Indicador

FORMAT / QUALITY Ultra-realistic 4K handheld smartphone footage, vertical 9:16, 10 seconds, 30 fps. Daytime indoor lighting with soft natural window light and realistic shadows. Slight natural handheld micro-shake throughout (authentic phone feel), subtle autofocus breathing with tiny focus-hunts on reactions, realistic motion blur, shallow depth of field. High-quality natural audio (room tone + breathing/snore + fart timing). No text, no captions, no logos, no watermark, no timestamps, no UI overlays. SCENE Cozy modern living room during daytime. Neutral walls, soft daylight from a window. Fabric couch/bed with visible weave texture and cushion compression under the dogs. Calm home ambience, background softly out of focus (minimal decor, not distracting). SUBJECTS (REFERENCE CONSISTENCY) Large black Rottweiler — lying on its side on the couch/bed, eyes closed, slow peaceful breathing, relaxed sleepy face, soft snore rhythm. Small tan Chihuahua — wearing a bright yellow shirt, sitting on top of the Rottweiler’s torso/back area, tiny size contrast, expressive face, mischievous but cute. Use the provided reference image to match exact look, fur, size, markings, and clothing. Keep both dogs consistent across the entire clip. EVENT The Chihuahua secretly releases a short, believable semi-transparent green vapor puff toward the sleeping Rottweiler’s face. The Rottweiler wakes slowly, confused and mildly offended, then gives a sleepy-angry deadpan with slight teeth showing (comedic offense only). Chihuahua pretends innocent, then gives a tiny smug side-eye to the camera for a perfect viral loop. ACTION (step-by-step, EXACTLY 10 seconds) 0.0–2.0s Close-up includes both dogs. Rottweiler asleep, slow chest rise/fall, soft snore. Chihuahua sits perfectly still with an innocent “angel” face. Slight handheld micro-shake, shallow DOF. 2.0–4.0s Chihuahua does a tiny “sneaky effort” wiggle: subtly scoots hips 1–2 cm, tail lifts slightly. Micro concentration face (small squint). Keep movement realistic and minimal. Autofocus breathes slightly. 4.0–5.0s A short semi-transparent green vapor puff appears (light mist, not thick smoke), drifting directly toward the Rottweiler’s nose and dissipating quickly. Add a soft comedic fart SFX audio only—still grounded like real recorded sound. 5.0–7.0s Rottweiler reacts subtly: nose wrinkles, whisker twitch, face twitches, slow blink. Eyes open gradually in sleepy confusion. Camera does a tiny natural handheld lean/push-in on the reaction. 7.0–8.8s Rottweiler slowly turns its head toward the Chihuahua and locks a deadpan stare. Expression shifts into a mild offended “sleepy-angry” grimace: upper lip lifts slightly, nose wrinkles deeper, a small amount of front teeth shows. No lunge, no bark, no snap—just comedic offense. 8.8–9.6s Chihuahua freezes in total “innocent statue mode,” looks away like “who, me?” Ears slightly back, tiny guilty swallow, but still cute and calm. No distress. 9.6–10.0s Chihuahua gives a tiny smug side-eye directly to the camera (eyes only, body still). Rottweiler holds the mild teeth-baring offended face in the background. Hard cut at exactly 10.0s for a seamless meme loop. STYLE / REALISM NOTES 100% photorealistic: real animals, realistic fur detail, natural lighting, real physics. Green cloud is subtle light vapor/mist, semi-transparent, believable, dissipates quickly. Viral pacing: micro handheld reactions, deadpan pause, loopable ending. CONSTRAINTS No violence, no harm, no distress. Reaction is mild comedic offense only. Teeth showing must be brief and mild (no snapping, lunging, chasing, contact, or biting). No CGI/cartoon look; realistic anatomy and textures. No on-screen text/subtitles/watermarks/timestamps/UI overlays. Keep both dogs consistent with the reference image across the whole clip.

video

FORMAT / QUALITY Ultra-realistic 4K handheld smartphone footage, vertical 9:16, 10 seconds, 30 fps. Daytime indoor li...

Sora 2video:sora2-text-to-video-standardTexto a video
21 feb 2026, 04:40

Ajustes

proveedor
Sora 2
Modelo
video:sora2-text-to-video-standard
Creada
21 feb 2026, 04:40
Créditos
70
Modo
Texto a video
Relación de aspecto
portrait

Crear algo similar

El generador a continuación preselecciona el modelo utilizado para este registro para que puedas remezclar la idea más rápido.

Las nuevas tareas aparecen aquí después de enviarlas.