Video AI Cinematic ShotsAdvanced180 minSaves 2+ hours

Crafting a Surreal Underwater Dream Sequence for Documentaries

Filmmakers can generate ethereal underwater dream sequences, complete with specific camera, lighting, and sound cues, to enhance documentary narrative depth.

Generate a cinematic underwater dream sequence for documentaries. Specify slow-motion figures, light shafts, and muted sound design. This prompt details shot list, camera movement, lenses, lighting, wardrobe, and score, empowering AI filmmakers to craft surreal, character-driven moments with a distinct filmic look.

Veo Kling Runway

READY-TO-USE PROMPT

Copy Prompt

prompt.txt

As a Film Director specializing in surreal, atmospheric sequences, your objective is to translate a conceptual underwater dream sequence into a detailed, actionable prompt for advanced video AI models.

### Context

I am developing a dream sequence for a documentary. The goal is to evoke a profound sense of suspension, introspection, and otherworldly beauty, rather than fear or danger. The visual narrative should feel like a fragmented memory or an internal psychological state, emphasizing the character's journey rather than external events. The aesthetic should be distinctly cinematic, with a focus on mood and texture.

### Task

Generate a 24fps video sequence, approximately 15-20 seconds in length, depicting a slow-motion human figure suspended underwater. The sequence must convey a dreamlike, ethereal quality, making explicit choices for camera, lighting, wardrobe, and sound design to achieve this effect.

### Constraints

* **Film Look**: Integrate elements like subtle film grain, slight chromatic aberration, and a shallow depth of field to achieve an authentic filmic aesthetic.
* **Color Palette**: The dominant colors should be cool blues, greens, and muted grays, counterpointed by warm, golden shafts of light penetrating the water from above.
* **Subject**: A single human figure ({{figure_description}}) in simple, flowing wardrobe ({{wardrobe_description}}). The figure's gender and age are less critical than their form and the passive nature of their suspension.
* **Movement**: The figure floats gently, with minimal deliberate movement; emphasize passive suspension and drifting. Camera movement should be slow, deliberate, and exceptionally smooth, mirroring the underwater environment.
* **Lighting**: Key lighting should consist of strong, defined shafts of light cutting through the water from an unseen surface above. Ambient light should be diffused, creating soft shadows and highlights.
* **Duration**: The final sequence should be between 15 and 20 seconds.
* **Frame Rate**: The output must be specified at 24 frames per second.

### Output Format

Provide a detailed cinematic scene prompt, including the following sections:

1. **Shot List**: Describe 3-4 distinct shots, specifying camera distance and angle (e.g., wide shot, medium close-up, low angle).
2. **Camera Movement**: Detail specific pans, tilts, dollies, or static shots for each segment.
3. **Lens Choice**: Suggest a focal length and aperture setting that supports the desired film look and depth of field (e.g., 35mm f/2.8, 85mm f/1.8).
4. **Lighting Scheme**: Elaborate on the light sources, their direction, and their effect on the scene and subject.
5. **Wardrobe**: Provide a detailed description of the figure's attire, emphasizing its texture and how it interacts with the water.
6. **Score Cue**: A brief description of the intended musical or ambient score, focusing on mood and instrumentation.
7. **FPS Spec**: Explicitly state the frame rate.

Estimated results

DifficultyAdvanced

Setup time180 min

Time saved2+ hours

Best modelsVeo, Kling, Runway

Best audienceFilm Production, Documentary Filmmaking

Editor's note

Why this prompt matters

Crafting abstract visual sequences that convey deep emotional or psychological states within a documentary often presents a challenge. Traditional filmmaking can find such scenes costly and time-consuming to realize, requiring specialized underwater crews, complex rigging, and extensive post-production. This workflow is designed for documentary filmmakers and visual artists who need to articulate highly specific, atmospheric, and surreal moments without the logistical hurdles of a live shoot.

This method allows for the rapid ideation and visualization of a nuanced underwater dream sequence, an element often critical for exploring a character's inner world or metaphorical journey. Instead of broad strokes, it focuses on granular control over every cinematic detail, from camera movement to the subtle interplay of light and shadow. Filmmakers should turn to this workflow when their narrative calls for a distinct, non-literal visual passage that requires precise aesthetic and technical specifications, ensuring the generated output aligns perfectly with their artistic vision for an internal, reflective segment. It enables the creation of evocative, mood-driven scenes that deepen a documentary's emotional resonance.

Anatomy

Prompt engineering breakdown

Role

Film Director specializing in surreal, atmospheric sequences

Context

Developing a dream sequence for a documentary, aiming to evoke suspension, introspection, and otherworldly beauty, focusing on a character's internal journey rather than external events.

Goal

Translate a conceptual underwater dream sequence into a detailed, actionable prompt for advanced video AI models, specifying all cinematic elements.

Constraints

Achieve a distinct film look (grain, chromatic aberration, shallow DOF); use a cool color palette with warm light shafts; feature a single human figure ({{figure_description}}) in simple, flowing wardrobe ({{wardrobe_description}}); emphasize passive, slow movement; employ strong, defined light shafts from above with diffused ambient light; duration 15-20 seconds; output at 24 frames per second.

Output format

A detailed cinematic scene prompt including Shot List, Camera Movement, Lens Choice, Lighting Scheme, Wardrobe, Score Cue, and FPS Spec.

Why this structure works

The prompt utilizes role priming by assigning the model the persona of a Film Director, which encourages a more creative and technically informed response. Explicit constraints on visual aesthetics, movement, and duration guide the AI toward the desired surreal, cinematic output. The structured output format ensures all necessary filmmaking elements are addressed, providing a comprehensive and usable input for video generation.

Pick your version

Prompt variations

BeginnerWorks with any model

For users new to AI video generation who need simpler guidance and fewer technical details to get started.

prompt.txt

Imagine you are guiding a visual artist. Create a simple plan for an underwater dream scene for a documentary. Show a person ({{figure_type}}) floating slowly in blue water, lit by sunbeams. We want it to feel calm and thoughtful, not scary. The scene should be about 15-20 seconds long, at 24 frames per second. Describe 3-4 shots, what the camera does (move slowly or stay still), what the person wears ({{simple_wardrobe}}), how the light looks (sun rays from above), and what kind of soft, quiet music should play. Focus on cool colors like blue and green, with warm light.

ProfessionalBest with veo

When a detailed, technically specific prompt is needed for advanced video AI models, mirroring a traditional film production brief.

prompt.txt

As a seasoned Film Director, develop a comprehensive cinematic prompt for an underwater dream sequence. This sequence, for a documentary, aims to convey deep introspection and a serene, otherworldly suspension, avoiding any sense of threat. The visual language should be distinctly cinematic, emphasizing mood, texture, and the character's internal state. Generate a 15-20 second, 24fps video prompt featuring a slow-motion human figure ({{figure_description}}) suspended underwater. Specify 3-4 distinct shots with camera angles and distances (e.g., wide, medium close-up), precise camera movements (slow dolly, subtle pan), a preferred lens (e.g., 50mm f/1.4 for shallow DOF), a detailed lighting scheme (shafts of golden light, diffused ambient blue), and specific wardrobe ({{wardrobe_description}}). Conclude with a descriptive score cue for ambient, ethereal sound design. The aesthetic demands subtle film grain, cool blue/green palette, and clear articulation of passive figure movement.

Short VersionBest with runway

For quick ideation or when interacting with models that perform better with concise, single-paragraph instructions.

prompt.txt

Create a concise 15-20 second, 24fps cinematic prompt for an underwater dream sequence. A single human figure ({{figure_details}}) floats in slow motion amidst cool blue water, illuminated by warm light shafts from above. The goal is to evoke serene introspection. Describe 3-4 shots, camera movement, lens choice, lighting, simple wardrobe ({{wardrobe_style}}), and a soft score cue, ensuring a filmic, ethereal mood with subtle grain. Focus on passive suspension and deliberate, smooth camera work to capture a fragmented memory aesthetic.

EnterpriseBest with kling

In large-scale productions where compliance, brand guidelines, and stakeholder review are critical, requiring explicit risk mitigation and approval pathways.

prompt.txt

As a Lead Creative Director for a major production studio, develop an advanced, compliance-reviewed cinematic prompt for a critical underwater dream sequence within a documentary project. This sequence must achieve a profound sense of serene suspension and introspection, aligning with our brand's narrative integrity and ethical guidelines. The goal is a 15-20 second, 24fps output depicting a slow-motion human figure ({{figure_identity}}) in an ethereal underwater environment. Ensure all visual and auditory elements (shot list, camera movement, specific lens choices like anamorphic or vintage prime, detailed multi-point lighting plan, ethically sourced wardrobe materials {{wardrobe_details}}, and a precise score cue) not only achieve the surreal, filmic aesthetic but also adhere to content safety protocols and avoid any potential misinterpretations regarding vulnerability or distress. The prompt must allow for clear stakeholder review and iteration cycles, emphasizing adherence to a pre-approved visual style guide, including subtle filmic aberrations and a cool/warm color interplay. All generated content must be reviewed for brand safety and intellectual property considerations before finalization.

What you'll get

Expected output

Here is a detailed cinematic scene prompt generated for an underwater dream sequence, focusing on a young woman in a flowing white dress:

Shot List:

Wide Shot: A lone human figure, a young woman with long dark hair, suspended motionless in a vast, deep blue underwater expanse. Her form is small against the expansive, muted environment. Shafts of golden light pierce the water from high above, creating shimmering columns.
Medium Shot: The figure drifts slowly, almost imperceptibly, towards the camera. Her face is serene, eyes closed, hair gently undulating around her. The focus is on her upper body and the subtle movement of her dress. The shafts of light play across her form, highlighting textures.
Close-Up: Focus on her outstretched hand, fingers slightly curled, catching the warm light. Tiny air bubbles rise past, distorted by the water. The texture of her linen dress is visible, flowing softly. The background blurs into an ethereal blue-green.
Tracking Shot: The camera slowly tracks alongside the suspended figure as she glides horizontally, revealing the graceful sweep of her dress and the gentle sway of her body. The strong overhead light creates a halo effect around her, contrasting with the deeper blue-green shadows below.

Camera Movement:

Shot 1: Extremely slow, almost imperceptible crane shot, descending gently as the figure comes into view, emphasizing her isolation.
Shot 2: Gradual, smooth push-in (dolly shot) towards the figure, maintaining a steady, dreamlike pace.
Shot 3: Static shot, holding the close-up, allowing the subtle movements of water and light to define the frame.
Shot 4: Ultra-slow side tracking shot, perfectly level with the figure, mimicking her gentle drift. The movement is fluid and uninterrupted.

Lens Choice:

For the wide and medium shots: 35mm f/2.8, to capture the expansive environment while maintaining a shallow depth of field on the figure, creating a slightly compressed, immersive feel with gentle bokeh.
For the close-up and tracking shots: 85mm f/1.8, to emphasize isolation and create a very shallow depth of field, blurring the background into an abstract wash of color and highlighting the figure's details with a soft, ethereal quality.

Lighting Scheme: The primary light source emulates natural sunlight piercing the water surface from directly above, creating distinct, warm, golden shafts that cut through the cool blue-green ambient light. These shafts should illuminate the figure selectively, creating high contrast on her form and the fabric of her dress, while leaving other areas in diffused, soft shadow. The overall ambient light is a deep, diffused blue-green, providing a subtle, almost phosphorescent glow from the distant water, ensuring no harsh black areas. The interaction of warm and cool tones is key.

Wardrobe: A simple, long-sleeved white linen dress. The fabric is lightweight and semi-transparent, designed to billow and flow effortlessly with the slightest water movement, creating soft, organic shapes around the figure. The texture of the linen should be subtly visible, enhanced by the shafts of light. The dress is unadorned, emphasizing purity and suspension, its folds catching and diffusing light.

Score Cue: A minimalist, ambient score featuring sustained, low-frequency synth pads mixed with sparse, reverberating piano notes and occasional, subtle hydrophone-like textures. The mood is contemplative, melancholic, and deeply serene. No distinct melody, rather a wash of sound that evokes vastness and introspection.

FPS Spec: 24 frames per second.

Under the hood

Why this prompt works

This prompt structure succeeds by employing several deliberate prompt engineering techniques. Firstly, role priming establishes the AI as a "Film Director specializing in surreal, atmospheric sequences." This immediately sets a high bar for creative interpretation and ensures the output is framed from an expert perspective, guiding the model to think in terms of cinematic conventions rather than generic image generation.

Secondly, the inclusion of detailed context regarding the sequence's purpose – to evoke suspension, introspection, and an otherworldly beauty – provides the AI with the essential "why" behind the request. This context is crucial for maintaining thematic consistency across all generated elements.

Thirdly, explicit constraints are fundamental to preventing output drift. Specifications for "Film Look," "Color Palette," "Subject," "Movement," "Lighting," "Duration," and "Frame Rate" act as guardrails, ensuring the AI adheres to the precise aesthetic and technical requirements. This level of detail minimizes ambiguity and focuses the model's creative space.

Finally, the demand for structured output with specific sections like "Shot List," "Camera Movement," and "Lens Choice" forces the AI to break down the complex creative task into manageable, interconnected components. This ensures a comprehensive and actionable response, detailing every aspect required to generate a coherent, high-quality cinematic sequence, far surpassing the capabilities of a single, less defined request.

Model fit

Best AI models for this prompt

Veo

Veo excels at generating highly detailed and coherent video sequences, making it suitable for complex scene descriptions like this dream sequence. Its ability to maintain visual consistency across longer clips and render subtle lighting effects will be crucial for the ethereal quality required. Limitations include occasional difficulty with highly specific character facial expressions or very rapid, complex action. See the full Veo hub for deeper guidance.

Kling

Kling is adept at producing stylized and aesthetically rich video, which aligns well with the "film-look" and atmospheric requirements of this prompt. It handles abstract concepts and mood-driven scenes effectively, offering strong control over color grading and visual texture. Users might find it less precise with intricate anatomical details or highly dynamic, fast-paced camera work. See the full Kling hub for deeper guidance.

Runway

Runway's strengths lie in its creative versatility and capacity for generating surreal and artistic video content, making it a strong contender for dream sequences. It can interpret descriptive language into visually compelling scenes, particularly when atmospheric and abstract elements are prioritized. Challenges may arise in achieving absolute photorealism or maintaining extremely tight continuity across extended, complex narratives. See the full Runway hub for deeper guidance.

When to use

When integrating abstract, emotionally resonant moments into a documentary narrative.
For creating non-linear segments that represent internal psychological states or fragmented memories.
To establish a profound sense of introspection or otherworldliness without requiring literal historical footage.
When existing archival footage cannot convey the specific surreal, atmospheric quality needed.
As a mood-setting interstitial or transition between more factual documentary segments.

When not to use

If your documentary requires strictly factual, verité, or observational footage.
For sequences demanding rapid action, clear narrative progression, or explicit plot points.
When the primary goal is to convey danger, fear, or a sense of urgency.
If your production pipeline cannot accommodate iterative generation and fine-tuning with advanced AI video models.
For projects where the aesthetic must strictly adhere to a raw, unmanipulated style.

Get more from it

Pro tips

1
Refine figure_description with emotional cues, like 'figure with a sense of peace' to guide AI's subtle facial expressions and posture. Prevents generic or misaligned character portrayal.
2
Experiment with wardrobe_description materials beyond 'flowing,' such as silk or linen, for varied underwater fabric dynamics and visual texture. Prevents flat or unrealistic fabric movement.
3
Specify exact color hex codes or detailed color names for blues, greens, and golds to ensure precise palette adherence. Prevents color shifts or unintended tonal variations.
4
Describe ambient sound elements like 'distant muffled thrum' or 'subtle water distortion' to complement the score cue effectively. Prevents a disconnected soundscape.
5
Iterate on lens and aperture combinations. A wider aperture (e.g., f/1.4) enhances bokeh, crucial for that dreamlike shallow depth of field. Prevents a flat, documentary-style focus.
6
Test various camera speeds and focus racks within the 'slow, deliberate' movement to find the most ethereal rhythm. Prevents static or unnatural visual flow.
7
Detail the quality of light shafts—'diffused edges,' 'sharp beams'—to guide the AI in rendering atmospheric particulate and refractions. Prevents generic or unconvincing light effects.

Don't ship this

Common mistakes

✗ The figure exhibits unintended active movements, disrupting the desired sense of passive suspension and dreamlike drifting.
Fix — Emphasize 'passive suspension' and 'drifting' explicitly. Add details like 'limp limbs' or 'minimal resistance' to avoid unintended gestures.
✗ The generated visuals appear flat, lacking depth and cinematic texture, breaking the surreal immersion.
Fix — Reinforce 'shallow depth of field' with specific f-stops (e.g., f/1.8) and describe blurred background elements, like 'softly diffused distant bubbles'.
✗ The accompanying score feels generic or fails to evoke the specified melancholic, serene mood, undermining the visual.
Fix — Provide specific instrumentation (e.g., 'sustained cello notes,' 'ambient synth pads') and emotional descriptors (e.g., 'subtle,' 'reflective').
✗ The resulting video lacks a distinct filmic quality, appearing too clean or digital for the intended aesthetic.
Fix — Explicitly request 'subtle film grain,' 'slight chromatic aberration,' and 'organic lens flares' in your prompt's visual style section.
✗ The underwater environment looks artificial, failing to convey natural water dynamics or light refractions.
Fix — Describe the water's characteristics: 'gentle particulate matter,' 'subtle internal currents,' and 'diffused surface ripples from above'.
✗ Lighting appears inconsistent or static, losing the dramatic effect of light shafts cutting through the water.
Fix — Detail how the 'warm, golden shafts' interact with the figure in each shot, ensuring dynamic shifts and visual continuity.

Frequently asked questions

Q.Can I use this for a short film instead of a documentary?

Yes, the prompt is designed for cinematic, atmospheric sequences, making it suitable for narrative shorts requiring surreal or introspective moments. Adjust the context slightly to fit your project's overarching story and character arc.

Q.How do I ensure the 'dreamlike' quality is consistently maintained?

Focus on consistent descriptions of slow movement, soft lighting, and muted colors. Reinforce these elements in every section of your prompt, from shot list to score cue, to guide the AI's interpretation and maintain mood.

Q.What if the AI model struggles with specific wardrobe details?

Simplify the wardrobe_description to focus on texture and how it interacts with water (e.g., 'flowing silk gown') rather than intricate patterns. Test simpler descriptions first, then gradually add complexity if the model handles it well.

Q.Can I introduce multiple figures into the sequence?

While the prompt specifies a single figure for focused introspection, you can experiment. Be aware that adding more subjects increases complexity for the AI and might dilute the sense of isolated suspension. Test with one figure first.

Q.How do I control the exact duration of 15-20 seconds?

AI models often generate clips of varying lengths. You'll likely need to generate several iterations and edit them together, or specify exact shot durations within your prompt if the particular model you are using supports that functionality directly.

Q.Will this prompt work for models not listed, like Midjourney video?

The prompt is optimized for leading video AI models (Veo, Kling, Runway) known for cinematic output. While you can try it with others, results may vary significantly in quality and adherence to specific cinematic parameters and stylistic nuances.

Version 1.0Last reviewed July 13, 2026

Reviewed by PromptInFlow Editorial Team