How to Read a Style Decode: A Complete Walkthrough
When AIKIZI decodes a style, it produces layers of data. This guide teaches you how to read each layer, understand what the AI found, and use the output to create new work.
01 — The Style Names
Three Names, Three Lenses
Every style decode produces three name variants. These are not synonyms — each captures a different axis of the style's identity.
Take Somber Cinema, decoded from 10 images:
Atmospheric Desaturated Realism
Cinematic Low-Key Tonalism
Soft Analog Noir
The first name describes the rendering approach — atmospheric, desaturated, realistic. The second names the lighting philosophy — cinematic low-key tonalism (dark overall with directed light). The third captures the emotional register — soft, analog, noir. Together they triangulate the style from three directions.
→
How to use this: When describing your style to another creator or to an AI model, use all three names. They are more precise than any single label.
02 — The Four DNA Layers
From Foundation to Finish
Style DNA is a four-layer stack. Each layer builds on the one below, from the foundational medium to the atmospheric finish.
Base Layer
The foundational medium. For Somber Cinema: Cinematic photo-realism — the world is captured as a camera would see it, with photographic fidelity.
Layer 1 — Technique
Low-key tonalism — the lighting approach. Dark compositions with directed, sculpting light. Most of the frame lives in shadow.
Layer 2 — Texture
Analog film grain — the tactile quality. Every surface carries fine grain that gives the images a material, archival quality.
Layer 3 — Atmosphere
Muted earth palette — the emotional finish. Desaturated ochres, slates, and warm blacks. The world seen through memory.
Understanding how the layers work helps you describe style more precisely in your prompts. The base layer defines the medium. Layer 1 defines the light. Layer 2 defines the texture. Layer 3 defines the mood. When one element feels off in a generated image, the layer system helps you identify which variable to adjust.
03 — Key Tokens
The Building Blocks
Tokens are two-word descriptors that capture the style's essential visual ingredients. They function as a vocabulary — use them in prompts, in conversations with other creators, or as a checklist when evaluating your own work against a reference style.
Chiaroscuro transitions
Muted tonality
Tactile grain
Atmospheric perspective
Desaturated palette
Cinematic lighting
Subdued contrast
•
Important: Tokens are not tags. "Chiaroscuro transitions" does not mean "use chiaroscuro." It means the style specifically uses gradual transitions between light and shadow, rather than hard-edge chiaroscuro. Precision matters.
04 — Color System
The Emotional Palette
The AI extracts the dominant 5-color palette with a named palette mode. For Somber Cinema: Low-saturation cinematic earth tones with warm amber accents.
These are not random samples. The AI identifies the structural colors — the ones that define the style's emotional character regardless of subject matter. A portrait and a landscape in Somber Cinema will share these exact tonal relationships.
05 — Rendering, Texture, Spatial
The Technical Fingerprint
Rendering
How the image was made: atmospheric filmic grain, soft feathered edges, tactile organic surfaces
Texture
What surfaces feel like: fine film grain, weathered organic quality, subtle analog noise
Spatial
How depth is handled: shallow focal depth, hazy nostalgic warmth, soft selective focus
Motion
The temporal quality: static contemplative stillness — time is frozen, not moving
06 — Putting It Together
From Analysis to Creation
1
Read the DNA layers
Understand the foundational medium and each overlay. This is your style's skeleton.
2
Study the tokens
These are your creative constraints. Every new image in this style should contain most of these tokens.
3
Apply the palette
Use the 5-color system as your tonal guide. It defines the emotional register of everything you create.
4
Use the master prompt
The decoded style includes a prompt template optimized for AI generation. Use it as-is or modify individual elements while keeping the structure.
How many images do I need for a style decode?
3 to 10 images that share a visual identity. More images produce more accurate results. The images should vary in subject but share the same aesthetic language.
Can I decode a style from images by different artists?
Yes. If images from different artists share a visual language, the AI will identify the shared DNA. This is useful for mapping aesthetic movements or trends.
What if my style has no name?
The AI generates names based on what it finds. Every visual identity gets named — that is part of the value. You may discover your work belongs to a tradition you did not know existed.
Start with Single-Image Decode
Decode Your First Image
Upload an image to AIKIZI Studio and run a single-image Decode to explore the analysis system. Style Decode (multi-image, 3–10 inputs) is available to select accounts. Single-image Decode is available to all users via Studio.