ai genshin impact

Okay, folks, let’s dive into my little experiment with “AI Genshin Impact”. I’ve been seeing a lot of buzz around AI-generated art lately, and being a huge Genshin fan, I thought, “Why not combine the two?”

So, I started, like most of us do, by just googling around. I typed in stuff like “AI art generator” and “Genshin Impact art style”. I found a few different tools, some free, some paid. I decided to start with the free options, just to test the waters.

My First Attempts (and Fails)

My first attempts were… well, let’s just say they weren’t exactly wallpaper-worthy. I tried feeding the AI some keywords like “Genshin Impact, anime girl, pyro”. The results? Hilarious, but not what I wanted. I got a lot of weird, distorted faces, and limbs that bent in ways they definitely shouldn’t. It looked like someone threw a bunch of Genshin characters into a blender and hit “frappe.”

I realized pretty quickly that I needed to be way more specific. Generic prompts were just going to get me generic (and messed up) results.

Getting More Specific

I started experimenting with different art styles, and using more specific characteristics. I started adding more detail, like, “Genshin Impact, Hu Tao, detailed face, liyue harbor background, anime, digital art style”.

This was much better! The AI started to understand what I was going for. I got images that actually looked like they could almost be from the game, The style and backgrounds looked closer to the actual game.

Playing with Different Tools

I also experimented with multiple free online AI art generators. Each one gave me slightly different results, even with the same prompts. Some were better at faces, others at backgrounds. It was a lot of trial and error, just seeing what each tool could do.

My “Aha!” Moment

My biggest breakthrough came when I started using image references. Instead of just typing in words, I uploaded some screenshots from Genshin Impact, showing the style and characters I liked. I was blown away by how much this improved the results. The AI could now “see” what I was aiming for, and it made a huge difference.

I was able to create more refined images, which are more alike to the style and the vibe of the game, and it started to feel like i was a real artist of the game!

Still Learning, Still Experimenting

I’m still very much a beginner with this stuff. I’ve learned that it’s all about finding the right tool, using the right prompts, and providing good reference images. It’s a lot of trial and error, but it’s also super fun to see what the AI comes up with. I have to tweak the details and try many times, but finally I could get some really satisfying images. I’m definitely going to keep playing around with this, and maybe even try some of the paid tools to see if they offer even better results.