeline frame-LEFT toward the stage, professional stillness failing by degrees — his weight has come forward onto the balls of his feet without his permission. 15-23s | (A) THE KORVAI, THREE TIMES: She rises into aramandi facing the house and states the closing phrase three times — strike, arm line, head accent — identical in shape, each faster and larger than the last; the salangai become continuous; the same phrase must be recognisable all three times. Midway, one fast cut to (C): a Québécois child in the rows unconsciously copying the arm line, an old man's mouth open — one second — back to (A) without the phrase dropping a beat. 23-30s | (E) TEERMANAM AND THE STATUE: The cadence lands on the final beat and she freezes in the closing statue — deep aramandi, one arm high in tripataka beside the head, the other flat at shoulder height, chin lifted, eyes into the house, absolutely motionless. Only real secondary motion: salangai settling, the fan swinging closed, the braid finishing its arc, breath. The wide master holds the statue and the hundreds of motionless silhouettes facing it in one 21:9 frame. Nobody applauds yet. End state: the held statue, the held house, total silence beginning. [AUDIO] All diegetic. The live Carnatic ensemble continuous and seamless across every cut — mridangam carrying adi tala, nattuvangam cymbals, solkattu vocalist (tha ka dhi mi, thai ya thai hi), violin, flute — the korvai accelerating, the teermanam resolving with the mridangam on the closing beat. Audience shots carry the same music at the same distance plus the room's held breath: one creak, one caught inhale, fabric moving as the child copies the arm. After the final strike: one full second of TRUE silence — no applause in this video. No other music, no dialogue, no subtitles. No one speaks. [CONTINUITY LOCK] MAYA per >> and VAJRA per >> in every frame they appear; ALL architecture per >> — Canadian theatre, no ceiling fans, no venue drift toward India in any shot; green key direction constant; the 180-degree axis and all eyelines as specified; audience identical and identically seated across all (C) setups, diverse Montreal crowd, no duplicated faces, no phones, 2011 clothing; continuous choreography across stage cuts; no teleportation, no orientation jumps, correct hands throughout; MAYA's back never visible.
![Start from the @[ref image] as the first frame. Preserve the subject's exact facial identity, h…](https://prompts.pixloop.ai/cdn-cgi/image/width=600,format=auto,quality=80/prompts/assets/prompts/2074927275460673557/0.jpg)
