Japanese Water Obstacle Challenge Show
This prompt is valuable because it treats a multi-camera sports show as one continuous physical course. Camera changes do not reset the contestant or the ven…
The idea / HolyCrab
Japanese Water Obstacle Challenge Show
A comprehensive prompt for generating a realistic Japanese-style water obstacle competition, including multi-camera angles and specific time-axis event instructions.

Source prompt
Generate a 30-second, 16:9 landscape, highly realistic live-action large-scale water obstacle competition program. The overall presentation has the live broadcast feel of Japanese large-scale sports entertainment special programs from the late 2000s to 2010s.
The stage is a huge outdoor water stadium at night, with the complete course built over a large competition pool, surrounded by large lighting towers, steel trusses, spectator stands, staff areas, and real TV production equipment. Multi-camera filming for real sports programs: ENG cameras, jib cameras, track tracking, telephoto, low-angle water surface, and fixed camera positions on the side of obstacles. The camera always serves the character's progress.
The screen maintains the feel of real TV broadcasts: natural Pan, Zoom, follow focus, character tracking, slight ENG camera movement, the slightly soft picture quality of early HDTV, natural TV compression, and real light and dark contrast formed by large live lighting. The entire process follows natural real speed and real-time physical laws. [Mandatory Character Consistency] The uploaded image is the only visual reference for the challenger.
Strictly maintain the character's identity, facial features, face shape, hairstyle, hair color, skin color, sense of age, height, body type, proportions, clothing, and overall temperament. From the first frame to the last frame of the video, it always remains the same character.
Across multi-camera switching, front, side, back, low-angle, telephoto, full-body action, running, jumping, falling, climbing, falling into water, and wet state, the character's identity, face, body proportions, and clothing remain consistent. The clothing remains the same outfit from the uploaded character image throughout, with fixed color, structure, material, and wearing style. [Site Continuity] The entire video takes place in the same continuous nighttime water arena.
Three obstacles are connected in physical space order: rolling drum obstacle -> Fishbone Reverse -> curved high wall. The character must move forward along the same track. Each camera cut continues the character's position and progress from the previous shot. The positional relationship of obstacles, pools, platforms, spectator seats, and large lighting remains continuous and stable.
HolyCrab
Editor's note
This prompt is valuable because it treats a multi-camera sports show as one continuous physical course. Camera changes do not reset the contestant or the venue; every cut must inherit the current obstacle, position, wetness, and progress from the previous shot.
Prompt breakdown
Identity lock
The uploaded contestant image is the sole identity reference across front, side, rear, telephoto, full-body, falling, climbing, and wet states.
Course graph
The rotating drum, reverse fishbone, and curved wall are connected in one direction. The contestant must keep moving forward through that fixed geography.
Broadcast grammar
ENG, jib, tracking, telephoto, water-level, and fixed obstacle cameras are assigned to readable sports coverage rather than decorative movement.
State continuity
Position, progress, clothing, water state, audience stands, lighting towers, and obstacle placement all carry across camera cuts.
How to use this prompt
- State that the character image defines identity and clothing only; the competition venue comes from a separate environment reference.
- Describe an observable end position for each obstacle before moving to the next one.
- Keep the course direction fixed and avoid asking every camera to move at the same time.
Seedance workflow
Seedance 2.5 Prompt Optimizer Skill
Download the Seedance-specific Skill for Codex to structure references, timing, camera movement, and audio more reliably.MiniMax H3