How to Use Midjourney: A Guide to Create AI Art
Guide of Commands, Features and Tricks.
Midjourney’s AI Art Generative Tool is an innovative and powerful AI tool that empowers your creativity by utilizing artificial intelligence and machine learning to produce stunning visuals.
Midjourney generates images based on a short text description called a Prompt with the primary “imagine” command.
Utilizing this command is quite simple. However, if you desire greater styles and variations over the images, you’ll need to master the advanced prompts and parameters. Today, I have prepared a comprehensive guide to
help you thrive in this creative journey! 🚀
As always, allow me to welcome you with a cup of coffee! ☕️
If you are just starting, check out this quick guide to set up Midjourney on Discord
Today, I will walk you through a list of parameters and commands to master Midjourney. From basic skills to advanced tips, you’ll unleash the full potential of this incredible AI tool.
Today’s content is separated into several parts:
- Part One: Prompt types
- Part Two: Parameters
- Part Three: Advanced Command
- Part Four: Version and Style
- Extra: Latest Features & Creative Use Cases
As you all know, to interact with the Midjourney Bot on Discord, we will need to use commands. Commands are used to create images, customize settings, track user info, and perform other helpful functions.
In order to create images in Midjourney, we need to input a prompt followed by the command “/imagine”.
That’s what we are going to start today — Prompt.
🌟 Part One: Prompt Types
A Prompt is a short text phrase that the Midjourney Bot interprets to produce an image and can be broken down into three major types in the following structure.
1. Image Prompt
Use images to impact a Job’s composition, style, and colors.
2. Text Prompt
Description of the image. Combine image prompts with text prompts for varied outcomes.
3. Advanced Text Prompt
3.1 Multi Prompt “::”
Using “ ::
” in the prompt lets Midjourney consider each part individually, allowing for assigning relative importance to the prompt.
For instance, “Space ship” results in a sci-fi spaceship, while “space::ship
” creates a sailing ship in space.
Prompt Weights
Besides, We can also add a number after it determines the relative importance of that specific part.
Modifying the prompt to “space::2 ship
” prioritizes “space” twice as much as “ship,” resulting in space-themed images with ships as secondary elements.
Order of the Prompt
Besides utilizing prompt weight, the order of your words also has a significant effect on the final image.
Begin with the key keywords that hold the most significance.
- Words 1–5: carry significant influence and are likely to appear prominently in the initial results.
- Words 5–20: also hold influence, but may need 2–3 rerolls to highlight them in the image.
- Words 21–40: remain relevant, but might require several rerolls to be noticeable.
- Words 40+: likely to be overlooked
3.2. Advanced Text Prompt — Permutation Prompt
Permutation Prompt allows users to generate prompt variations using a list of options separated with commas ","
within curly braces “{ }
” to create different combinations.
Prpmpt: A Fox { Inkpaint, Pop art, Cyberpunk}
🌟 Part Two: Parameters
Parameters are additional choices incorporated into a prompt, which modify the generation process of an image.
Parameters go at the end of the prompt, usually in this way
--parameter <value>
1. Aspect Ratios --ar [ width : height ]
Here are common aspect ratios used in various contexts:
--ar 1:1
Default aspect ratio.
Landscape:--ar 5:4
Common print ratio.--ar 3:2
Common in print photography.--ar 16:9
Televisions and computer monitors
Portrait:--ar 2:3
Good for Posters and Social Media Posts--ar 9:16
Good for Stories and Video
The aspect ratio impacts the shape and composition of a generated image. V5 brings in various aspect ratios, including unconventional ones like 100:1.
2. Chaos --c < 0–100 >
Higher values result in more unusual and unexpected results and compositions.
3. Image Weight --iw < 0–2 >
Image prompt weight relative to text weight. The higher the iw value, the closer the final result will resemble the image.
Use --iw 2
if you want the result to look close to the original image. It is great for creating your own avatar and replicating images.
Prompt: [ Image URL ], a handsome guy with short brown hair smiling, keep the consistency of action, expression, clothing, shape and appearance of the photos, photography from disney pixar studio, anime, super detail
--iw 2 --s 180
More detialed guide can be found here:
4. Stop --stop < 10–100 >
Stop a job partway to create less detailed and blurrier results.
5. Stylize --s < 0-1000 >
Application of Midjourney’s default aesthetics including artistic color, composition, and forms. High stylization values create images that are very artistic but less connected to the prompt.
6. Weird: --w < 0–3000 >
Adds quirky and offbeat qualities to your generated images, resulting in unique and unusual aesthetics.
🎨 Note that using “ — stylize and — weird” together with the same value can result in strange yet pretty images.
7. No: --no [ ]
Negative prompting, for instance --no red
remove the red color from the image.
To eliminate more than one element, separate it with “,” like this
Your prompt --no item1, item2, item3
8.
Seed --seed < 0–4294967295 >
Using the same seed number will produce a similar style.
9. Tile: --tile
Generate images that can be used to create seamless patterns for print-on-demand products or packaging
Other Parameters
Modes: --relax,--fast & --turbo
Run a job using Relax/Fast/Turbo Mode.
Quality: --q <.25, .5, or 1>
Higher values use more GPU minutes.
Repeat:--r < 1–40 >
Create multiple Jobs from one prompt.
Video: --video
Create a short video of image creation.
🌟 Part Three: Advanced Command
1 ./blend
Allow you to upload 2–5 images and merge the aesthetics.
More examples can be found here:
How to Combine Images in MidJourney
Combine images with one single command easily
bootcamp.uxdesign.cc
2./describe
Generate four prompts based on an uploaded image. It allows you to learn more about prompts writing as well!
3 ./prefer option:
Create a custom option. Set up parameters that you can add to the end of prompts, making your workflow more efficient.
For example: /prefer option set
my
--s 500--ar 16:9
creates an option called "my” that translates to --s 500--ar 16:9
.
/imagine prompt: “Cat — my” will become “ Cat — s 500 — ar 16:9” directly
- /prefer option list: View current custom options.
- /prefer suffix: Specify a suffix at the end of every prompt.
4./shorten:
Analyze a prompt and return shorter prompts omitting redundant words
5. /remix:
Remix Mode can alter prompts, parameters, model versions, or aspect ratios in variations. It takes the general composition of your starting image and incorporates it into the new Job.
🌟 Part Four: Versions and Style
Midjourney provides an easy way for you to switch between various versions and styles either by selecting in the setting panel or adding a parameter.
1. Settings
/settings: Using the settings command will open a panel that allows you to switch between version, style, quality, and modes.
2. Changing Parameters
Alternatively, you can switch versions and styles through the use of parameters.
--v [1/2/3/4/5/5.1/5.2/6]:
Switch between versions
--style raw:
Reduce default aesthetic
3. Niji Styles
Niji Mode is an anime-specific model with a vast knowledge of anime, styles and aesthetics.
--style [cute/ expressive/ scenic]
Besides the original style that is great for modern anime flair, Niji also comes with
- Cute: An adorable style that produces gentle pictures
- Default: The most creative style, great for drawing new concepts
- Expressive: A stylish, chic look with a Western influence
- Scenic: A cinematic look with strong dynamic lighting
Here is a demonstration of the 4 different styles of Niji Journey V5 with the same prompt:
The most updated version is Niji 6 now. Niji also has a mobile app.
🔥 Midjourney V5 New Features🔥
1. Vary (Region) — Inpainting
The “Vary (Region)” option allows you to select and regenerate specific parts of an image. This allows users to remove unwanted objects or fill in missing parts which works similarly as inpainting.
2. Zoom Out
It allows you to expand the canvas of an upscaled image by 1.5x, 2x or in custom aspect while keeping the original content intact.
3. Panning
Panning extends your image in four different directions ➡️⬅️⬆️⬇️. You can keep panning multiple times, resulting in a large and detailed panoramic image.
4. Upscale
Upscalers allow you to upscale your image instantly to 2x and 4x larger.
Example Usage of the 3 features
Let’s look at the example below where I regenerated a new scene with the same character
To add more settings, I used “Zoom-out 2x” and pan left with a custom prompt here.
V6 brings in a new “Pan” which works even better.
5. Style Tuner
Style Tuner allows you to personalize the appearance of your Midjourney images. It lets you customize and optimize Midjourney’s model to consistently generate images with colors, textures, character details and more to your liking.
/tune
is only compatible with Midjourney V5.2/tune
is only available in Fast Mode.
🔥 Midjourney V6 NEW Features🔥
1.
Style Reference --sref
Work similar to image prompts where you give a URL to one or more images that ‘describe’ the consistent style you want to work over.
2. Character Reference --cref
Character reference allows you to create a consistent character across images and in different styles.
3. Style Random --sref random
Style random allows you to explore different styles and new visuals to inspire you.
If you type --sref random
it'll replace it in your prompt with something like --sref 123456
(also known as sref code) and then if you type that random number back you'll be able to get back to that style again
To learn the latest feature of Midjourney V6, be sure to check out this list:
👉🏻Digital Guide: Master Midjourney 🎨🚀
If you are looking for more ideas and advanced tips for generating captivating AI art with Midjourney, I have also created a “Master Midjourney: Beginner to Pro” digital guide.
It comes with:
- 🎨 1,000+ expert prompts
- 💡 Advanced Midjourney prompt writing, structure and core techniques
- 🖼 Specialized guidance for graphic design, branding, logo, UI, photography, and more commercial uses
- 🚀 Pro tips to recreate styles, make avatars, generate anime artwork, and more
- 💰 AI Arts Monetisation Strategies
- 💼 Top AI tools to edit Midjourney creations
Hey friends 👋 I am Christie — An AI Art Educator & Creator 😊 Also a coffee lover ❤️☕️
If you’re craving more Insider Tips, Prompts and Tricks for creating and generating income with AI art, join me on this journey.
🎨 Hit that “Follow” button to stay in the loop for more AI wisdom
💌 Stay tuned for more tutorials!
Let’s explore the limitless possibilities of AI Art and build passive income together! 🚀💰