Welcome to our guide on using MidJourney AI to create stunning images!

What is MidJourney?

MidJourney is a powerful visualization tool that allows users to generate unique images using artificial intelligence.

Whether you're looking to explore portrait ideas or use it as a starting point for overpainting, this tool is a must-have for any creative. In this post, we'll share tips and tricks for getting the best results from MidJourney AI, so you can create impressive images that will wow your friends and family.

Plus, with its fast iteration steps, even beginners will find it easy to get started with MidJourney. So, let's get started!

How can I use MidJourney?

To use MidJourney, users simply type specific prompts into the Discord web community, which guide the program's visualization creation process.

It's important to be clear and specific in your prompts and using commas for soft breaks and double colons (::) for hard breaks can help improve the results.

In order to generate high quality outputs, users can add "high quality" or "4K" to the end of their prompts, and selecting a well-known subject for their portrait can also improve the results.

The V1, V2, V3, and V4 buttons can be used to make variations on the initial output, and emojis can be used to add detail and convey mood.

The upscale button can also be used to improve the quality of the image before making variations.

• U = Upscale: Creates a larger version of the image.
• V = Variation: Creates a new image based on the one you selected.
• Refresh: Get four new images from the same prompt.

Brevity and directness are good. However, it also helps to be specific.

For the most predictable text-to-image portrait outputs, your prompt should basically be a formula of:

Creating Portraits in MidJourney

Generating high-quality portraits can be a challenge for MidJourney AI, but there are ways to improve the results.

One factor that can affect the quality of your output is the subject of the portrait. Choosing a well-known individual can provide the AI with more images to draw inspiration from. The fame level of the subject, or the number of pictures of them available online for the AI to reference, can also impact the quality of the results. It's worth noting that the AI's image source databases currently do not go beyond 2019.

When you see exceptional, unique portraits from MidJourney, it may be because the user generated many different iterations to achieve the desired result. To improve your own outputs, try using the V1, V2, V3, and V4 buttons to make variations on the initial image, and consider tweaking your input by making subtle additions or subtractions to see how it affects the output.

To get the most specific and powerful results from MidJourney's text-to-image AI, it's helpful to be specific in your prompt instructions. While "portrait" is a default setting that will generate a head-and-shoulders image, there are many other options you can choose from to give your portrait a specific look. Some options include

• headshot (commercial headshot, legit headshot, theatrical headshot, etc.)
• selfie / self portrait
• glamour shot
• passport photo
• yearbook photo
• driver’s license photo
• polaroid
• ID photo
• head and shoulders picture
• wanted poster
• environmental portrait
• mugshot

Emojis in MidJourney

Emojis can also be a useful tool for conveying mood and adding background details to your portrait. By using these strategies, you can help MidJourney AI create a portrait that more closely matches your vision.

Additional end tags

One way to get creative with MidJourney AI is to play with different aspect ratios. The default image is a square, but you can change the aspect ratio by adding "--ar" and the desired aspect ratio to the end of your prompt. You can also add images to your prompts to make them more visually interesting.

You can include more than one image and a prompt in your input, and you can also use a suffix with the URL for the image you upload (e.g., "--iw 1.8") to specify the "image weight" of your reference graphic, which can help MidJourney use it as a target. These strategies can help you create more dynamic and visually appealing images with MidJourney AI.

Sometimes, using the initial upscale feature in MidJourney can help you achieve the result you want. However, in other cases, it may add too much detail and detract from the overall quality of your generated image.

To limit this potential issue, you can add "--stop 80" as a parameter at the end of your initial prompt. This can help prevent MidJourney from going too far in its rendering of the image.

After you have a variation you like, you can then upscale it as usual and follow it up with a light upscale and a light upscale max. Including "--stop 80" from the start of the process can help eliminate unnecessary or distracting details like strange nostrils, and adding "symmetrical" to your first prompt input can also help improve the overall quality of your image.

Upscaling Images

Before generating variations of your image with MidJourney AI, it might be helpful to try an upscale first. This extra step can sometimes help you achieve your desired result more quickly than simply making random variations. The upscale feature can add a bit of extra visualization effort to your image, which may help to improve the overall quality and bring it closer to your desired outcome.

Using Timeframes

Consider the time period when using MidJourney to create a portrait. Adding information about the desired time period in your prompt can help the tool make decisions about the subject's hairstyle, clothing, and other details. Want a futuristic look? Add it to your prompt. Want a throwback to the 1980s or even the 1680s? Include that in your prompt as well. Time period information is a useful way to give MidJourney more context and help it generate a portrait that accurately reflects your vision.

Clothing

Clothing can sometimes be a challenge for MidJourney's AI to render accurately, particularly in full body compositions. To avoid odd collars, strange necklines, and other clothing issues, you can use your AI prompt to specify that the subject should be draped in robes or wear loose clothing.

For example, if you wanted to create a portrait of Bryan Cranston as Walter White from Breaking Bad, you might ask MidJourney to dress him in a yellow hazmat suit.

Another way to improve the accuracy of your portrait is to use the word "symmetry" or "symmetrical" in your AI prompt. Additionally, a forthcoming update to MidJourney's default image generation setting is expected to produce images that are less distorted, more colorful, and generally more aesthetically pleasing. In the meantime, you can use this setting by adding "--test" to the end of your prompt.

Getting Ideas for Prompts

Using prompts is a great way to get started with creative AI, as they provide you with a specific goal to aim for.

The community page on MidJourney's website is a good resource for finding prompt ideas and seeing examples of how others have used them to achieve specific looks. For example, you might find a prompt that asks you to create an image that looks like it was painted by Van Gough.

By looking at examples of other people's work, you can get a sense of the style you want to achieve and what tools and techniques you need to use.

Once you've gained some experience using AI to create art, you can start experimenting with different styles and techniques. Collaboration is also an important aspect of using MidJourney AI, as it allows users to learn from and inspire each other.

Additional tools to help you with Prompts

There are several tools available to help you create and edit prompts for use with MidJourney AI. Prompter for MidJourney is a free Google Sheets-based tool created by Shane McGeehan that allows you to save and organize your prompts. PromptMANIA is another free prompt builder that includes a complexity/speed calculator to help prevent overloading the text-to-image bot with too complex prompts, and it can also send prompts directly to Discord without the need for copy-and-pasting.

The MidJourney Styles & Keywords Reference is a Github repository managed by willwulfken that provides information about the differences between past versions of MidJourney and the latest version 3 release.

It's important to proofread your prompts carefully to avoid typos and other errors that can negatively impact the quality of the generated image. MidJourney's text-to-image bot is not as advanced as Google and will not be able to make educated guesses if you misspell something in your prompt, so it's important to take the time to check your work before hitting the enter button.

Copyright Issues

Artificial intelligence (AI) has the potential to create unique and original art, but the question of who owns the copyright for AI-generated art remains controversial and largely unresolved. In the United States, copyright law generally requires that a work be created by a human in order to be eligible for copyright protection. However, AI-generated art is typically created by a machine using algorithms, rather than a human artist. As a result, it is unclear whether AI-generated art qualifies for copyright protection under current law.

Some argue that the creator of the AI algorithm should be considered the author of the resulting art and be entitled to copyright protection. Others argue that the human user who inputs the data and prompts the AI to create the art should be considered the author. Still others argue that the AI itself should be considered the author and be granted copyright protection.

The controversy surrounding AI and copyright has led to a variety of different approaches to the issue. Some artists and companies have sought to assert their ownership over AI-generated art by including language in their terms of service or contracts stating that they own the copyright to any art created using their AI algorithms. Others have sought to license their AI algorithms to users, allowing them to create and own the copyright to the resulting art.

Controversy in the Artistic Community

The art community has also weighed in on the issue, with some arguing that AI-generated art represents a threat to traditional artistic practices and should not be granted copyright protection. Others argue that AI has the potential to expand the boundaries of artistic expression and should be welcomed and encouraged.

Ultimately, the question of who owns the copyright to AI-generated art is likely to be resolved by courts or through legislative action. Until a clear legal framework is established, the controversy surrounding AI and copyright is likely to continue.