An Introduction to Midjourney: For Absolute Beginners & Beyond
Are you ready for endless possibilities? Then join the exploding Cosmic Paint Can of creativity. The Cosmic Paint Can has room for all forms of expression, all earlobe dimensions, pixels of various hues, and no matter what color your shoelaces, thou art welcome.
This post may contain affiliate links, which means I’ll receive a small commission if you purchase through my link at no extra cost to you. All my content is 100% FREE, thanks in advance for your support Please read the full disclosure here.
It’s been let out of the lab, set free from its geeky prison of code and secrecy, and now ordinary people like you and I are playing with the new technological genie that will create almost anything you wish. But you are not limited to 3 wishes, but 3 billion. Are you ready to create unlimited artistic possibilities? This introduction to Midjourney will give you all the essential basics for creating remarkable AI Art.
Imagine the Possibilities, for They Are Endless!
The sheer scale of AI’s growth and astounding abilities continues to defy our imaginative embrace. Is this the democratization of art? Are we on the edge of redefining art?
If 2021 was a big year for artificial intelligence in art, with both established and emerging artists using AI-generated visuals to create unique pieces that captivated viewers around the globe. Then 2022 was the year when AI art truly came to the fore of many people’s thoughts. 2023 and beyond—how exciting to see the new technology continue to astound many!
The land was in turmoil as different nations fought for control. Then one night, a small group of pollen mages descended from the heavens with glorious wings and honey-crystal faces – all of them carrying wisps of living golden pollen, creating beautiful specks resembling snowfall that had immense healing energy within them. – From my series: Return of the Honey Mages
The art in this post is created with a delightful mix of Midjourney AI, Photoshop, and imagination. AI sometimes still struggles to draw hands (LOL, much like human artists), but it will undoubtedly improve over time; these are still the early days of this fantastic technology—to read more on AI tools for home and business, see my Best AI Copywriters Software article. Enjoy!
AI Art is Becoming Rather Famous
If you can imagine it, you can create it!
As Artificial Intelligence (AI) technology continues to become more advanced, so do the tools associated with it. Text-to-Image AI tools are skyrocketing to fame (or infamy, according to others).
This article mainly features references to Midjourney version 4. And now that they have released Midjourney v5 there is a whole new level of awesome to appreciate! While the basic principles are the same, how we write prompts now in v5 is more like natural language, rather than a string of words separated by commas.
Global use and adoption will continue to rise as users continue to produce phenomenal images unheard of a year ago. Some even consider it one of the world’s most innovative and exhilarating art forms.
AI Art harnesses algorithms and neural networks to create visually stunning works of art that can sometimes appear more lifelike than traditional artwork, thus the reason (among others) for a vocal anti-AI-art crowd.
The golden blimp was a sight to behold as it floated over the streets of Weathersteam. The citizens held their breath as they expected to witness something amazing, and it did not disappoint. As it approached, they saw the royal guards mounted atop Pegasi accompanied by a squadron of Volarian guards leading the way – an impressive spectacle! – From my series: The Battle for Weathersteam
It also allows artists more freedom to explore new aesthetics not possible in pre-existing mediums like painting or photography. Such flexibility has opened up AI Art to a wide range of styles, from hyperrealistic images to abstract shapes and patterns unlike anything seen before.
While there are many different options available, four stand out from the rest: DALL-E 2, Stable Diffusion, Night Cafe, and Midjourney.
DALL-E 2 is a deep learning model that provides users with exceptional results in terms of image generation from text-based data. It can generate detailed images via a CLIP (Contrastive Language-Image Pre-training) process, also created by OpenAI, using simple text descriptions that accurately represent what is described in the input.
Stable Diffusion is an AI platform powered by Latent Diffusion Model (LDM) and designed for use in natural language processing tasks such as summarizing texts and recognizing patterns within them.
NightCafe Creator is another fantastic app for modern art enthusiasts. Combining the power of advanced AI algorithms with an easy-to-use interface, it offers multiple methods of creating AI artwork quickly and simply. With neural style transfer technology, you can take a photo and turn it into an impressive piece of digital art. What’s more, NightCafe Creator is incredibly affordable and accessible to everyone.
Mystic crests of the cosmos glowed in a gentle iridescence of wonderous fire, calling out to those who sought adventure and exploration. Many brave voyagers passed under their majestic presence and were instantly transported away, across time and space, to realms unknown and unimaginable throughout the known universe. – From my series: Crests of the Cosmos
My main experience has been with Midjourney as a paid member (along with dabbling in BlueWillow and NightCafe). If you don’t have any experience creating art with Midjourney yet, this brief article will get you started and might just save you some initial headaches as you get up to speed.
Creating art with Midjourney AI (and yes, it is art) is an exciting and entertaining experience that can be enjoyed at any level of expertise. Whether you’re a seasoned artist or just getting started, the Midjourney AI software offers a variety of tools and features to help you get creative without having to worry about too much technical know-how.
The Cat is (Definitely) Out of the Bag
If AI is now ‘out of the bag’ and in the public conversation, then I might as well let the cat out of the bag also—I’m personally very excited about this incredibly innovative technology. Some argue that the release of a public-domain text-to-image AI that is capable of creating artworks is going to disrupt many lives or steal from artists or create mere lifeless mimics etc.,
After a lot of reading and research, I really do not think this is, or will be, the case (and I have been a digital artist and dabbler for many years from the early versions of Photoshop), and the advent of Midjourney, especially version 4 (and wow and more wow with v5 now released), has amplified this excitement.
James Scott, commenting on an ArtStation blog article, says:
The genie is out of the bottle. I was a professional airbrush artist in the ’90s and early 2000s. When the vinyl wraps were introduced, my life was turned upside down. 15 years of perfecting my craft, gone. What did I do? I bought a computer and learned 3D modeling and web graphics; long story short, every industry in the past 150 years has been disrupted by new emerging technology.
The dense forest of Zinar was known for its mysterious beauty and tranquility as if time itself had stopped here. Few knew, however, that a gift from the distant past lay within these ancient woods. – From my series: TARDIS Trees
Minjourney AI technology is not going away and has reached a point where it can create stunning and vivid artworks with only minimal input, and this is leading some to worry about the impact on traditional art forms.
And just in case you think I am unaware of some of the issues that are still associated with Midjoyrney- take a look here.
However, while many are concerned about how this type of AI could potentially damage their profession and livelihoods, others remain cautiously optimistic about its potential. I am in the latter group.
While it’s true that this new technology could lead to profound changes in the way we create and define artwork, it also comes with the promise of greater accessibility for those who want to make something creative or artistic but lack the skill or training. It could open up avenues for people from all walks of life to express themselves through artwork that previously might have been out of reach for them.
What and Where is Midjourney?
At its simplest—it is a bot on a Discord server.
You give the bot commands on Midjourney’s Discord server, and it does what you tell it to do.
To some, the interface on Discord can be a little tricky, but once you get the hang of it, you will be creating masterpieces in no time. Discord also makes exploring new techniques and styles easy while honing your existing ‘prompting’ skillset.
You are allowed to generate 25 images for free to try the software, and all the prompts are public in a ‘Newbie’ channel thread. So that is where to learn.
But when you subscribe, you can access the AI bot via your Discord direct messages and locate your generated artwork a LOT EASIER! This is way better than scrolling through multitudes of posts in public threads trying to locate your recent masterpiece.
Midjourney itself is an independent research lab that produces the artificial intelligence program under the same name that creates images from textual descriptions.
She identified herself as the warden of the outer worlds, her aura brave and fearless. Everywhere she went, a golden-blue light illuminated her path to guide those who encountered her on their journey of self-discovery. – From my series: Psytrance Wardens
What is Discord, And What is it Used For?
To create art with Midjourney, you need to join Discord and click the ‘Join the Beta’ button on Midjourney.com.
Discord is a free application that allows users to communicate with each other in real-time. It was originally released in 2015 and has become one of the most popular communication tools for gamers, streamers, and communities. Discord is designed to be user-friendly and provides users with access to many features such as text chat, voice chat, video chat, servers, bots, custom emojis, and more.
Whether you’re an aspiring artist, an experienced one, or just curious about the latest in AI technology, the Midjourney Discord channel is the perfect place to connect with like-minded artists. The server offers plenty of opportunities to share artwork with other members and discuss different techniques, writing prompts, and styles used in digital art.
Plus, you can take part in activities such as creating unique AI-generated art via daily prompts or exploring new digital creative forms that help bring your ideas to life.
You’ll have access to both real-time discussion threads and more structured forums dedicated to curating artwork from diverse backgrounds. Via Midjourney’s Discord channel, showcase your work while gaining valuable feedback from other artists around the world.
Mysteriously, they were always there—the Archons of the Eternal, invisible lords of realms unseen. They blessed us with their dangerous yet graceful presence without fail, and I had no idea why or how these creatures managed to exist in this world without being seen by anyone. – From my series: Archons Eternal
How to Use Midjourney: Very First Steps
I believe there is an ‘art and science’ to the generator prompt, and it is a skill to learn in and of itself.
Midjourney AI eliminates the guesswork of creating digital artwork quickly and easily. Maybe the prompts will be the secret sauce of the future that everyone will guard and protect. This may separate the dabblers from those serious about using the platform for serious art production. Maybe the prompts themselves will become a form of income? Who knows?
To begin creating stunning artwork with Midjourney, you must use the
/imagine prompt within Discord and type in what you want the AI to create – whether it be a landscape, an abstract piece, a lavender dragon eating a small house, or something more specific, like a bubble taking a walk in the snow—there are literally a million ways to create!
So get comfortable, and let’s begin:
The default option is to create 4 square images with each prompt request sent to the AI; the numbers go from 1 to 4 and stand for their positioning.
The amazing thing about this generative technology is that it keeps learning from each image that it creates, meaning that every time you use it, you will get better results.
I have found this to be the case when making self-portraits from seed images. At first, they look nothing like me, but then they become increasingly creative! So give it a few tries before throwing your hands up in the air.
The generated images vary in complexity depending on how much time and effort you put into them, but even if you’re just getting started, they can still look more than impressive.
When I first started experimenting with Midjourney, I used it for almost 11hrs straight (I am forever thankful for a good Secret Lab chair) to learn the nuances of the system, such as lighting effects, material properties, design styles, art styles, and color properties.
The Angelica Model VI glowed in the evening sky, a sight that should have been impossible given the night’s darkness. But there it was: an iron-clad warrior standing tall and proud on the edge of town, watching protectively over every citizen who walked beneath its protective wings. – From my series: Angelicas
Writing Prompts and Basic Tips for Successful Usage
Here is my little cheat sheet and some basic keyword guidance for improving image output by crafting better prompts:
I want a Landscape Image
For a landscape image: right at the end of a prompt, add
--ar 3:2 The ar stands for aspect ratio. You can also use
I want a Portrait Image
For a portrait image: right at the end of a prompt, add
I want a Larger Image
Upscale an image by pressing a U1, U2, U3, or U4 option underneath the returned images from your prompt.
Remember, the images are always returned in this 2 x 2 grid:
I want some variation of the images I just created
Create variations of the results by pressing the V1, V2, V3, and V4 options underneath the returned images from your prompt.
I want a new set of images without typing the prompt all over again
Simply click on the ‘reroll’ button, and you will ask the AI to generate a new set of images based on the previous prompt.
I want to seed the bot with my own image
This is fun for making self-portraits or caricatures (or anything else). Seed an image with your own image file by making sure the image is first on a server somewhere. I use pCloud for this as a lifetime member as I then have an easy record of all the images I use, or you can simply upload your image directly into Discord.
Pro Tip: I have found portrait mode
--ar 2:3 sometimes works a little better for self-portraits in Midjourney and it may take at least several attempts as we train the AI to get what you want.
Then get the direct URL to the image and, in the prompt, add it right at the beginning, then add the rest of your prompt text. On Discord simply right-click the image. (see here for a little more explanation)
I want to mix two images
Blend images together by using the
/blend command instead of the
/imagine command. You will need the direct URLs again for each image.
Argh, My Trial has ended!
Burned through all the free 25 credits? That was fast! Now you need to subscribe, yes? Type
/subscribe to have the bot send you a direct link so you can subscribe to Midjourney. Read a little more about what subscription means.
I want to switch GPU modes
Switch between Fast GPU Mode and Relaxed GPU Mode with
/relaxed These different modes are explained more below.
I need some basic prompt help
Try these words in prompts to get really creative: wizard, goblin, angel, emperor, necromancer, greek gods, urban, queen, spirit, TARDIS, temple, ocean, car, robot, mountain, deck of cards.
Try describing a style you would like to see generated: “a cyberpunk dragon,” “a mountain from Middle Earth,” “a cat astronaut.”
Experiment with lighting styles with words like Cinematic Lighting, Illumination, Studio Lighting, Spot Lighting, Frontlight, Backlight, Strobe Lights, etc.
Use some of the color model words for interesting results: RGB, sRGB, CMYK, HDR, Adobe RGB, and Pantone.
Type in artist styles to get a particular style returned: “a landscape in the style of Albert Bierstadt,” “a poppy field in the style of Claude Monet,” “an 18th-century princess in the style of Elisabeth-Louise Vigee Le.” For a detailed list of possible Midjourney artist styles, take a look at 130 AI painting techniques.
I need better-quality images
How much rendering quality time do you want to spend? Use
--q <.25, .5, 1, or 2> at the end of a prompt. The default value is 1. Higher values cost more and lower values cost less.
I need greater style variety
This Midjourney Bot has been trained to produce images that favor artistic color, composition, and forms. The
--s parameter influences how strongly this training is applied. The default is 100. With ranges for version 4 being 0 – 1000.
I need to know what these TWO COLONS ‘::‘ mean
You will see in prompts the addition of two colons to prompts—such as
big red car::5 blue car::1 --ar 3:2 These are called multi-prompts and creators use them so the Midjourney Bot considers two or more separate concepts individually. You can also give weight to the concepts by adding numbers after them, this assigns relative importance to that part of the prompt. The higher the number, the more important the subject.
Most times commas
red car, white vespa, on a road and natural language
a red car and a white vespa on a road work just as well.
I need A Little more chaos!
Do you want more unusual and unexpected results? Try a little bit of chaos! Use the
--chaos parameter. It accepts values 0–100, with 0 being the default.
I want to use my own image
Use your own images to influence a returned composition, style, and color. The image prompt URL goes at the front of a prompt.
And the images created from the above example are seen below. Notice the subject’s shape, general color, and style, plus the castle library background, are all there but different. So now you are on your way to being a master!
Keep a record of your best prompt ideas in a notebook (either online or in a physical notebook). I use Notion and have pages of prompt ideas collected with a sample image next to each one of the outputs that can be expected.
Try these prompts to kickstart your AI fun and experimentation:
Need more help?
Always keep learning and experimenting and if you want a very detailed and well-thought-out guide on all the artistic possibilities of Midjourney – take a look at the excellent GitHub resource: Midjourney Styles and Keywords Reference
The older clowns had a sweet air of wisdom about them. When you talked to them, it was like being taught by the most gentle grandfather teacher in the world – their words and advice were so kind, almost as if they came from some higher knowledge.. – From my series: Clowns Can Be Kind
Photoshop Skills Are Still Needed
Photoshop editing skills continue to be a necessary tool for photographers and digital retouchers alike. Editing images with Adobe Photoshop can often help to bring out the best in an image, and there is no substitute for having a good understanding of how this powerful software works. However, Midjourney images require a special set of skills that go beyond just basic photo editing.
Midjourney images, which generally give you amazing concepts and art are kinda in between stages of production or development and require more than just simple cropping and color correction. These intermediate images may need additional touch-ups such as skin smoothing, removal of 29 extra fingers, weird object removal, or facial restoration work that goes beyond what regular photo editing programs can offer.
Most of the nagging issues will likely be resolved in an upcoming release (hopefully). But these are some of the issues you are likely to encounter:
Midjourney’s AI-generated art is still super impressive, but not perfect–so taking some time to learn some Photoshop skills will be a great advantage to producing super AI art.
The devils were writhing in the bathtub, their scaly skin reflecting sickly blue and purple hues under the harsh lights. The sound of their cackles was downright sinister as they dipped themselves into the murky water, And they never seemed to get any cleaner no matter how much time passed. – From my series: Bathtub Devils
Benefits of Midjourney
Midjourney is truly a revolutionary platform where users can create stunning visuals with just one simple step—inputting text.
Through Midjourney’s service, AI art enthusiasts can generate perfect images tailored to their preferences and specific needs. The platform leverages its advanced natural language processing capabilities and wide range of algorithms to produce rich visual content in any desired style.
Connect with other AI-generated art lovers via the Discord channel or via Facebook groups such as Ai Art Universe
As a web and graphic designer, I am always looking for a source of new, fresh, and original images about ‘anything’. Let’s be honest; stock images are nice, but they are really starting to look a little tired and repeated ALL OVER the internet. Google knows this, and it may help SEO and ranking if we include new original artwork in your posts, along with well-written, engaging content!
As seen below, I needed to create a particular style for a client on a project and created the following with Midjourney, with post-processing and editing in Adobe Photoshop.
You can also benefit from prototyping original ideas for layout and design from websites, logos, menus, UI/UX, etc. For example:
Are you looking for the perfect way to showcase your new business cards and make a great first impression? With Midjourney, you can create professionally designed and meticulously detailed business card ideas in just minutes.
You love art, so now you can not only explore the Midjourney AI to increase your understanding of style and design but also book cover design. I have designed many of these in the past, but this takes it all to a whole new level. What an amazing garden to play around in as you test all those little nuances and styles you have not quite experimented with yet.
Theoretically, you now have unlimited images on any subject imaginable at your fingertips.
So you want to share interesting content? Then Midjourney can provide that in a way no other site can. So why not enhance Instagram, Twitter, and Facebook posts with stunning and captivating artwork?
Possible monetization of original works of art via print production and other merchandising.
You can use an AI-upscaling service online to increase your image to a higher resolution so that you can print it as a large poster or print. I use ON1 Resize AI, and it can produce very large images. This super-resolution AI image resizer will allow any photographer to enlarge photos while quickly maintaining and recovering incredible detail and sharpness.
Aria shimmered with a grace that could only come from a kind and gentle soul. All who encountered her were imbued with joy, hope, and truth simply by being in her presence. Her beauty radiated like the light of a thousand stars, capturing those around her in profound moments of awe. – From my series: Watercolor Maidens
Experience the Magic: Become a Member of Midjourney
You will become addicted to Midjourney and probably burn through the free/trial hours in about an hour or so!
The solution to your newfound addiction – become a paying member!
Midjourney has three subscription tiers.
Pay month-to-month or for the entire year for a 20% discount. Each subscription plan includes access to the Midjourney member gallery, the official Discord, general commercial usage terms, and more.
You can upgrade or downgrade plans at any time.
Payments are only via Stripe support cards like Mastercard, VISA, American Express, and others.
I think the best value is the Standard Plan, as the Basic Plan is still fairly limited. The good thing with Standard and Pro Plans is that they offer a generous Fast GPU Time and unlimited creations via a Relaxed GPU Time per month.
How to Subscribe?
/subscribe command once you are logged onto Discord to generate a personal link to the subscription page. Or, go to Midjourney.com/accounts. You can also select Manage Sub from the sidebar while logged into the Midjourney website.
You may cancel your subscription anytime, with the cancellation being effective at the end of the current billing cycle.
The Golden Goddess of Aerethia was a gracious but mysterious entity. She held great power over the realms beyond our understanding and safeguarded the most sacred TimeSpire Gate. – From my series: Timespire Gate
What is Fast and Relaxed GPU time?
With the help of some of the world’s most advanced processing chips—Graphics Processing Units (GPUs), Midjourney can interpret and process each prompt from users quickly and efficiently to create those unique works of art. When you purchase a subscription to Midjourney, you are purchasing time on these GPUs. Now you can really bring your creativity to the fore.
GPU Processing Times:
I have used both Fast and Relaxed modes and have noticed only a fraction more time taken to generate images.
I use Fast Mode for upscaling if I want a quick result. Otherwise, I am in Relaxed Mode.
Just make sure you use all your allotted Fast GPU time each month, as it does not roll over. You can always check your account status by typing the command
/info to view information about your profile.
As I mentioned above, Midjourney runs on GPU minutes/hours, and every user, free or subscribed, is allocated a certain amount of GPU time for their subscription level. One hour of GPU time will allow you to generate or upscale roughly 60 images or about 200 image variations.
What is Fast GPU time? A highest priority processing mode on the bot GPU.
What is Relaxed GPU time? A lower priority processing mode on the bot GPU. Midjourney uses a queue system for its relaxed mode. Wait times for Relax are dynamic but generally range between 0–10 minutes per job, although I have never had to wait more than 40 seconds.
In wintertime, the Piddlypot villages come alive with frosty festivities. The streets are filled with laughter as snow falls over little shops and homes like powdered sugar in a bakery window. – From my series: Seasons Over Piddlypot
Its a Gamechanger
The roll-out and availability of Midjourney, backed by high-profile tech players, signals an exciting and potentially game-changing moment in the world of digital art. Midjourney executive is David Holz, who was previously the founder of Leap Motion and a researcher at NASA. On the advisory board includes Jim Keller, Lead Silicon at Apple, AMD, Tesla, and Intel; Bill Warner, inventor of non-linear video editing; Nat Friedman, the CEO at Github; and Philip Rosedale, the CTO of RealNetworks.
While it is sure to create some disruption in the art world, it is also a sign of the times and a testament to our ever-evolving technologies on the planet. AI-generated images are no longer just a curiosity but a viable tool for creative expression that will continue to grow in potential and capability.
We are witnessing the birth of a new and democratic artistic medium, and I agree with what Stephen March said in the Atlantic: “The art that is half-born is the most exciting. Nobody knows what a new art form can be until somebody figures out what it is. Figuring out what AI art is will be tremendously difficult and tremendously joyful. Let’s start now.”
And if you’re feeling overcome by generosity and want to help the universe align on the side of intergalactic peace and harmony, you could buy me a coffee at www.buymeacoffee.com/moonpixel. Your support would greatly help in keeping me motivated and writing articles that you love.
Steve Roberts is a full-time blogger, digital marketer, and freelance designer who also helps young Jedis make informed lightsaber decisions. Join Steve on this incredible journey, learning how to scale your business and entrepreneurial vision. His clients include organizations and businesses from Europe, Asia, Australia, and the USA.