AI Video Prompts for Landscape Video
Seedance 2.0 prompts for landscape cinematography — wide establishing shots, layered atmospheric perspective, foreground-to-horizon depth staging, camera reveal movements, and the specific time-of-day and weather conditions that transform a landscape from a background into a subject. Each technique names the physical and geometric conditions that define cinematic landscape footage.
Landscape cinematography is one of the most photographed subjects in the world and one of the most difficult to make visually distinctive in AI video. The challenge is specificity: "beautiful mountain landscape" or "scenic nature scene" describes what the viewer sees without instructing how the camera sees it — angle, altitude, depth layers, light direction, atmospheric condition, and movement. The difference between a generic landscape background and a directed landscape shot is entirely in the geometry: where the camera is relative to the terrain, what depth layers exist between foreground and horizon, what the light is doing at the specific moment of capture, and whether the camera moves or holds. Camera altitude is the first and most consequential decision in landscape cinematography, and it determines the entire visual grammar of the shot. Low camera angles — at or near ground level, 30–50 cm above the terrain — produce the most dramatic foreground-to-horizon relationships: a field of wildflowers, a frost-covered rock, a cracked desert floor becomes enormous in the foreground while a mountain range or coastline dominates the background, compressing the middle ground between them. The vocabulary: "camera positioned 30 cm above the ground, wide 16mm equivalent lens, foreground of frost-covered granite rock sharp and detailed, mid-ground of pine forest in shadow, distant snow-capped peaks at the horizon fully in focus, all three planes simultaneously sharp via deep focus." This three-layer foreground/middle-ground/horizon structure is the fundamental visual grammar of cinematic landscape photography. Drone altitude — 300 to 1,000 meters — produces the opposite grammar: the landscape reads as a map, terrain becomes abstract pattern and color, rivers and roads become lines, and individual elements (trees, buildings) become texture rather than subjects. Both are valid, but they require completely different vocabulary to produce. Atmospheric perspective is the physical phenomenon that makes deep landscapes legible as three-dimensional space on a two-dimensional screen: distant objects scatter blue light from the atmosphere between them and the camera, causing far-away terrain to appear lighter, bluer, and lower in contrast than near terrain. This depth haze is the automatic depth cue in outdoor landscapes, and it scales with distance — at 1 km the effect is subtle, at 50 km it renders mountain ranges as silhouettes of graded blue. "Five layers of mountain ridges visible from foreground to horizon, each ridge progressively lighter and more blue-gray than the one in front of it — the nearest ridge dark green with visible tree texture, the second ridge medium gray-green with compressed detail, the third ridge pale gray, the fourth pale blue-gray almost matching the sky, the furthest ridge barely darker than the sky behind it." This vocabulary — naming each atmospheric layer with a specific tone and detail level — gives Seedance a contract for spatial depth rather than a backdrop. Without naming the atmospheric layers, landscape prompts typically produce visually flat results where all terrain reads at the same contrast level regardless of distance, losing the sense of three-dimensional space. The landscape reveal is the most cinematically powerful camera move in landscape shooting: the moment when a wide, extraordinary scene is exposed to the viewer for the first time. Reveal structures include the crane-up (camera starts at ground level on an obscuring foreground element — a hillcrest, a ridgeline, a boulder — and rises to expose the landscape beyond), the pull-back (starts tight on a foreground detail and retreats to reveal the surrounding landscape), and the pan-to (starts on a neutral sky or rock face and sweeps to expose the landscape at the end of the move). The crane-up is the most powerful and requires: "camera begins at ground level behind the ridgeline, the ridge completely filling the upper frame, a slow crane upward over 8 seconds reveals an enormous valley 500 meters below, camera continues rising until the full 20 km of valley is visible, ending with a wide view of the entire mountain range in the background." The critical element is specifying what is hidden before the reveal and what is exposed during it — the anticipation-exposure structure is what makes the reveal a narrative event rather than just a camera move. Weather and time-of-day vocabulary transforms a landscape from a backdrop into an active narrative element. Each atmospheric condition produces a completely different visual grammar: overcast diffuse light (no shadows, flat contrast, colors saturated and accurate, all surfaces equally lit — useful for emphasizing color and pattern over form), dramatic storm light (one band of direct sun breaking through storm clouds onto a narrow section of landscape while the surrounding terrain is in dark shadow — the most dramatic landscape condition, requires naming the sunlit band and the dark surroundings separately), golden-hour raking light (long horizontal shadows from all terrain features, warm 3,000K directional light from low angle striking only the tops of ridges and trees, valley floors remaining in shadow), and post-storm clearing (rain-soaked surfaces reflecting sky light, mist lifting from valley floors in wisps, cloud shadows moving rapidly across the landscape creating shifting light patterns). The post-storm vocabulary in particular: "mist lifting from the valley floor in irregular columns as the sun returns after rain, dark storm clouds retreating to the right while clear blue sky advances from the left, wet foliage glistening with residual moisture, the valley floor alternating between shadow and direct light as the clearing clouds move."
No prompts here yet — browse the full gallery.
More use cases
Frequently asked questions
What are the best AI video prompts for landscape?
The best Seedance landscape prompts specify four elements beyond the subject: (1) camera altitude and angle — "camera 30 cm above ground, wide 16mm lens" or "drone at 400 meters, looking straight down" — because altitude determines the entire visual grammar; (2) depth layers — name at least three planes between foreground and horizon with distinct tone and detail levels; (3) atmospheric condition — overcast diffuse, direct sun with long shadows, post-storm clearing, or golden-hour raking light, each requiring different vocabulary; (4) camera motion or hold — a slow crane-up reveal, a lateral pan across terrain, or a locked static wide. Combining these four elements produces a directed landscape shot rather than a generic scenic backdrop.
How do I write a Seedance prompt for atmospheric depth in a landscape?
Atmospheric perspective — the blue haze that makes distant terrain lighter and less saturated than near terrain — is the primary depth cue in wide landscape shots. To activate it: name each depth layer explicitly with its tonal value and detail level. "Foreground of dark green grassland with visible individual blade texture, first mid-ground ridge in medium gray-green with compressed tree texture, second ridge in pale gray with no visible detail, third ridge in pale blue-gray matching the lower sky, distant mountain peaks as pale blue silhouettes barely distinguishable from the sky." Five named layers are enough to build convincing three-dimensional space. Without naming the tonal progression, landscape prompts typically render all terrain at the same contrast level, producing flat results with no sense of depth.
How do I prompt a dramatic landscape reveal in Seedance 2.0?
A landscape reveal needs three elements: what is hidden, the camera move that exposes it, and what is revealed. For a crane-up reveal: "camera begins at ground level on the near side of a ridgeline, the ridge completely filling the upper half of frame, a slow crane upward over 6 seconds rises above the ridge and exposes the landscape on the far side — a wide river valley 300 meters below, flanked by forested slopes, ending with the full valley visible and the distant mountain range at the horizon." For a pull-back reveal: "camera starts tight on a single wildflower in sharp focus, slow backward dolly over 5 seconds pulls away to reveal the surrounding meadow, continuing backward until the meadow fills the frame and the mountain range appears at the horizon." Specify the duration of the move (in seconds), what the camera starts on, and exactly what is exposed and in what order as the camera moves.