Skydiving to Work · An Illustrated Girl Drops from the Dawn Sky to Her Office Tower

Midjourney V8.2 + Seedance 2.52D AnimationAnimationUrbanCreativeCamera MovementSeedance 2.5Midjourney V8.230s
Midjourney sets the art style, then Seedance 2.5 takes over the animation: a girl with headphones leaps from a high perch, glides across a lavender-blue dawn skyline, opens her parachute, and lands squarely at the entrance of her office tower.
PROMPT · Video Prompt
Create with AI
Use @[ref image] as the direct reference for the girl's identity, face, hair, headphones, clothing, backpack, and the entire visual treatment. Preserve its hand-drawn 2D anime illustration, fine black ink contours, flat cel shadows, deep navy clothing, lavender-blue atmospheric city, and sparse peach dawn highlights. Extend the pictured skyline into a coherent city seen from new angles. The tall stepped tower on the right of the reference is her workplace and the destination throughout. An exhilarating morning commute: she stretches on a high perch, jumps, freefalls, opens her parachute, and steers across the waking city to work. Six cinematic shots, with a rhythm that moves from sleepy ease into vertigo, opens into a beautiful glide, tightens through a gust, and resolves in an ordinary arrival. Preserve her appearance and the drawn visual language through every angle. Her existing backpack houses the parachute; its harness is already secured around her torso and legs beneath the loose clothing. The departure perch stands well above the route and destination street, allowing a steadily descending journey. Shot 1. Begin with her seated on the high perch in the reference pose. She stands, spreads both arms out to her sides, stretches her shoulders and back, and yawns. The camera rises beside her and eases backward, revealing the enormous drop beyond the edge and the distant stepped tower across the city. Wind stirs her short black hair and loose sleeves. She finishes the stretch and looks toward the tower with the relaxed familiarity of someone heading to work. Shot 2. Cut to a low rear three-quarter tracking view as she commits to a clean outward leap. Show her shoes leaving the edge, with open air between her and the structure. The camera follows over the lip and pitches into the drop, exposing the immense street depth beneath her. The ledge races upward out of view as her descent accelerates; the quiet dawn ambience gives way to rushing air. Shot 3. Cut to a lateral falling tracking shot, keeping her face, streaming clothing, and the rising building facades readable. Let her fall freely for a distinct beat before she deploys the parachute from her backpack. Within this same shot, the camera widens to show the orange-and-cream rectangular canopy unfurl above her, fill with air, and pull the suspension lines taut. The harness takes her weight, her descent visibly slows, and she swings into an upright hanging position beneath the fully open canopy. A sharp fabric snap punctuates the change. She takes the steering toggles, focused and composed. Shot 4. Cut to a spacious aerial wide shot. She guides the open canopy over layered rooftops, soft morning haze, and streets catching peach light. The camera sweeps in a broad arc beside her flight path, letting nearer rooflines slide across the distant skyline to reveal the city's depth. Give this view room to breathe. Her small steering inputs produce visible gentle turns toward the stepped tower, which steadily grows ahead as she loses altitude. The canopy remains inflated and attached, her weight suspended beneath it throughout the journey. Shot 5. Cut closer on the approach to the same tower. A sudden crosswind ripples and tilts the canopy, drifting her away from the broad street beside its entrance. Her relaxed expression sharpens. She responds through the steering controls and banks into a clear, sweeping correction around the tower's outer corner. The camera tracks from outside the turn, keeping the tilted canopy, her suspended body, and the rapidly sliding facade in a readable relationship. Maintain open clearance for the entire canopy. The building's scale and the sudden sideways displacement supply the danger; her controlled response restores the route. As she levels out, the open landing area at the tower's base appears ahead and below. Shot 6. Cut to a low street-level view beside the tower's office entrance, looking up to receive her descending approach. The camera tilts down with her as she slows into a soft landing on the wide, clear pavement. Her shoes visibly meet the ground; a few quick balancing steps carry her forward as the canopy settles and collapses behind her. She gathers the loose fabric over one arm and heads toward the office doors among arriving workers, her sleepy morning ease returning. Finish with her approaching the entrance, the same tower rising above her and the parachute fabric still visibly carried. Sound follows the action: high-altitude wind, the accelerating rush of freefall, canopy opening and fabric flutter, then footsteps and the waking street. No dialogue, captions, or title cards. No background music or score.
✍️ Editor’s Notes

This prompt first locks the art style and character in one move with @[ref image] — line art, cel shading, palette, headphones, backpack — narrowing the variables down to just camera movement. The narrative is written as six Shots forming a clear emotional arc: sleepy stretch → committed leap → freefall into parachute deployment → open glide → crosswind correction → ordinary landing. Each Shot states the camera position first, then describes the action — the changing angles themselves build the rhythm. Plausibility details are seeded in too: the parachute is hidden in a backpack she already has, and the harness is stated as 'already secured,' so gear doesn't sprout out of nowhere mid-shot. The sound keeps only wind, fabric flutter and footsteps, explicitly stating 'no score' — grounding realism in ambient sound.

Create with AI
Image Prompt
PROMPT · Full Prompt
Create with AI
dawn --ar 16:9 --raw --sref 2116795492 452368747 --stylize 500
✍️ Editor’s Notes

This is a classic Midjourney mood-lock prompt: the body text is just one word, 'dawn,' leaving style, composition and detail entirely to the two --sref reference images, while the prompt itself only pins down the time of day (dawn) and the frame ratio/style parameters (--ar 16:9 --raw --stylize 500). This 'minimal body text + parameters doing the work' pattern is common for follow-up generations once a style reference is already locked in: there's no need to re-describe the character or scene (other prompt blocks already handle that) — a single keyword is enough to adjust this particular image's lighting mood. Worth copying: once --sref has locked the look, the body text only needs to carry the one variable that actually changes — here, the time of day — and leave everything else to the parameters and reference images.

Create with AI
Related Works
׋›
↑