top of page

Get the Midjourney Images You Want: A Guide to Key Parameters

Apr 28

6 min read

0

3

0

midjourney blog post image
A Midjourney generated image using Midjourney Automation Suite

Ready to take your Midjourney creations to the next level? Going beyond simple text prompts lets you fine-tune your images exactly how you imagine them. Understanding Midjourney's parameters is the key to controlling the look, style, and composition of your generated art.

This guide walks you through the essential parameters and techniques. You will learn how to direct Midjourney more effectively.

Starting Simple: Your First Prompt

Let's start with a basic prompt using the `/imagine` command. We will use a simple phrase like:

`/imagine a castle from a fairy tale`

Midjourney quickly generates images that visually interpret this text. By adding parameters, you change how Midjourney creates these images.

Controlling the Look: Essential Parameters

Setting the Scene: Aspect Ratio

The `--ar` or `--aspect` parameter changes the shape of your image grid. The default is a square (1:1).

You can set different ratios:

`/imagine a castle from a fairy tale --ar 3:2`

This gives you a horizontal image. You can use many different ratios, like 7:4 or even input decimals as whole numbers (e.g., 139:1 instead of 1.39:1).

Embrace or Reduce Variation: Chaos

The `--chaos` or `--c` parameter influences how varied the initial image grid is. A high chaos number generates more unusual results. A low number produces more standard, predictable images.

Values range from 0 to 100. Use 0 for low chaos and 100 for high chaos.

`/imagine a castle from a fairy tale --c 10` (Low Chaos)

`/imagine a castle from a fairy tale --c 90` (High Chaos)

You might get unexpected images with high chaos, even if they don't strictly match the prompt text.

What to Leave Out: Using `No`

The `--no` parameter tells Midjourney what you do *not* want in your image. This helps remove unwanted elements that might appear based on your main prompt.

If your fairy tale castles always have trees, you can remove them:

`/imagine a castle from a fairy tale --no trees`

You can exclude multiple things by listing them, like `--no trees, sky, towers`.

Detail Level: The Quality Parameter

The `--quality` or `--q` parameter controls how long Midjourney spends generating an image. Higher values take longer and add more detail. It does not change the image resolution, only the detail level.

Default quality is 1.

Accepted values for current models are 0.25, 0.5, and 1. Higher numbers are rounded down to 1. The quality setting is applied during the initial image generation.

`/imagine a castle from a fairy tale --q 0.25`

Lower quality can sometimes be better for abstract or gestural art. Higher quality can be good for detailed things like buildings.

Get More Options: The Repeat Parameter

The `--repeat` or `--r` parameter runs a prompt multiple times at once. This is useful for exploring many visual options quickly, especially when combined with parameters like `--chaos`.

The number of repeats depends on your subscription plan (e.g., 2-4 for Basic, 2-40 for Pro/Mega).

`/imagine a castle from a fairy tale --r 2`

This simple command saves you from re-typing the prompt.

Consistency is Key: Using Seed

The `--seed` parameter is powerful for getting similar results. Midjourney uses a random seed number for each image grid it creates.

If you find an image you like, you can get its seed number. React to the image with the envelope emoji (✉️). Midjourney will send you a message with the seed number and job ID.

Use this seed number with your prompt to generate images with the same base structure or composition:

`/imagine a castle from a fairy tale --seed 12345`

Using the same seed with different parameters (like `--ar`, `--q`, `--stylize`) lets you see how those parameters change a consistent base image.

Want to streamline your Midjourney workflow, especially when managing seeds and variations? Check out the Midjourney Automation Suite from TitanXT. It can help you organize and scale your creations.

Exploring Different Aesthetics

Changing the Output Style: Model Versions

Midjourney releases new model versions regularly. Each model has a slightly different understanding of prompts and a different visual style. The current default model is 5.2.

You can specify a different model using `--version` or `--v` followed by the version number:

`/imagine gangster's paradise --v 4`

Or use the `/settings` command to set a default model.

Exploring Style Parameters

Some model versions offer specific styles. The `--style raw` parameter on models 5.1 and 5.2 reduces the influence of Midjourney's default artistic aesthetic, offering more control or a more photographic look.

`/imagine gangster's paradise --style raw`

The Niji model (often used for anime/illustration styles) has its own specific styles that can be used with `--niji`, such as:

  • `--style cute`

  • `--style scenic`

  • `--style expressive`

  • `--style original`

`/imagine gangster's paradise --niji --style cute`

Adding Artistic Flair: Stylize

The `--stylize` or `--s` parameter influences how strongly Midjourney's training on artistic principles is applied. Low `--stylize` numbers stick closely to the prompt but are less artistic. High numbers are more artistic but might stray from the prompt.

Values range from 0 to 1000 (for current models).

`/imagine gangster's paradise --s 0` (Low Stylize)

`/imagine gangster's paradise --s 1000` (High Stylize)

Special Use Cases

Patterns for Everything: The Tile Parameter

The `--tile` parameter generates images that work as seamless repeating patterns. This is perfect for creating textures, wallpapers, or fabric designs.

`/imagine abstract pattern --tile`

See Your Art Move: The Video Parameter

Adding the `--video` parameter creates a short video showing the image generation process, from the initial noise to the final result. It's like a time-lapse of your image being made.

`/imagine spaceship in space --video`

After the images are done, react with the envelope emoji (✉️) to get a link to download the video file.

Inject Some Unexpectedness: The Weird Parameter

The `--weird` or `--w` parameter adds quirky or offbeat qualities to your images. It's experimental and can produce unique outcomes.

Values range from 0 to 3000. 0 is the default (no weirdness). Note that `--weird` might not work perfectly with the `--seed` parameter.

`/imagine gangster's paradise --w 500`

Advanced Techniques

Using Images as Inspiration: Image Prompts

You can use images as part of your prompt to influence the style, composition, or colors. Place the image URLs at the beginning of your prompt.

Prompt structure:

`/imagine [image URL] [image URL] [text prompt]`

Or:

`/imagine [image URL] [text prompt]`

For example, if you have an image URL of a red car and an image URL of a spaceship, and want a spaceship in space:

`/imagine [URL of red car] [URL of spaceship] spaceship in space`

Your output might combine elements or styles from the source images with your text idea.

Edit Specific Areas: Vary Region

The "Vary Region" tool lets you select a part of an image and regenerate only that area with a new prompt. This is often called inpainting.

First, upscale an image. Then click the "Vary (Region)" button below it. A box appears allowing you to select areas you want to change. Type your new prompt for that area.

For example, on a picture of a person, select the head and type "antlers". Midjourney will try to add antlers to that spot.

Make specific changes to complex images without starting over. For fine-grained control over your image editing workflows, explore the capabilities provided by the Midjourney Automation Suite from TitanXT.

Changing Your View: Pan and Zoom Out

After upsizing an image, buttons appear that let you expand the view.

  • Zoom Out: Adds content to the edges of your image, making the original image appear "zoomed in" on a larger scene. You can zoom out 1.5x or 2x. You can also use a custom zoom.

  • Pan: Expands the image specifically in one direction (up, down, left, or right), keeping the existing content and adding more around it.

You can use these tools multiple times on the same image to continuously expand the canvas and scene.

Weighting Your Words: Multi-Prompts

Sometimes you want Midjourney to treat parts of your prompt as separate concepts or give more importance to certain words.

Use a double colon `::` to separate concepts.

For example, `/imagine space ship` gives you a spaceship. But `/imagine space:: ship` tells Midjourney to think of "space" and "ship" as distinct ideas, potentially giving you a ship in outer space.

You can also add weight to a term by adding a number after the double colon (e.g., `::2`). A higher number gives that term more importance.

`/imagine space::2 ship` (Space is twice as important as ship)

Conclusion

Learning Midjourney parameters unlocks a new level of control over your creative process. From basic aspect ratios and quality settings to advanced techniques like using seeds, image prompts, and varying regions, these tools help you move from beginner to experienced creator.

Experiment with different combinations and see how they change your results. Happy creating!

Apr 28

6 min read

0

3

0

Related Posts

Comments

Share Your ThoughtsBe the first to write a comment.
bottom of page
Midjourney Automation Suite - Automate your image generation workflows on Midjourney | Product Hunt