Consistent AI Image Characters: The Key to Better Visual Storytelling (Patreon)

i! I’ve always loved working with visual art and storytelling through images. Back in college, I spent way too much time playing around in iMovie, making scrappy little video montages out of my favorite photos. I was definitely not a pro (or well-funded 😂), but I made it work with the tools I had. I always dreamed about creating animations and short films, but it never felt truly accessible — until the last few years with AI art.

Getting to explore this space now, and to share that creative journey with all of you, honestly feels like a dream. So thank you for being part of this experiment and exploration with me.

I’ve had a few requests asking how I get better character consistency and age progression in my short films and music videos, so I wanted to share a few of the tips that have worked well for me. I’ll keep things high-level here, but if you’d like my exact prompts, formulas, and a full walkthrough, you can find those over on my Patreon.

My very first video project, The Patchwork Path (which you can watch on my Patreon for free here), was my first real attempt at keeping a single character consistent across multiple scenes and environments. And, naturally, I chose the most challenging option possible — a quirky, stitched-together character that’s part bear, part patched doll, with lots of very specific details. I fully expected it to fail… but I’m stubborn and don’t back down from a challenge easily. Below, you’ll see the original character and a few examples of what Midjourney generated from my prompts.

So this was my character reference image, which in retrospect I would highly recommend not using an image with so much going in the background. Midjourney really loved adding those flowers to my scenes lol. AI generated image created with Midjourney AI. 

Here are examples of some scenes I made for my little patchwork doll character. You can find exact prompt formulas for different scenes by joining my patreon community here. 

One of the biggest lessons I learned using Midjourney V7 is that character consistency starts with a very clean reference image. Minimal background, no dramatic lighting, and no distracting effects. Just a clear, well-defined version of your character. I then use that same image consistently across all scenes. This works especially well when your character stays the same age and is perfect for illustrated, animated, anime, or stylized video projects.

My reference images for the music video: The End Didn’t Take Us (watchable above)

Age progression takes a slightly different approach. Instead of being ultra-specific, it helps to guide Midjourney gently with descriptive phrases like “visible aging,” “subtle wrinkles,” “deep wrinkles,” “graying hair,” or “age spots,” paired with a general age range such as “late 40s” or “early 50s.” Sometimes you really have to hold its hand a bit , Midjourney is very committed to youth, smooth skin, and freckles 😅.

My favorite aged image of the two pictured above as a couple:

Also, quick honesty moment: I am not a professional video editor. Most of my videos are currently put together using Canva, simply because it’s accessible and gets the job done. If you have favorite tools, apps, or workflows you love (especially beyond Canva), I’d genuinely love to hear about them — feel free to share!

For more in-depth tips, prompt formulas, and a full guide on creating multiple consistent characters in the same scene, join this Patreon as a paid member! It's $5 a month and you can cancel at any time.