DIRECTIVE — 30s cinematic single continuous take, 16:9, 24fps filmic motion cadence.
SUBJECT & STORY — [FORMAT] 16:9 aspect, 24fps, ARRI Alexa 35 capture, 17 stops of latitude, creamy highlight rolloff, natural skin rendition, Cooke S4/i glass, warm 'Cooke look', gentle skin falloff, soft creamy bokeh, 40mm near-normal, natural human perspective, mild depth separation, T2.8 aperture, photoreal live-action cinematography, real human skin texture, natural imperfections
[SCENE] INT. NEON PORT ROAD — NIGHT. LOCATION LOCK — NEON PORT ROAD: Wet asphalt reflects a dizzying kaleidoscope of neon signs as speeders blur past under a perpetually drizzling sky. This exact set, unchanged, in every shot of the scene. The duo escape in their muscle car, a blur against the city's electric glow.
[SUBJECTS] KAI — KAI, a stoic and formidable man in dark, utilitarian gear, with a scarred face and intense, watchful eyes. — wearing KAI in a reinforced charcoal grey tactical jumpsuit, obsidian boots, and darkened chrome gauntlets, ready for silent action. (identity locked to reference, do not alter face, hair, age or build) | NOVA — NOVA, a captivating pop star in a sleek, dark tactical suit, with glowing cybernetic elements and a fierce gaze. — wearing NOVA in a form-fitting obsidian tactical bodysuit with pulsing neon blue cybernetic accents and mirrored visor, a lethal pop sensation. (identity locked to reference, do not alter face, hair, age or build)
[BLOCKING] ONE CONTINUOUS TAKE: this 30-second clip is a single uninterrupted shot. No cuts, no edits, no dissolves, no jumps in time or place — the camera never stops rolling. The camera may move — dolly, pan, reframe, follow the actors — but it stays one operator, one roll, one piece of film. Play the action straight through in real time so the take can be cut afterwards in the edit. STORY CONTINUES: the previous scene ended with — NOVA and KAI ambush a convoy in a flash of neon and metal.. This take picks the film up from there: same characters, same wardrobe, same world, later in the same story. Do not re-introduce anyone and do not restart the film. It has to hand over cleanly to the next scene — NOVA performs amidst the kinetic chaos of the container yard battle. — so end the take on a frame that can cut straight into it. PERFORM TO @audio1: the attached audio is the exact 30-second window of the film's song that this take covers. Lock the performance, lip movements and cutting rhythm to it — singing, breaths and body movement land on the beat, and the take starts and ends exactly with the music. [REFERENCE MAP — bind each reference image to exactly one thing and never mix them] @image1 = KAI (Silent, Hyper-Competent Partner / Driver) — canonical face, hair, body and age of this actor; @image2 = NOVA (Pop Sensation / Master Thief) — canonical face, hair, body and age of this actor; @image3 = KAI's ONLY approved wardrobe — Shadow Driver Gear; exact garments, materials, accessories, fit and colors; @image4 = NOVA's ONLY approved wardrobe — Cybernetic Siren Ensemble; exact garments, materials, accessories, fit and colors; @image5 = the LOCATION of this scene — Neon Port Road (set, not a person). Each @image is that one subject only: do not merge two references into a new person, do not swap who is who, and do not invent anyone who is not in this map. [CAST + WARDROBE LOCK — SOLE AUTHORITY] The reference map contains the canonical identity portrait and the ONE approved wardrobe plate for every actor: KAI (Silent, Hyper-Competent Partner / Driver), NOVA (Pop Sensation / Master Thief). Reproduce those exact faces, hairlines, bodies, skin tones, ages, garments, accessories, materials, fit and colors for the whole take and every scene. Never re-cast, re-age, beautify, re-style, change clothes, remove accessories, blend identities or invent a face. LIVED-IN WORLD: this is a real place on a real day, never an empty set. Populate the exterior with background life at cinematic depth: pedestrians crossing and reacting, parked and moving cars, taxis, a police cruiser, delivery vans, street signage, traffic lights, litter, wet asphalt reflections. Background people are unnamed extras: out of focus or turned away, correctly scaled, never mugging at camera and never resembling the named cast. The named cast in the reference sheets stay the foreground subjects, sharp and clearly readable at all times. ABSOLUTE CONTINUITY: same wardrobe, same light, same set, same time of day for the entire take.
[CAMERA] ARRI Alexa 35 capture, 17 stops of latitude, creamy highlight rolloff, natural skin rendition, Cooke S4/i glass, warm 'Cooke look', gentle skin falloff, soft creamy bokeh, 40mm near-normal, natural human perspective, mild depth separation, T2.8 aperture
[LIGHT] motivated Night light in a Cyberpunk, grimy, electric, fast-paced space, practical sources visible in frame, soft falloff into shadow
[ACTION] ACTION (one take, straight through):
The duo escape in their muscle car, a blur against the city's electric glow.
[DIALOGUE] MUSIC VIDEO — NO SPOKEN DIALOGUE. Nobody talks to camera and nobody holds a conversation. If a performer is on screen they sing/mouth the lyric to the track with believable lip movement and stage energy; everyone else performs physically — movement, choreography, gesture, reaction. Cut and move to the musical pulse: performance energy on the beat, visual hook repeated. No narration, no subtitles, no on-screen text.
[GRADE] kinetic, dangerous, triumphant, neon action thriller, film-grade contrast curve, protected highlights, no digital clipping
[NEGATIVE] identity drift, different face between shots, wardrobe change mid-scene, extra fingers, deformed hands, warped faces, duplicated people, subtitles, captions, watermark, logo, text overlay, cartoon, 3d render, video game look, plastic skin, over-smoothed skin, oversaturated colours, fisheye distortion, motion blur smear, low resolution
CINEMATOGRAPHY — shot on ARRI Alexa 35 with anamorphic primes, T2.0, shallow depth of field, organic lens flares, motivated practical lighting: soft key, negative fill, hard rim for separation. Filmic contrast curve, rich blacks with retained shadow detail, subtle 35mm grain, natural skin tones, no CGI plastic sheen.
GRADE — low-key cinematic colour science, teal-leaning shadows, warm practical highlights, film-print halation on speculars.
CAMERA — one unbroken take, micro-handheld breathing, no cuts, no transitions; performance carries the shot.
SOUND — diegetic in-camera ambience, close-mic detail on the hero action, no music bed unless specified, no voice-over unless specified.
AVOID — a single slow drifting take, static talking-head boredom, warped hands or faces, morphing geometry, on-screen text, watermarks, logos, subtitles, stock-footage look, over-saturated HDR.