How to Create Wes Anderson-Style Shots with AI? Pastel Fairy-Tale Videos? NadouPro Prompt Practice

A perfectly composed image is already telling a story before any character speaks. One type of visual style pushes this precision to the extreme: strictly symmetrical frames, soft pastel tones, and scenes arranged as meticulously as miniature models—the audience feels as if they are walking through a fairy-tale dollhouse. This article breaks down every component of this style and provides prompt methods for implementing them one by one in NadouPro.

What Are the Visual Genes of This Style?

The core of this kind of stylized image is a highly controlled visual language: lighting, camera movement, color, and set design are all precisely planned. The image is usually symmetrical and flat, with compositions like picture-book illustrations. Every prop stays exactly where it belongs, and the camera acts as a restrained observer.

To recreate it, your prompt must cover five layers: composition, color, camera movement, lighting, and set design. Below, we break them down layer by layer.

How to Write Symmetrical Composition into a Prompt

Symmetry is the most eye-catching feature of this style. The subject is centered or arranged in a mirrored layout, naturally drawing the viewer’s attention to the sense of order in the frame.

Key points for converting it into prompts:

• Flat composition: explicitly write “flat composition” and “no deep perspective” to prevent the model from freely adding depth of field

• Centered subject: give characters, props, doors, and windows a clear positional instruction such as “in the exact center of the frame”

• Eye-level camera: the camera is level with the character’s gaze, and the character faces the lens head-on

• Geometric order: balanced arrangement of straight lines and circular elements; “rigorous geometric composition” is an effective control phrase

A usable basic description:

A perfectly symmetrical flat composition, the character standing in the exact center of the frame, facing the camera head-on, eye-level camera position, a rigorous sense of geometric order.

Pastel Palette: Name Specific Colors, Don’t Say “Beautiful Tones”

Color defines the mood of a scene. This style often uses exaggerated yet restrained color combinations: low-saturation mustard yellow, faded pink, soft teal, and vintage brown to create nostalgia; sometimes, bright colors are also used to contrast with a downbeat plot, intensifying emotional tension.

The key method: give the colors specific names. “Soft tones” is an ineffective instruction; “faded pale pink and low-saturation mustard yellow” can be executed more accurately. You can prepare fixed palettes for different scenes and reuse the same set across the whole sequence, immediately establishing stylistic consistency.

Two reference palette directions:

Palette 1: soft cyan-blue, coral pink, warm orange, teal, suitable for bright and cheerful scenes
Palette 2: olive green, gray-green, off-white, pale pink, suitable for quiet and nostalgic scenes

Place the color instruction near the front of the prompt, and clearly state that “the entire sequence uses this palette consistently.”

Camera Movement: A Restrained Camera Feels More Refined

In this style, the camera almost never shakes; it moves along an invisible track. Each type of movement has a narrative function:

• Static Shot: objective and detached, letting the composition speak for itself

• Horizontal Dolly: sweeping across the scene on a perfectly horizontal path to reveal information in sequence; for a “perfectly horizontal dolly shot,” you must lock in “horizontal”

• Slow Dolly In: brings the audience into the character’s inner world

• Slow Dolly Out: reveals the absurdity or loneliness of a larger environment

There are two ways to control camera movement in NadouPro: write clear action verbs in the prompt, such as dolly left/right, dolly in, dolly out, or static, and let a model that supports camera blocking execute them—Qizhi-VGOF has strong instruction understanding and stable camera scheduling, making it the first choice for this kind of precise camera movement requirement; when camera positions are more complex, use 3D Director for visual multi-camera design and timeline animation, with a maximum output of a complete 90-second video.

Multi-Shot Storytelling: Write the Structure into the Prompt

A single shot cannot carry a complete narrative. The way to write a multi-shot prompt is to treat it as a structural blueprint: what the opening image is, what action happens in the middle, where the camera finally stops, and the shot size and movement of each shot should all be clearly explained.

NadouPro’s multi-shot capabilities: Jimeng-Seedance 2.5, Tongyi-WAN 3.0 (intelligent multi-shot), MiniMax H3 (multiple shots in one generation), and Happy Horse 1.1 (intelligent multi-shot) all support shot switching within a single generation; when you need strict execution according to a storyboard script, first use Storyboard to split the script into shots with one click, then generate shot by shot or create a one-click storyboard video, and finally assemble it in Clip.

A structural example of a multi-shot prompt:

Three-shot narrative: first shot, a perfectly horizontal dolly shot sweeping across the interior of a vintage train carriage, with a low-saturation cyan and vintage brown color scheme; second shot, static eye-level camera position, two characters sitting upright and motionless; third shot, a slow dolly out revealing the entire carriage. Ambient sound and background music match the emotion of the image.

Sound Design: Dialogue and Ambience in One Pass

All NadouPro video models support audio-visual synchronization, so ambient sound and background music can be generated together with the image. For dialogue needs, use the Voice Dubbing capability (Qisheng Omni): emotional expression is natural, and Voice Dubbing and Voice Conversion are completed in one integrated process.

Key points for sound planning: plan resolution, visual details, and sound direction as three separate things; do not squeeze them into one sentence. Write ambient sound semantically, such as “ocean waves,” “wind,” or “quiet room.” Run dialogue separately through the Voice Dubbing workflow and then composite it for higher control precision.

Character Consistency: No Face Changes Across Shots

Stylized storytelling often needs the same character to run through multiple scenes. NadouPro’s method has three steps:

1. Build anchors: use Multi-Grid to generate Character Face 3-View, Character Body 3-View, and Character Expression 9-Grid, fixing the character’s appearance, clothing, and temperament

2. Pass references: every time you generate a video, input the finalized character image as a reference image; Multi-reference mode can lock both character and scene at the same time

3. Manage centrally: organize all shots on the same canvas and reuse character reference assets globally

Lighting and Set Design: Details Create the Fairy-Tale Feeling

Lighting: soft diffused light is the standard choice, removing harsh shadows and creating a picture-book feeling. In the prompt, write “soft volumetric light,” “diffused studio light,” and “no harsh shadows.” On the image side, if you already have a satisfying composition but the lighting is wrong, you can use Relight to redesign the lighting.

Set design: costumes and props are silent character descriptions. Guide the vintage texture with specific material words—corduroy, tweed, pastel knitwear; name specific scene elements—vintage wallpaper, old-fashioned furniture, neatly arranged props. The scene must feel fully controlled, with no random objects; every item should serve the geometric composition.

Negative space: empty space is just as important as solid objects. Clearly write “large area of negative space above the character” and “spacious, clean background” to prevent the frame from being filled with clutter. Negative space emphasizes the subject’s isolation or smallness and is a powerful emotional tool for this style.

Ten Scene Prompts You Can Use Directly

The following ten prompts are written according to five elements: “scene + composition + palette + camera movement + lighting.” You can replace the subject and colors as needed.

Symmetrical hotel lobby: a perfectly symmetrical vintage hotel lobby, eye-level flat composition, faded pale pink and low-saturation mustard yellow palette, a concierge standing in the exact center of the frame facing the camera, soft volumetric light, static camera.

Vintage train carriage: three-shot narrative, interior of a vintage train carriage, soft cyan-blue and vintage brown color scheme; opening with a horizontal dolly shot, cutting to a static eye-level shot of two people sitting upright, ending with a slow dolly out; retain the ambient sound of the carriage in motion.

Character office monologue: load the finalized character reference image, the character wears a vintage tweed suit and sits in a neatly arranged mid-century office, with a large area of negative space above the head, completely static camera position, heavily diffused studio light, the character delivers a monologue to the camera.

Pastel bakery storefront: eye-level camera facing a vintage bakery storefront, off-white pale pink and soft mint green palette, completely static camera, soft light removing shadows, paired with bright and cheerful background music.

Symmetrical restaurant: formal restaurant scene, low-saturation mustard yellow and soft teal color scheme, a rectangular dining table located in the exact center of the frame, extremely slow dolly out, characters seated quietly, with whisper-like dinner conversation in the background.

Vintage library: a perfectly horizontal dolly shot sweeping across the archive area of a library, teal-green bookshelves, an old rotary phone and neatly arranged books on display, diffused studio light, text on the book spines clearly readable.

Ocean research vessel: four-shot sequence, brightly colored research vessel deck, main color is soft cyan-blue; opening with a static extreme wide shot, slow dolly in, symmetrical close-up, and finally dolly out; the ambient sound is continuous ocean waves.

Phone booth conversation: the character wears a vintage corduroy jacket and stands inside a bright red phone booth, with a large area of negative space around the booth, eye-level static camera position, load the character reference to maintain consistency, the character completes a full dialogue scene.

Mountaintop observatory: an old-fashioned observatory backed by an open sky, olive green and faded pale pink color scheme, locked flat composition, lighting with no dramatic shadows, building sign text is clear, ambient sound is mountain wind.

Symmetrical garden maze: two-shot narrative, a neatly trimmed garden maze, deep green and vintage brown color scheme; first shot is a static extreme wide shot with the character centered, second shot slowly pushes toward the character, maintaining facial consistency, with lyrical background music.

For scenes that require clear text, such as book spines and signs, on the image side it is recommended to use Qizhi-IGI 2 (precise text and layout) or Qizhi-INB 2 (good text rendering) to generate a still frame first, then enter the Image to Video workflow. This is more controllable than generating directly.

Clarity Check Before Delivery

The finished video’s texture is determined by a few small things: whether the focus is solid, whether the composition is clean, whether details are readable, and whether the lighting matches the style. For scenes containing text, check the text first, then check the motion. NadouPro’s Frame Interpolation / Ultra HD capability automatically completes upscaling and frame interpolation; run it once before delivery, and the image finish will improve noticeably.

Frequently Asked Questions

Q: Why is the generated image not “symmetrical” enough?
In most cases, the prompt contains words that imply depth and angle at the same time. Check whether you wrote “eye-level camera position,” “flat composition,” and “centered subject,” and remove conflicting instructions such as “perspective,” “high angle,” and “dynamic angle.”

Q: What should I do if the palette keeps drifting?
Write the palette as a fixed short phrase and place it in the same position in every prompt; when creating a series, use character profile images and a four-grid scene mood board to lock the visual tone first, then use them as reference images throughout subsequent generations.

Q: What if a horizontal dolly shot has shake or perspective changes?
Emphasize “perfectly horizontal,” “constant speed,” and “no perspective change” in the prompt; for model selection, prioritize Qizhi-VGOF because its camera scheduling is stable. If the result is still not ideal, generate in segments and then connect them in Clip.

Q: Is this style suitable for commercial content?
The style itself can be used for commercial creation, but pay attention to two points: do not imitate the signature personal style of a specific living artist for commercial claims, and avoid visual confusion with existing brands. The scope of commercial licensing is subject to the NadouPro platform terms.

Q: Is the workflow different for image versions and video versions?
On the image side, use Text to Image plus Photo Parameters—Focal Length, Depth of Field, Color Science, and Lens Set, suitable for Qizhi-IGI 2, Qizhi-INBP, and Qizhi-INB 2—to lock the composition and palette first; after you are satisfied, use it as the first frame for Image to Video. Creating the image first and then the video significantly reduces rework.

Explore more features on NadouPro: NadouPro

Contains AI-generated information that may be incorrect. Please verify independently.