First Step of Graphics | Crystalled Bullets - Dev Diary

This article is a transcript of The Crystalled Bullets adjusted as an IndieDB/Itchio article.

Join the Discord Server, and come hang out with us!

People say that graphics aren't important, but You can't deny that it is one of the key factors that give the first big impression of the game when people see the screenshots or video footage of the game. For example, If you saw a screenshot of the game that was poorly drawn with MSPaint, or organized from free stock assets, you probably speculate that the developer put a bugger all in terms of the effort they put into it to make the game worth playing.

So, Let's talk about my game "The Crystalled Bullets"'s graphics, how I chose the art style, and how I expressed the uniqueness in terms of the visual of the game. I choose this topic to talk about in this series first. Because A: My other key elements of the game, like the core gameplay loop or unique game mechanics, have room to polish, and B: While the quality of sprites in the game could be improved over time, art style won't change that much in progress through the development, and C: is the programming language invented in 1972. but I digress.

There are multiple styles in terms of visuals. In indie games developed by small teams or individuals, the most common style is pixel 2D art style. Because It requires the least amount of effort to make it look anywhere decent. I don't think it's a bad thing. because as less effort goes into graphics, more and more could go to the better gameplay design. The other popular style is hand-drawn art styles. notable examples are Spiritfarer, Castle Crashers, and Cuphead. and there's a game that combines both pixel art and hand-drawn art. OMORI is one of the popular examples of that term. and I think the latter method is much more common. Because there's more additional effort is required than pixel art does. In some cases, it needs to scale those sprites depending on the game world, and sometimes could affect performance as well. Especially VRAM in GPU. And more recently, compared to previously mentioned methods. Games with Low-Poly 3D Models are becoming common in indie spaces. For example: Compound VR, Dusk, and Neon Rain. This could apply to the same reason why pixel art is popular, It's easier to reach the quality of "not bad!" And, I think I covered most of the popular styles. In Indie games, at least. It's not a significant number, but there are some cell-shaded ones, and realistic graphics ones, but both of them can be difficult to make. Especially for small independent teams, or solo developers like me.

Anyway, In the Crystalled Bullets, I used a hand-drawn art style with 3D environments. Firstly, I want to make my game stand out in this world of indie games where the majority of it is drawn with pixel art. Secondly, I really can't be bothered with understanding how to use Blender to do low-poly 3D style. Thirdly, I wanted to cope with NOT to convince my Pen Tablet is a bloody waste of money! So, I decided art style, There are still some problems. As I previously mentioned, It takes a lot of time to draw each sprite. So, I had to make some corner cuts to save time. As a great man said, "Work smarter. Not harder." I have to do not only drawings, but some coding, audio design, animations, and all other busy works to make my game functional.

For example, Take a look at these weapon view model sprites. then you may notice that this glove is weirdly detailed. To tell the truth, I used an actual gloves photo instead of drawing it by myself! I'm not very good as far as an artist goes, When I try to draw hands, It usually ends up looking like spaghetti.

 In the other corner cut example is this female mage and male soldier for enemies. If you look at them carefully, you'll notice that the hands are re-used. but even with this and all of the other corner cuts, It still takes a lot of bloody time to draw. And you may thought, "What about AI to save time?" Three Problem viewers, First: Currently It has multiple issues of Copyrights. Second: It usually ends up very bland, and soulless looking Third: I DON'T KNOW HOW TO USE THAT STUFF!

Like, I tried to run Stable Diffusion on my PC, and I don't know how to use those thing! So I'd stuck with drawing, like a caveman! grrrrrr! Since It's a cutting-edge technique, It's better to learn for sure, but I think it's not too late to wait for tech to get stable. Since "cutting-edge" has the potential to stab me like a knife.

Still, I had to learn Blender to make the map slightly look nicer. all of those lanterns, and trees, (probably, I should learn more about Blender to improve all of those 3d Models' look.)but most of the time, I had to deal with drawing. and I spent the majority of development time drawing all of those sprites. I think it's good timing to investigate the technical side of stuff. I draw all of those enemy sprites on 4K resolution but If the game loads all of those, the file size of the game will be unreasonably big, and lead to experience some stutters for loading all of those sprites. So, I had to crash down all of those sprites resolution to 20% of 4K. 796 to 700. so, the file size can be reasonable, and performance can stay steady. And you may think, "Why don't you draw them by that resolution? isn't it a waste of time to change the resolution of the image, every time?" Well, Not at All!

Have you ever heard of "Downsampling"? Oh yes! I'm a tech nerd who knows DLSS, FSR, PBR, and many other words with 3 big letters related to tech. so I can bore another person when I'm talking! Ha Ha!

 Anyway, Downsampling is, in short: It's a technique that prepares the image on a much higher resolution, often the resolution higher than the monitor resolution, then displays it on a much lower resolution. and the result is jaggy less smooth edges!

So, as far as I can tell with the current build of the game, how do I feel about all of these? The answer from me, Is... Yes? On the one hand, Yes My Game's look is distinguishably different enough from other game's visuals. And yeah, I succeded in that term. on the other hand, I probably should spent more time on the core gameplay loop's design first. Since I want to make my game another selling point is that unique gameplay uses guns with special abilities and guns and enemies' elementals. Considering Hi-Fi Rush and Portal made core gameplay loop first, then unique assets.

Well, for my excuse, I had to make all of the assets first, so I had to get some idea of how the game's visual was gonna look like, and I don't wanna waste time making placeholder assets. But the most ironic thing about all this is that later I had to redraw several assets of enemies and guns! so yeah, I ended up wasting some time with ugly assets.

Moral of this story: Just use Place Holder first, Graphics Later. Afterword: So... Yeah. That's all with today's Dev Diary, Dev Memo, Dev Essay, or whatever you want to call it. This is the first time I'm writing the script for the video with 1000+ words. and It's not even my mother tongue language. (I'm Japanese, by the way.) Anyway, If you're interested in my game project, You know the drill, leave a like, subscribe, join the Discord, and blah blah blah blah, bloody blah!