프롬프트

Ultra-realistic POV GoPro underwater found footage, starting from the exact composition of the reference image. Camera is fixed to the line and orientation is locked: no rotation, no sideways drift, no zoom, only minor vibration transmitted by the taut line. The fishing line remains under constant tension for the entire 10 seconds — never slack, never floating. Hard constraint (must follow) The lure must move continuously toward the camera for the full 10 seconds, monotonic approach (it gets closer every moment). No pauses, no stops, no hovering at any time. Motion is driven by steady reeling with small realistic speed variations (micro-accelerations), but always forward. Motion + timing (10 seconds) 0–1s: Lure starts farther away, centered. It immediately begins approaching the camera. 1–2s: Continuous retrieve; lure wobbles naturally side-to-side; hooks swing. 2–3s: Lure advances closer; visible water drag and particle flow streaks toward the lens. 3–4s: Slight increase in retrieve speed (like a half-turn faster), still continuous forward motion. 4–5s: Predator fish silhouette appears behind the lure, tracking from the haze. 5–6s: First fast chase—predator accelerates from behind; lure keeps coming forward (no stop). 6–7s: Near-miss strike from behind; turbulence surges forward; lure jitters but continues moving toward camera. 7–8s: Retrieve continues steadily; predator repositions behind the lure, centered. 8–9s: Final aggressive approach; predator closes distance quickly while lure keeps advancing. 9–10s: Explosive strike timed with the moving lure; hard cut just before contact (no slow motion, no fade), perfect loop. Style: raw GoPro underwater physics, realistic motion blur only from speed, real visibility limits, natural colors, anti-AI realism, documentary fishing footage.

video

Ultra-realistic POV GoPro underwater found footage, starting from the exact composition of the reference image. Camer...

Sora 2video:sora2-pro-text-to-video텍스트를 영상으로
2026년 2월 2일 오전 12:21

설정

제공자
Sora 2
모델
video:sora2-pro-text-to-video
생성됨
2026년 2월 2일 오전 12:21
크레딧
675
모드
텍스트를 영상으로
화면 비율
portrait

비슷한 것을 만들다

아래 생성기는 이 레코드에 사용될 모델을 미리 선택하므로 아이디어를 더 빠르게 리믹스할 수 있습니다.

제출한 후 여기에 새 작업이 표시됩니다.