ConveYour's EIGHT Steps to Create Superior Stock Images With Midjourney AI

Related Article: The Power of Using Midjourney AI in Your Training Content

Take a look at the images below.

Would you have been able to guess that they were entirely created by artificial intelligence? Probably not. Specifically, they were created by prompting the Midjourney generative AI with no manual edits done. Clearly, the future is now - AI can create images in a variety of different styles with stunning accuracy. Photorealistic portraits, stylized cityscapes, beautiful vistas…the floodgates have officially opened and the possibilities are limitless.

Of course, just typing anything into the MidJourney prompter is not likely to get you the results that you want. As with any artificial intelligence, the prompts that you provide and practices that you follow will determine the quality of your results. Don't worry, though - we're going to break down some of these best practices so you can hit the ground running creating incredible AI-generated stock images for your business.

Are you a sales leader struggling to create useful stock art for your team? We wrote a Midjourney guide SPECIFICALLY for people working on sales-related stock art. Check it out here!

Using AI service like MidJourney is all about prompts. Since computers are very literal, the more detailed and specific you can be about the image the happier you are going to be with the results. You need to come prepared with ideas around the subject of the image, the style, the mood, lighting, colors, and any specific details that you need included. It is also important to have an aspect ratio in mind so that the generated image fits where you want it to go. You can easily incorporate aspect ratios into your prompt by ending your query with the dimensions you want, such as square (1:1), widescreen (16:9), or portrait (3:4).

If you are just getting your feet wet with AI prompts, start with some simple, direct, and clear instructions and take note of how MidJourney interprets them. Even if you've used generative AI before, each piece of software is different so it’s a great idea to get a feel for how they work before seriously diving in. Take note of keywords used or styles you requested and how MidJourney incorporates them into the imagery. Focus on keeping your prompts short and clear.

Once you start feeling more comfortable with how MidJourney operates, you can start adding different parameters and details to the generated images. You can add language around style, detail, vividness, or color profile.

Note that there is a difference between prompts with a lot of detail and prompts with unnecessary words. Here is an example directly from Midjourney’s documentation:

Instead of: “Show me a picture of lots of blooming California poppies, make them bright, vibrant orange, and draw them in an illustrated style with colored pencils”

Say: “Bright orange California poppies drawn with colored pencils.”

Both prompts convey essentially the same information, but one was written much more clearly. This is always going to be preferable when prompting AI. Of course, if it was important to make sure there were still “lots” of poppies, that can still easily be added to the revised prompt.

If the AI isn’t successful at capturing what you want on the first try, that’s okay. Try to identify what you would like changed and add (or remove) language from your prompt as needed. Iterating on images is an expected part of AI image generation as well as a great way to learn more about the power of your prompts and how the AI interprets them.

Still think your prompt is too long? The /shorten command within Midjourney will analyze a given prompt and make it more succinct.

You may want your images done in the style of a particular artist. This can create issues for a generative AI for various reasons. A better practice is to figure out how to describe the artist’s style, and to use that. For example, instead of naming a particular painter, you could request an image in an “Impressionist style.” You can also use emotional cues (optimistic, dreamy, melancholy, etc) to nudge the AI in the right direction.

Still not getting the results you want? You can always ask Midjourney to describe an existing image for you. Doing so will generate four prompts which you can use as guidelines to describe the style in a way that Midjourney will receive properly.

Once your images have been generated, you have a few different options. Using Midjourney AI, you can select the image that you like, and generate an upscaled copy of it. From there, you can re-generate variations of that same image, helping you to zero in on what you want. Midjourney also gives you the ability to “Zoom Out” to varying degrees, expanding the barriers of your image and explore what’s happening just outside of frame.

Midjourney is new, exciting technology, and the community of users is growing by the day. Spending some time on the Midjourney dashboard will reveal the limitless creativity that people have been pouring into this software over the last few months. There are also a variety of message boards, forums, documents, and subreddits available online to discuss the software and your creations. In fact, Midjourney itself currently exists on Discord, and the Discord server used to generate images has a built-in community support channel. Don’t be afraid to ask questions!

Keep in mind that AI image generation technology is very new, and guidelines around ethical and legal considerations are still being worked out. Midjourney will generally not stop you from referencing copyrighted materials in your prompts, but it is important to consider the potential future consequences of doing so. Again, regulations have not been fully hammered out in this area, and you do not want to open yourself up to copyright issues in the future.

Similarly, remember to follow ethical guidelines when generating imagery with Midjourney or any other AI service. Generated imagery is available to viewing by the wider community, so caution is especially important when creating imagery of people or elements that could be considered sensitive or controversial.

Midjourney, and really any other generative AI, has different plans available to customers with different usage limits. These are subject to change over time as the technology develops. It can be easy to get carried away, and plan limits are generous, but it’s important to keep usage limits in mind to avoid unexpected costs - especially if you share a plan with your team.

Check out our article about creating superior business stock art with Midjourney AI for additional tips.

It really does feel like the future - the magic of AI now allows you to quickly generate stunning, imaginative, and useful visuals that stand out, and not in that “bad AI” type of way. By following the tips and practices outlined above, you and your team can easily become pros at prompt-writing and image-generating. Experimentation and constant learning are a must as we move into an even more mind-blowing future.

ConveYour is an all-in-one software platform that combines recruit organization, onboarding automation, training creation, and SMS messaging.

It’s astonishingly easy to use whether you have 100 or 10,000 reps.

‍Click here to learn more.