🎨 HOW TO MASTER MIDJOURNEY: FROM A TO Z | PART 2

In HOW TO MASTER MIDJOURNEY: FROM A TO Z | PART 1

1. Introduction

2. How to start with Midjourney?

3. Your first image!

4. Understanding the variations.

You think you have an idea for an excellent image and are eager to see what it will look like. However, you don't have any experience in creating or drawing images. So, how can you bring your unique image to life?

Imagine typing a few words and, voila, you get a stunning image! That's Midjourney for you. It's like having a magic art studio in your pocket. Perfect for beginners and pros, it turns your words into visual wonders. Get ready to unleash your creativity.

If you do not know where to start, please continue reading. Maybe you will become an expert in using Midjourney after finishing reading this passage. 😊

READ MORE: HOW TO MASTER MIDJOURNEY: FROM A TO Z | PART 1

PARAMETERS

The "parameter" of the prompt in Midjourney refers to additional instructions or details that you can include to guide the AI in generating the image. Parameters can specify things like image style, colors, level of detail, or any specific characteristics you want to emphasize in the generated image. By providing parameters, you can fine-tune the AI's output to better match your creative vision.

Aspect Ratios

• -aspect, or -ar Change the aspect ratio of the generated image

SEED

In Midjourney, the "seed number" serves as the starting point for image creation. These seed numbers are generated randomly, but you can also select one by using special parameters. Using the same seed number and prompt will produce similar images.

This is the prompt used, and the resulting picture has been assigned the seed number 2900061700.

a girl with curly hair, medium hair, light smile, turtleneck sweater, long skirt, snow, white scarf, sunset, bloom, top of hills, from side, hands on hips, looking away, looking at sunset, beautiful scenery, masterpiece, best quality

This is what you will get if you use the same prompt and the parameter --seed 2900061700:

a girl with curly hair, medium hair, light smile, turtleneck sweater, long skirt, snow, white scarf, sunset, bloom, top of hills, from the side, hands on hips, looking away, looking at the sunset, beautiful scenery, masterpiece, best quality --seed 2900061700

NO

--no is a negative prompting parameter. If you do not want something in your image, you can use this. For example, this is the image of the prompt “dinner table”.

As you can see, there are many candles in those pictures and if you do not want candles in your picture, you can use the --no paramater:

dinner table --no candles

This is what you get: different images with no candles

USING COMMANDS FOR SPECIAL THINGS

BLEND

The /blend command in Midjourney lets you combine multiple images into one. However, the /blend command doesn't work with text prompts.

How to use the Blend command:

To use the Blend command, you first type /blend in the Midjourney chat box and then choose at least 2 images and no more than 5 images to upload.

You can choose any image you want, but remember they must be image files, not URL links.

You can also add in specific blending parameters, Midjourney will prompt you to those. You can define how the images should be blended together. This may include specifying the blending mode (e.g., overlay, multiply, screen), adjusting opacity, or choosing other blending options.

This is what you get if you merge the two images above:

One blending tip: To get the best results, make sure the pictures you upload have the same ratio as the final image you want.

DESCRIBE

The /describe command in Midjourney is used to provide a text description of the image generated by the AI. It allows you to upload an image and generatefour prompts based on that image.

This description helps you understand the content and details of the image, especially if it's a complex or abstract composition. This feature is a fantastic way to expand your vocabulary and delve into various artistic styles and movements.

How to use /describe command:

You type /describe in your private chat with Midjourney Bot and then attach an image file to it:

For example, you choose this image:

These are four prompts that Midjourney might give you. The prompt consists of a text prompt and the aspect ratio of the uploaded image.

1️⃣ a table is full of food, in the style of 2d game art, hallyu, realistic hyper-detail, solarizing master, thick paint, storybook illustration, swiss style --ar 85:128

2️⃣ tray with many foods on, in the style of speedpainting, hiroshi nagai, steve henderson, range murata, high resolution, warmcore, lively illustrations --ar 85:128

3️⃣ the table has a lot of food on it, in the style of speedpainting, ken sugimori, high dynamic range, paul corfield, hallyu, solarizing master, bright palette --ar 85:128

4️⃣ a table filled with utensils, fruits and plates, in the style of cartoon realism, animecore, realistic hyper-detail, sun-kissed palettes, storybook illustration, thick paint, dansaekhwa --ar 85:128

REMIX MODE

The /prefer remix command in Midjourney allows you to generate new variations of existing images.

How to use /prefer remix command:

You can turn it on by typing /prefer remix in the chat box and then you should follow the instructions:

And then you should follow the instructions - clicking the variations buttons and changing your prompt according to your new idea. This is an original prompt and original picture you have gotten:

a person on Midjourney boat, background is a new world with weird animals and clouds, underwater have many weird fish and creatures

If you want to change the prompt to create variations of the first pictures, you choose the V1 button, Midjourney will open a small window below so you can change the prompt a little bit:

This is the new prompt:

a person on Midjourney boat, background is a new world with weird clouds shapes, weird-shape birds

Midjourney will provide you with new images:

Finally, you need to turn the Remix mode off by running /prefer remix again.

EXPERT TIPS - ADVANCED MIDJOURNEY PROMPTING TECHNIQUE

IMAGE PROMPT

IIn Midjourney, you can use images in your prompts to control how your generated images look, including their composition, style, and colors.. You can use these image prompts alone or with text prompts, allowing you to experiment and create exciting results.

Different from the /blend command, the image prompt can use more than 5 images but to add images to the prompt, you need to type or paste the web address where the image is stored online and the address must end in .png, .jpg, or .gif. The image prompt is at the front of the prompt and then you can add text prompt and parameters after it

If you want to create an image similar but different character to this image:

You can copy the web address of the picture and create an image prompt:

MULTI PROMPT BASICS

Adding a double colon (::) separates each part of the prompt, treating them as independent concepts.

For the prompt "hot dog", all words are considered together, and the Midjourney Bot produces images of bread with sausages.

If the prompt is separated into two parts, "hot::dog", both concepts are considered separately, creating an image of a dog with high temperature background or a dog have high temperature.

In other words, the double colon (::) tells the Midjourney Bot to treat each part of the prompt as an independent concept. This can be useful for creating more complex and nuanced images.

TIPS AND SOURCES FOR MIDJOURNEY PROMPTING

TIPS FOR MIDJOURNEY PROMPTING

• Be clear and specific: The clearer and more specific you are, the better Midjourney can meet your needs. This is the difference between the prompt “Iron Man” and

“Envision a delightful version of Iron Man as if he's been brought to life in a Pixar movie. His iconic armor is now softened and rounded, making for a charming, adorable interpretation of this classic superhero. The signature red and gold hues of his armor remain, bright and appealing in this new style. The artistic style should emulate the works of Pixar artists such as Pete Docter or Daniel Arriaga, filled with soft shapes, vibrant colors, and appealing textures. This is the blend of the bold superhero aesthetic with the heartwarming and cute Pixar touch. The lighting should be soft and diffused, contributing to the sweet, charming Pixar atmosphere. It should highlight Iron Man's softened features, his bright, appealing armor, and the gentle glow of his Arc Reactor. Artistic techniques should include using texture brushes for character rendering to capture the charming Pixar look. Care should be taken to preserve the iconic elements of Iron Man while reimagining them with a cute, appealing Pixar style. Reference: Pixar movies like Toy Story or Monsters, Inc. for the style, and classic Iron Man comics for character details. --ar 16:9”

Using Metaprompt for prompting: You can use ChatGPT as your prompt Generator for Midjourney. If you find it hard to create a very specific prompt to create a picture, you can let CHATGPT do the rest. It will save a lot of time and still generate a good image. For the first example:

Generate an "imagine prompt" that contains a maximum word count of 1,500 words that will be used as input for an AI-based text to image program called MidJourney based on the following parameters: /imagine prompt: [1], [2], [3], [4], [5], [6]

In this prompt, [1] should be replaced with a random subject and [2] should be a short concise description about that subject. Be specific and detailed in your descriptions, using descriptive adjectives and adverbs, a wide range of vocabulary, and sensory language. Provide context and background information about the subject and consider the perspective and point of view of the image. Use metaphors and similes sparingly to help describe abstract or complex concepts in a more concrete and vivid way. Use concrete nouns and active verbs to make your descriptions more specific and dynamic.

[3] should be a short concise description about the environment of the scene. Consider the overall tone and mood of the image, using language that evokes the desired emotions and atmosphere. Describe the setting in vivid, sensory terms, using specific details and adjectives to bring the scene to life.

[4] should be a short concise description about the mood of the scene. Use language that conveys the desired emotions and atmosphere, and consider the overall tone and mood of the image.

[5] should be a short concise description about the atmosphere of the scene. Use descriptive adjectives and adverbs to create a sense of atmosphere that considers the overall tone and mood of the image.

[6] should be a short concise description of the lighting effect including Types of Lights, Types of Displays, Lighting Styles and Techniques, Global Illumination and Shadows. Describe the quality, direction, colour and intensity of the light, and consider how it impacts the mood and atmosphere of the scene. Use specific adjectives and adverbs to convey the desired lighting effect, consider how the light will interact with the subject and environment.

It's important to note that the descriptions in the prompt should be written back to back, separated with commas and spaces, and should not include any line breaks or colons. Do not include any words, phrases or numbers in brackets, and you should always begin the prompt with "/imagine prompt: ".

Be consistent in your use of grammar and avoid using cliches or unnecessary words. Be sure to avoid repeatedly using the same descriptive adjectives and adverbs. Use negative descriptions sparingly, and try to describe what you do want rather than what you don't want. Use figurative language sparingly and ensure that it is appropriate and effective in the context of the prompt. Combine a wide variety of rarely used and common words in your descriptions.

The "imagine prompt" should strictly contain under 1,500 words. Use the end arguments "--c X --s Y --q 2" as a suffix to the prompt, where X is a whole number between 1 and 25, where Y is a whole number between 100 and 1000 if the prompt subject looks better vertically, add "--ar 2:3" before "--c" if the prompt subject looks better horizontally, add "--ar 3:2" before "--c" Please randomize the values of the end arguments format and fixate --q 2. Please do not use double quotation marks or punctuation marks. Please use randomized end suffix format.

And this is the prompt for Midjourney you will get:

/imagine prompt: a grand library, an imposing library teeming with towering bookshelves laden with myriad books from across the ages, rich mahogany accents underline a room abundant with classic and modern literature, somber yet inspiring, the atmosphere is tinged with a scholarly respect and an awe-inspiring silence, soft ambient lights hanging from the intricate ceiling illuminate each corner of the room casting subtle shadows that flicker on the hardbound spines, types of lights are warm LED pendants, lighting style leans towards accent lighting that softly brightens the room without glare, the global illumination is subtle yet clear with gentle shadows enhancing the cozy nooks where one might lose themselves in a world within a world --ar 3:2 --c 17 --s 678 --q 2

With that prompt, Midjourney will generate for you these images:

For the second example, you can use this Metaprompt:

You will be generating prompts for Midjourney, a Generative Adversarial Network (GAN) that can take text and output images. Your goal is to create a prompt that the GAN can use to generate an image. To start, only ask and wait for a subject from the user. The subject can contain an optional parameter '--p' which specifies that the generated image should be a photograph. For example, 'a lone tree in a field --p'. If the '--p' parameter is not entered, then assume the image to be an illustration of some kind.

When an object is submitted, begin the response with the prompt with the start command required by the GAN: '/imagine prompt:'. Next, take the subject and expand on it. For example, if the subject was a lone tree in a field, a description may be: 'A lone tree in a field stands tall with gnarled branches and rugged bark. The surrounding open space provides a sense of peace and tranquility.'

Next, specify an appropriate artist and artistic style, such as 'a watercolor on canvas by Constable'. Multiple artists can be referenced.

Next, describe the lighting effects in the image, including direction, intensity, and color of the light, whether it's natural or artificial, and the source of the light.

Then, describe the artistic techniques used to create the image, including equipment and materials used. Then, include any reference materials that can assist the GAN, such as a movie scene or object. For example, 'reference: the Star Wars movies'.

Finally, decide on an appropriate aspect ratio for the image from 1:1, 1:2, 2:1, 3:2, 2:3, 4:3, 16:9, 3:1, 1:3, or 9:16. Append the aspect ratio prefixed with '--ar' and add it to the end of the prompt, for example: '--ar 16:9'.

Return the prompt in a code box for easy copying. After generating the prompt and displaying it, ask for further instructions in a code box:
N - prompt for next subject
R - regenerate the previous prompt with different words
A - return the exact same prompt but change the artist
M - return the exact same prompt but change the artist and add several other artists. Also change the artistic techniques to match the new artists
O - return the exact same prompt but omit the artists and style
X - return the exact same prompt but change the artist. Choose artists that don't normally match the style of painting
S - random subject
P - change the image to a photograph. Include the manufacturer and model of the camera and lens. Include the aperture, ISO, and shutter speed.
Help - list all commands.

With just a simple subject “flowers”, you can get this prompt from CHATGPT:

/imagine prompt: Flowers, A bouquet of mixed flowers including roses, daisies, and lilies in full bloom with vibrant colors ranging from deep reds to pure whites and yellows, The setting is a rustic wooden table set against a pale blue background, The mood is uplifting and refreshing, akin to a breath of fresh air, Atmosphere is light and cheerful, evoking feelings of a warm spring day, Lighting includes diffused daylight coming from the left, casting soft shadows on the table, while a small focused LED light highlights the texture of the petals, Artistically, the image draws inspiration from the works of Claude Monet and Georgia O'Keeffe, painted using oil on canvas, Techniques include textured brushwork for the flowers and softer blending for the background, Reference: Claude Monet's Water Lilies series, --ar 4:3

And with this second prompt, Midjourney will provide you with these images:

To know more details about Meta prompts, you should check this article: Prompt Engineering Mastery: The Complete Guide from A-Z

SOURCES FOR PROMPTING

If you do not have any idea how to create an image with Midjourney, you can go to these websites and look for some unique images and learn some prompts from them. It also helps you save time and create more pictures for yourself.

• https://lexica.art/

• https://civitai.com/

• https://leonardo.ai/

• https://ideogram.ai/login

MIDJOURNEY PROMPTING CHEAT SHEETS

Here are some Midjourney Cheat Sheets for you!

CONCLUSIONS

In summary, Midjourney is your go-to platform for generating stunning visuals from a few words, called “text prompts”. Whether you're a beginner or an experienced artist, Midjourney can turn your creative ideas into reality.

READ MORE: HOW TO MASTER MIDJOURNEY: FROM A TO Z | PART 1

For more details on how to use Midjourney, please check here.

If you are interested in other topics and how AI is transforming different aspects of our lives, or even in making money using AI with more detailed, step-by-step guidance, you can find our other articles here:

• Top 12 AI Prompting Guides of 2024 - Simplified Overview

• 20 AI Startups Raise $460M+ (Mar 28, 2024 - Apr 5, 2024)*

• Building an Animated YouTube Channel using Free AI tools - Earn Money by Views*

• Creating Consistent Characters with Midjourney V6: A Game-Changer for Creatives

• Guide to Creating Pet Videos that Make Money from Affiliate Links*

_{*indicates a premium content, if any}