AI Video Prompts for Fog & Mist
Seedance 2.0 prompts for fog and mist — volumetric god rays, depth-gradient landscapes, horror ground fog, morning valley mist, and the atmospheric diffusion techniques that give AI-generated video its cinematic depth.
Fog is one of the most technically powerful atmospheric elements in AI video prompting, but it is almost universally underspecified. Most prompts that attempt a foggy scene simply say "foggy morning" or "misty atmosphere" and get a generic haze without depth, directionality, or light interaction. The difference between a flat foggy scene and a cinematically atmospheric one is always in the specifics: how dense the fog is at different distances from the camera, how it interacts with available light sources, and what physical system generates it (ground fog that clings to the earth, high-altitude valley fog, sea fog rolling in from water, interior architectural fog from low-lying surfaces). Fog is not a uniform grey filter — it is a depth-stratified medium that interacts differently with every light source in the scene, and the most effective Seedance fog prompts treat it as a three-dimensional object rather than a background texture. The depth gradient is the core structural fact of fog as a cinematography element. Objects at different distances from the camera are rendered with different levels of atmospheric haze: subjects in the near field (0–5 metres) are sharp, with full colour saturation; mid-ground subjects (10–30 metres) are partially obscured, with reduced saturation and softened edges; far-field subjects (50+ metres) dissolve into the fog entirely, readable only as tonal shapes rather than distinct forms. Naming this gradient explicitly — "near-field subject is sharp and fully saturated, mid-ground trees are partially dissolved in blue-grey haze, far-field buildings visible only as darker grey shapes against the fog" — gives Seedance the depth map that produces a three-dimensional atmospheric scene rather than a flat filter applied uniformly. The gradient also shifts with fog density: thin morning mist shows the full gradient compressed to 10–50 metres; heavy storm fog compresses it to 5–20 metres. Volumetric god rays through fog are the highest-leverage single technique in fog cinematography. When a strong directional light source (the sun at low angle, a spotlight, car headlights, a window in a dark interior) shines through fog or mist, the particles scatter the light sideways, making the beam itself visible as a solid shaft of light through the air. This is the phenomenon that produces the dramatic light-shaft images in forest fog, morning mist over water, and dusty interior spaces. Prompting god rays requires naming three elements: the light source (sun at 15-degree angle, single spotlight, car headlights), the fog or particle medium through which the rays travel (morning mist, dusty fog, dense sea fog), and the geometry of the rays (near-horizontal shafts breaking through gaps in the tree canopy, vertical beams falling from a high window, forward-directed light cones from car headlights). Without all three named, AI video defaults to a bright scene with atmospheric haze rather than visible light shafts. Fog as a diffuse light source is the second major cinematographic function of fog. When fog is dense enough, it acts as an enormous omnidirectional diffuser, softening every hard light source in the scene into a soft, directionless wrap. Morning fog converts the harsh sun into a large, soft key light with no discernible direction — all shadows disappear, all surfaces receive equal illumination, skin is rendered at its most flattering, and the overall palette shifts to neutral grey-white-blue. This is the quality used in romantic outdoor scenes, dream sequences, and landscapes that convey peace or meditation. To prompt this quality, name the fog's diffusion function explicitly: "dense fog acting as a diffuse filter on the early morning sun, eliminating all hard shadows, all surfaces rendered in soft neutral grey-blue-white light, subject's skin lit from all directions without a discernible key." This is different from "foggy morning," which can produce either a hazy scene with a still-visible sun or a flat grey filter. Ground fog — fog that sits in a layer near the earth's surface while the upper atmosphere is clear — is the most cinematically distinctive fog type and requires specific vocabulary to produce. Ground fog is generated in real environments when cold air drains into low-lying areas (valleys, depressions, river banks, fields) and the cold air stabilises a fog layer at 0–2 metres height while the sky above remains clear or has light cloud. The visual result is that figures walking through the fog are visible from the waist up, with their feet and lower legs dissolved into the white layer, while the sky above is deep blue or shows stars in the early morning. In horror and dark fantasy, ground fog is the defining atmospheric element: a single figure walking through knee-height ground fog with only a foggy layer at the base has become a genre archetype. Prompting this: "ground-level fog layer sitting at knee height, 1.5 metres thick, subject visible from waist up above the fog, sky clear and blue above, fog layer illuminated from below by moonlight giving it a faint blue-grey luminescence." Fog in different genre registers produces different visual and emotional effects, and the prompting vocabulary differs by register. Morning landscape fog (romance, drama, meditation): soft diffuse grey-white light, colour palette shifted to cool neutral, fog layer visible in valley with mountain peaks emerging above, dew on surfaces, slow camera movement matching the stillness. Horror fog: cold blue-grey ground-level fog, hard practical light sources (single lamp, car headlight, torch) cutting through the fog and visible as light cones, fog movement made visible (rolling, drifting slowly), subject partially obscured below the fog line. Sea fog and coastal fog: white-grey fog rolling in from the ocean, fog erasing the horizon line so sea and sky merge, foghorn audio cue, lighthouse light visible as a rotating beam through the fog. Industrial fog and steam: fog rising from ground-level sources (river, factory grate, cooking surface), hot air rising visible as thermal shimmer at the fog edge, steam and fog distinguished by their opacity and rise pattern.
No prompts here yet — browse the full gallery.
More use cases
Frequently asked questions
What are the best AI video prompts for fog and mist?
The best Seedance fog prompts specify fog as a three-part depth system: (1) the depth gradient — "near-field subject sharp and saturated, mid-ground partially dissolved, far-field visible only as grey tonal shapes"; (2) the light interaction — name the light source and how fog interacts with it ("morning sun at low angle creating visible god-ray shafts through the forest fog", "ground fog illuminated from below by moonlight"); (3) the fog type and height — "ground-level fog layer sitting at knee height" vs "dense valley fog filling the entire frame" vs "sea fog rolling in from the right, erasing the horizon." These three layers together produce atmospheric depth rather than a flat grey filter.
How do I write a Seedance 2.0 prompt for fog or mist effects?
A fog prompt has four layers. First, name the fog's physical type and height: ground fog (clings to the earth, knee-height), valley fog (fills the low points while peaks emerge above), sea fog (rolling horizontal), interior mist (steam rising from a surface). Second, describe the depth gradient: how sharp is the near field versus the far field? Third, name your light source and the fog's interaction with it — god rays through directional light, diffuse soft wrap from dense overcast fog, or colour-temperature shift (blue-grey fog under moonlight, warm orange-yellow fog under sodium streetlights). Fourth, optionally add fog movement: "fog drifting slowly from right to left" or "ground fog rolling forward toward camera." These four layers convert a mood description ("misty morning") into a directed atmospheric specification.
Can Seedance 2.0 generate horror fog, morning mist, and sea fog differently?
Yes, and each requires distinct vocabulary. For horror fog: name the fog type ("ground-level cold fog, 1.5 metres thick"), the light source cutting through it ("single practical lamp casting a hard cone through the fog layer, making the fog visible as scattered light particles"), and the subject's relationship to the fog height ("figure visible from chest up above the fog, lower body dissolved"). For morning mist: name the diffusion quality ("dense morning fog eliminating all hard shadows, sun visible only as a bright white disc through the mist, no discernible light direction, all surfaces in soft grey-white wrap light"), the colour palette shift ("blue-grey-white tonal range, all colour saturation reduced by 60%"), and the depth gradient ("near field sharp, mid-ground partially dissolved, far-field trees visible only as darker grey shapes"). For sea fog: name the fog origin ("fog rolling in from the ocean on the right, moving toward the coast"), the horizon erasure ("sea and sky merged into a single grey-white plane"), and any remaining landmarks ("lighthouse at mid-ground visible only as a flashing beam through the fog, no structural detail").