Prompt | CinePixel

BMX

51Likes1.9KViewsJul 26, 2026

Reference prompt

Create a 15 second, 16:9 photorealistic cinematic restaurant video built around one seamless visual idea: a diner uses an invisible mixed-reality interface to seemingly summon a complete meal. ONE CONTINUOUS CAMERA ANGLE. Camera positioned slightly behind and above the left shoulder of a seated young adult diner with short natural brown hair, wearing a clean off-white long-sleeve shirt. Only the back of the head, shoulder and right hand are clearly visible; never reveal a full frontal face. The diner sits at a pale natural-wood restaurant table inside a warm modern casual café. Background: beige upholstered booth, narrow planter filled with realistic green foliage, muted wall artwork and a small dark wall outlet. Background softly defocused. CAMERA: full-frame 45–50mm lens aesthetic, shallow realistic depth of field around f/2.8, camera height approximately seated eye level. Mostly locked-off observational framing with tiny human operator breathing drift, never perfectly robotic. Natural 24fps motion, 180-degree shutter, subtle motion blur, realistic focus behaviour, gentle fine sensor grain. No artificial camera orbit, zoom or dramatic movement. 0–1.5 sec: empty tabletop. Diner casually raises the right hand and extends the index finger into the space above the table. 1.5–4.8 sec: a refined translucent augmented-reality control panel softly appears approximately 40 cm above the tabletop. It resembles premium contemporary glass UI rather than science-fiction holograms: several small rounded icons, two tall translucent vertical sliders, faint white symbols, subtle warm reflections and realistic light transmission. Interface remains properly anchored in 3D space. The diner naturally taps one control, drags a slider slightly and makes one final deliberate selection. Finger contact and UI reaction must align perfectly with no hand deformation. 4.8–5.3 sec: interface rapidly but elegantly fades away. 5.3–8.5 sec: almost immediately, a real restaurant server’s arm in a plain black sleeve enters from camera-right carrying a rectangular rustic wooden serving board. The server smoothly places it in the centre of the table. On it: one highly realistic gourmet burger with glossy toasted brioche bun, beef patty, lettuce and visible fresh fillings; crisp golden fries in a small metal container; a ceramic cup of red ketchup. Accurate food weight, shadows, reflections, crumbs and natural imperfections. Server adjusts the tray slightly, then withdraws. 8.5–10 sec: short quiet hold. Diner remains still, subtly looking toward the meal. 10–12 sec: server’s hand returns from frame-right carrying a condensation-covered classic dark glass cola bottle with a red vintage-style label. Place it upright immediately to the right of the serving board. Realistic glass reflections, liquid density and contact shadow. 12–15 sec: satisfying final hero hold on the complete meal from the same over-the-shoulder viewpoint. Slight natural diner head movement, warm restaurant ambience, no posing. LIGHTING: warm practical tungsten café illumination around 3000–3400K, soft overhead key, gentle highlights on burger bun and bottle, realistic shadow falloff, restrained contrast, believable skin and wood tones. Avoid CGI gloss, excessive bloom, oversaturation, warped fingers, floating food, unstable objects, fake text, visible cuts or impossible physics. AUDIO/MUSIC: understated modern lo-fi electronic groove with soft bass, muted percussion and tasteful minimal synth texture. Add subtle spatial UI taps/soft glassy clicks, faint restaurant room tone, serving-board contact on wood, gentle fries/plate movement and a convincing glass-bottle placement sound. Music remains sophisticated and low enough for tactile sound effects to sell the realism.

Output details

Media type: Video prompt
Task type: Text To Video
Dimensions: 1280 × 720
Aspect ratio: 16:9

Model: seedance-2

Continue exploring

All Prompts

Prompt details: Create a 15 second, 16:9 photorealistic cinematic restaurant video built around one seamless vis

Reference preview

More like this