Midjourney Features

Repeat with —repeat (—r)

icon
The --repeat or --r parameter in Midjourney allows to create multiple job in a single prompt.
Repeat is particularly helpful when you’re experimenting with a prompt and want to see a wider range of possible outputs without the need to manually reissue the command multiple times.

a young girl in a cafe in Milan --style raw --sref random --r 3

—sref random assigns a random style reference to the image

Here are the key details about this parameter:

  1. Usage: You add –repeat or –r followed by a number to your prompt.
    Range: The parameter accepts values from 1 to 40.
  2. Function: It tells Midjourney to run the same prompt multiple times, generating a new set of images each time.
    Purpose: This is useful for quickly rerunning a job multiple times without having to manually re-enter the prompt.
  3. Syntax example: your prompt here --repeat 5 This would generate 5 separate sets of images from the same prompt.
  4. Efficiency: It’s a time-saving feature, especially when you want to explore multiple variations of the same prompt.
  5. Subscription impact: Each repeat counts as a separate job, so it will use up your subscription’s job allowance accordingly.
  6. Compatibility: This parameter works across different Midjourney versions and can be combined with other parameters
Permutation

icon
Allows users to generate multiple variations of an image prompt using a single command. This is achieved by using curly braces {} to enclose a list of options, separated by commas, within the prompt. Each option within the braces results in a separate image being generated, allowing for efficient exploration of different creative possibilities.

How Permutations Work

Basic Structure: You create permutations by placing options within curly braces {}. For example, the prompt a {red, blue, yellow} car will generate three separate images of cars in red, blue, and yellow.

Midjourney Prompt to try:


a {red, blue, yellow} car

Prompt with permutation

  • Combining Options: You can combine multiple sets of permutations in a single prompt. For instance, a {watercolor, oil painting, digital} portrait of a {Victorian lady, modern-day businessman, futuristic cyborg} would generate nine different images, covering all combinations of the specified styles and subjects[1].
  • Nesting Permutations: It’s possible to nest permutations by placing curly braces within other curly braces. This allows for even more complex combinations and a greater number of variations.

Midjourney Prompt to try:


a {watercolor, oil painting, digital} portrait of a {Victorian lady, modern-day businessman, futuristic cyborg}

Prompt with combined permutation

Use Cases

  • Artistic Exploration: Permutations are particularly useful for artists and designers who want to explore different styles, subjects, or settings without manually creating each variation. This can include experimenting with different artistic styles, subjects, or even times of day for a scene.
  • Efficient Workflow: By automating the creation of multiple images from a single prompt, permutations streamline the creative process, saving time and allowing for rapid iteration and comparison.

Limitations

  • Mode Restrictions: Permutations are only available in “Fast” or “Turbo” modes, not in “Relax” mode. The number of permutations you can run simultaneously depends on your subscription level, with higher tiers allowing more concurrent jobs.
  • Subscription Limits: Basic subscribers can run up to 4 permutations at a time, while higher-tier subscriptions allow for more, up to 40 for Pro and Mega subscribers.
Character Consistency with —cref

Character Description

[age] [hairstyle] [hair color] [eye color] [any other facial features]

icon
CREF (Character Reference) QUICK CHEATSHEET
refURL = url link of the character image prompt — cref {refURL} — cw 0 | maintains only the face
prompt — cref {refURL} — cw 25 | maintains face and hair, not the outfit
prompt — cref {refURL} — cw 50 | maintains face, hair, and some parts of the outfit
prompt — cref {refURL} — cw 100 | maintains face, hair, and outfit

icon
The --cref parameter in Midjourney is used for maintaining character consistency across different images.

It allows users to provide a character reference image, ensuring that the generated images consistently depict the same character traits and appearance in various scenes or contexts. This feature is particularly useful for storytelling, branding, and creative projects where uniformity in character depiction is crucial.

Key Features of -cref

  1. Character Consistency: By using the -cref parameter, users can ensure that a character maintains the same appearance across multiple images. This is essential for projects that require a consistent visual narrative or branding.
  2. Reference Image: The parameter works by using a URL of a character reference image. Midjourney analyzes this reference to replicate the character’s features in new images.
  3. Adjustable Fidelity: Users can adjust how closely the generated image matches the reference using the cref cw (character weight) parameter. This allows for flexibility in how strictly the AI adheres to the reference image.
  4. Best Practices: The -cref feature works best with AI-generated images, such as those previously created with Midjourney. Using real photographs may not yield accurate results due to potential distortions.
  5. Version Compatibility: As of the latest updates, the -cref parameter is designed to work with Midjourney version 6, which is the default version for all users.

Usage

To use the --cref parameter, include it in your prompt followed by the URL of the character reference image. For example:


a hero in action --cref <https://example.com/character.jpg>

This command will generate an image of a hero in action, maintaining the character’s consistency as defined by the reference image.

Overall, the --cref parameter is a powerful tool for creators who need to maintain consistent character designs across multiple images, enhancing storytelling and brand identity through visual uniformity.

Unfiltered Style —style raw
icon
The --style raw parameter in Midjourney is to generate images with a more realistic and unfiltered appearance. This mode is designed to strip away some of the default stylization that Midjourney typically applies, resulting in images that are closer to a natural or photographic look.

Key Features of –style raw

  • Realism Over Stylization: The primary function of the –style raw parameter is to produce images that are more literal and less stylized. This makes it ideal for users who want a realistic depiction of their prompts without the artistic embellishments that Midjourney might otherwise add.
  • Literal Interpretation: When using –style raw, the AI focuses more on accurately representing the input prompt without adding creative or artistic flair. This can be particularly useful for detailed and specific prompts where accuracy is prioritized over artistic interpretation
  • Flexibility in Usage: Users can activate –style raw either by adding it to individual prompts or by setting it as a default mode in their Midjourney settings. This flexibility allows users to switch between stylized and raw outputs as needed
  • Impact on Prompt Length: The effect of –style raw can vary based on the complexity and length of the prompt. For shorter prompts, it tends to produce straightforward and realistic images, while longer, more detailed prompts can benefit from its ability to capture intricate details accurately

Overall, the –style raw parameter provides a tool for users who prefer a more natural and realistic output, offering greater control over the final image’s aesthetic by minimizing the default stylization applied by Midjourney.

Stylization with —stylize or —s
icon
The --stylize (or --s) parameter in Midjourney is a tool that allows users to control the level of artistic interpretation applied to the images generated from their prompts.

This parameter influences how much the default artistic style of Midjourney is incorporated into the final image, affecting aspects like color, composition, and form.

How -stylize Works

  • Range and Default Value: The -stylize parameter can be set to any integer value between 0 and 1000, with the default being 100. This default setting provides a moderate level of artistic flair while still adhering closely to the prompt.
  • Effect of Different Values:
    • Low Values (0-100): Lower stylize values result in images that closely match the prompt with minimal artistic embellishment. This is useful when accuracy and clarity are prioritized over artistic expression.
    • Medium Values (100-250): These values provide a balanced approach, adding some artistic elements while maintaining a reasonable connection to the prompt.
    • High Values (250-750): Higher values introduce more artistic flair, with images becoming more stylized and potentially diverging from the literal prompt. This can result in more creative and abstract visuals.
    • Very High Values (750-1000): At these levels, the images are highly stylized, with the AI taking significant creative liberties. This can lead to outcomes that prioritize artistic expression over the prompt’s specifics.

Usage

To use the --stylize parameter, you simply add it to your prompt with the desired value. For example:


Futuristic cityscape with flying cars --s {0, 1000}

This command will generate two images of a Futuristic cityscape with flying cars:

  • one with the lowest level of artistic interpretation.
  • one with the highest level of artistic interpretation


Practical Applications

  • Creative Exploration: The -stylize parameter is ideal for users looking to explore different artistic styles and effects, allowing for a wide range of visual outputs from a single prompt.
  • Customization: By adjusting the -stylize value, users can tailor the artistic intensity to fit their specific needs, whether they require a realistic depiction or a more abstract and expressive image.

Overall, the --stylize parameter in Midjourney offers users a powerful way to influence the artistic direction of their generated images, providing flexibility and creative control.

Batch Command

Use this parameter below to test several style weights


--sw {20, 40, 80, 100, 250, 400, 600, 800, 1000}

Variety with —chaos
icon
What is —chaos?The --chaos option in Midjourney is a parameter that influences the diversity and variation among the images generated in response to a prompt. When you use this parameter, you can control how similar or different the four images in the grid will be.

How -chaos Works

  • Range and Default: The -chaos parameter accepts values from 0 to 100, with the default being 0. A value of 0 means the images will be very similar, while a value of 100 will result in highly varied and diverse images.
  • Effect of Higher Values: Increasing the chaos value leads to more unusual and unexpected compositions. This means that the images will be more distinct from each other, offering a broader range of creative possibilities.
  • Use Cases: Higher chaos values are useful when you want to explore a wide array of artistic interpretations or when seeking inspiration through diverse visual outputs. Lower values are preferable when you need consistency and reliability in the generated images

-chaos Values

0 to 100


0
means the images will be very similar

100 will result in highly varied and diverse images


—chaos 0


—chaos 50


—chaos 100

Anime style with —niji
icon
The --niji style in Midjourney is a specialized mode designed to generate anime-style art. 

Features of the -niji Style

  • Anime Focus: The -niji style is specifically tailored for creating images with anime aesthetics. This mode is ideal for users looking to generate art that resembles traditional anime and manga styles.
  • Distinct Styles: Within the -niji mode, users can choose from several distinct styles, each offering a unique take on anime art. These include:
    • Original: The default style of the Niji model before updates.
    • Cute: Emphasizes adorable features, often with exaggerated characteristics like larger eyes and smaller bodies.
    • Scenic: Focuses on detailed and beautiful backgrounds, creating a cinematic feel.
    • Expressive: Highlights facial expressions and can blend 2D and 3D elements for a more dynamic look.
  • Customization: Users can further customize their outputs by combining the -niji style with other parameters and prompts, allowing for a wide range of artistic expressions.

How to Use -niji Style

To use the --niji style in Midjourney, you can add the --niji 6 parameter to your prompts.

—niji’s default version is 6.

Both work: --niji and --niji 6

For example:


a girl drawing and smiling --niji


a girl drawing and smiling —s 200 —v 6.1

a girl drawing and smiling, anime style —s 200 —v 6.1

a girl drawing and smiling —s 200 —niji 6


Seamless Pattern with —tile
icon
STEAM FOR SEAMLESS PATTERNS S: SUBJECTT: None
E: None
A: MOOD (Optional)
M: ART STYLE and other modifiers.


icon/imagine prompt: [SUBJECT], [MOOD], [ART STYLE] —ar 1:1 —tile

Example


papaya, watercolor art, minimalist on white background --chaos {0, 10, 15, 20} --ar 1:1 --tile --style raw --stylize {80, 100, 250} --v 6.1


baby room giraffe and elephant, cartoonish, minimalist on white background --chaos {0, 10, 15, 20} --ar 1:1 --tile --style raw --stylize {80, 100, 250} --v 6.1


black shapes, Boho Japandi style, black minimalist, on white background --chaos {0, 10, 15, 20} --ar 1:1 --tile --style raw --stylize {80, 100, 250} --v 6.1


Modifiers for seamless patterns images

Abstract: Creating non-representational, abstract designs that focus on color, shape, and texture.

Geometric: Utilizing geometric shapes and patterns that can repeat seamlessly, such as hexagons, triangles, or other tessellating shapes.

Floral: Designing patterns with flowers, leaves, and other botanical elements that repeat naturally.

Minimalist: Focusing on simple, clean lines and shapes for a modern and understated look.

Vintage: Incorporating elements from historical or retro styles, such as art deco or Victorian motifs.

Organic: Using natural forms and textures, such as wood grain, water ripples, or stone textures.

Cultural or Ethnic: Drawing inspiration from traditional patterns found in various cultures, such as Moroccan tiles or Japanese wave patterns.

Fantasy: Creating whimsical or fantastical designs that might include mythical creatures or imaginative landscapes.

Illustrative: Using detailed, hand-drawn elements to create intricate and engaging patterns.


The --tile option in Midjourney is a parameter used to generate images that can be seamlessly repeated to create patterns. This feature is particularly useful for creating backgrounds, textures, and designs that need to be continuous without visible breaks or seams.

How the -tile Option Works

  • Seamless Patterns: When you include the -tile parameter in your prompt, Midjourney generates an image designed to repeat seamlessly. This means that the edges of the image align perfectly with each other, allowing for a continuous pattern when tiled.
  • Application: The -tile option is ideal for creating designs for wallpapers, fabric patterns, game textures, and other applications where a repeating pattern is needed.
  • Compatibility: The -tile parameter works with Midjourney models V1, V2, V3, and V5, V6 but it is not compatible with V4 or the Niji model
  • Usage Example: To use the -tile option, you simply append it to your prompt. For example:

/imagine prompt: geometric pattern, vibrant colors --tile

This command will generate a geometric pattern that can be repeated seamlessly.

Practical Use

  • Testing and Verification: After generating a tileable image, you can use tools like the Seamless Pattern Checker to verify the continuity of the pattern and ensure that it tiles correctly without visible seams.
  • Creative Applications: Tiled images can be used in a variety of creative projects, such as designing custom wallpapers, creating unique fabric prints, or developing game textures. They are also popular for use in digital art and design projects where a repeating motif is needed.

Ut id nisl quis enim dignissim sagittis. Curabitur at lacus ac velit ornare lobortis. Nam commodo suscipit quam. Aenean massa. Nunc nonummy metus.

Copyright © 2023 World Masterclass. All rights reserved