3 Major Prompt Parameters for Midjourney AI to Accurately

Midjourney AI has excellent ability to generate artistic styles makes it particularly well-suited for creative fields like illustration, concept design, and advertising. Three prompt parameters help you better control your AI-generated image style and improve your work efficiency.

Style Reference：--sref

--sref is the core parameter for Midjourney AI consistent style.

Like a meticulous art director, it accurately extracts the core stylistic characteristics of a reference image (such as brushstroke texture, color scheme, and lighting and shadow atmosphere).

This allows all generated content to remain firmly locked into a single style, easily resolving the issue of inconsistent styles across a series of artworks.

Case Study: The Creation of a Van Gogh-Inspired Cat

1\. Prepare a reference image: Upload Van Gogh's "Sunflowers" as a reference image and prepare the image URL;

2\. Enter the prompt: End the prompt with "--sref reference image URL."

3\. Rendering: The generated image will be based on the "Sunflowers" reference image, generating Van Gogh-esque brushstroke textures and rich, oil-painted colors.

Character Reference：--cref

--cref is the guardian of character consistency.

It deeply extracts the core features of a character from a reference image: hairstyle, facial features, clothing details, and even facial expressions, ensuring that the character remains the same regardless of the scene or action. It's especially suitable for creating comic character designs, game characters, or IP images.

Core Value: Say Goodbye to "Change Scene = Change Face"

With traditional character generation, changing a character's actions can completely change their appearance. By using cref as a reference for character designs, whether you're "battle in a space station" or "reading a book in a cafe," AI accurately preserves the character's appearance details, clothing, equipment, and other core features, even preserving their usual expressions.

Overall Reference：--oref

--oref is a versatile tool for capturing both style and composition.

It can borrow the stylistic characteristics of a reference image while replicating its compositional logic, content elements, and atmosphere. It's ideal for scenes that require a complete emulation. For example, you can use the composition of a travel photo to create an illustrated version of the same scene.

Core Value: More Flexible Stylistic Application

Using oref as a reference, the AI will preserve the foreground, middle, and background layers while applying the visual style according to the prompts.