Creating graphic stories like comics, graphic novels, children’s books, and even short animations has never been more accessible, thanks to AI programs like Midjourney. This technology empowers anyone with a story to tell to produce rich visuals, even if they lack traditional artistic skills. However, diving into AI-powered storytelling can be more challenging than it seems, and most tutorials available on YouTube lack the depth necessary for storytellers.
This guide aims to fill that gap by focusing on the features of Midjourney that benefit storytellers the most.
Setting Up Midjourney
To get started with Midjourney, you’ll need a Discord account. If you don’t have one, create it first. Once your Discord and Midjourney accounts are ready, log in to the Midjourney website, click on your account name, and manage subscriptions. Midjourney offers four subscription plans. For beginners, the Standard plan is usually sufficient, providing unlimited slow generations and the ability to upgrade later if needed.
Next, navigate to Midjourney’s Discord page by clicking the World icon. Here, we’ll create our own server to keep our assets private. Select “Create My Own,” choose “For me and friends,” and give it any name. Go back to Midjourney, find the Midjourney bot, and select “Add to Server,” then choose the server you just created. Now, you’re ready to generate images in your server without interruption.
Before we generate images, let’s configure your settings. Type /settings and hit Enter. Set Stylize to medium, turn on Remix mode, High Variation Mode, and Fast Mode. Now, you’re ready to start creating.
Midjourney Basics
For storytellers, generating images usually involves prompting a subject, a modifier, and an art style. The subject is the main element, the modifier describes what the subject looks like and what it’s doing, and the art style defines the overall look. For example, a prompt might look like this:
/imagine man holding a sword, 8-bit
Here, “man” is the subject, “holding a sword” is the modifier, and “8-bit” is the art style. Once you generate an image, you’ll see U & V buttons underneath. U stands for upscale, which enlarges the image, and V stands for variation, creating similar images based on the selected one.
Upscaling and Inpainting
After generating images, upscaling introduces new buttons like Vary (Region), also known as inPainting. This feature allows you to change specific sections of an image without starting from scratch. For example, if you want to change a sword into an axe, you can use the lasso tool to paint over the sword, modify the prompt, and generate the change.
Storyboarding
Storyboards are essential for visualizing the sequence of scenes in a story. They typically come in black-and-white drawings, but feel free to change the art style. Use the variation technique to maintain consistency across scenes. For example, if your first thumbnail is close to your vision but lacks detail, generate variations until you find the perfect one.
Creating a Storyboard
- Generate the first thumbnail.
- If it’s not perfect, generate variations.
- Select the best variation and upscale it.
- Repeat the process for subsequent thumbnails.
The final output should resemble a cohesive sequence that lays the foundation for more advanced storytelling workflows.
Character Design
Creating characters involves a specific prompting formula. For instance, if you want to illustrate a medieval fantasy story, your prompt might look like this:
/imagine fullbody portrait of a knight, wearing armor, standing in a forest, ar 2:3
This prompt includes the subject (“fullbody portrait of a knight”), modifiers (appearance and clothing), and an aspect ratio (ar 2:3). The sweet spot for prompts is around 15-35 words to ensure accuracy.
Once you have your character, you can create different poses using the character reference parameter (--cref). This allows you to reduce the description of the appearance and focus more on the activity.
Scene Composition
With character poses ready, you can now create a scene. Use the pan and zoom features to adjust the canvas and add new parts to the image. For example, if you want to extend the image to show more of the character’s body, click the down arrow to pan downward.
Zoom out generates new regions in all directions, expanding the scene. For instance, you might set the zoom to 1.5x to increase the canvas size. Modify the prompt when using these features to guide the AI in generating the desired elements.
Generating a Background
Generating a background that matches your scene is crucial. Use the style reference parameter (--sref) to ensure consistency with your character designs. Set the aspect ratio to match your layout, for example, 16:9 for a widescreen format.
If the background is too dark or not to your liking, use the stylize parameter (--s) to make it more artistic. Adjust the stylize value between 0 and 1000, depending on how much you want to prioritize artistry over accuracy. For instance, setting it to 350 or 500 can enhance the background’s appearance.
Advanced Techniques
Discover Infinite Art Styles
To explore various art styles, use the Style Tuner by typing /tune followed by your prompt. Choose the number of style directions you want. For example, selecting 16 style directions can provide a broad range of artistic variations. Once your Style Tune is ready, you can combine multiple styles or select individual ones to use in your prompt.
Using Style and Character References
Combine style and character references to maintain consistency across different characters and scenes. Use the style reference parameter (--sref) with your selected art style and the character reference parameter (--cref) to generate new poses and scenes. This approach ensures that all elements in your story have a unified look.
Conclusion
Patience and persistence are key when using AI image generation for storytelling. While the process can be challenging, especially for beginners, there’s always a way to achieve the results you want. Keep experimenting with different prompts, parameters, and techniques, and continue learning to improve your skills. With practice, you’ll be able to create compelling visual stories using Midjourney and other AI tools.
