Dev Log, Week 1: Pixelart and a Brief History Lesson

I began working on graphics for Solid Light before summer began, mostly because I know I’m not an artist and it would take time for me to make some sprites (images used in game) that looked anywhere close to good. I’ve had some interest in pixel art before, but more in the NES style, which had pretty severe limitations that artists had to work around. But, I’ve always seen limitations as a challenge, and so I took the time to mess around until I had some decent-looking sprites. Below, the top 3 images comes from Castlevania 1, and the bottom 3 are sprites I made.

Clearly, the ones made by professionals are better, but the point is that they look pretty similar because they’re both working under the same limitations – the strictest of which is that only 3 colors could be used in one sprite at once. (Some characters in NES games broke this limitation by actually being multiple sprites at once, but that’s beyond this explanation.)

Anyway, most NES games don’t hold up very well graphically because of these limitations – even the most impressive ones. Sticking with Castlevania for this post, the third game pushed the NES to its limits, and is considered to be one of the best looking NES games of all time. The screenshots below reveal just how much farther there was to come, however. Note especially the glitching out near the top of the second screenshot – what’s being displayed is still almost too much.

I knew I wanted something a bit more modern looking than that for Solid Light, so I started looking into the SNES’s limitations, only to find them to be incredibly loose. As long as you were running something at the same resolution as the NES, you could display almost anything you tried hard enough to. Below is a screenshot of Castlevania 4, and I think the difference speaks for itself.

I knew I likely wasn’t up to the task of matching that myself, so I tried to look for and/or create some middle ground, and this is what I decided on:

  • Adjust the screen size to fill a proper widescreen monitor, based on Shovel Knight’s decision to do so. (If you haven’t played Shovel Knight, by the way, do that. It’s worth your money.)
  • Limit to 15 colors within each sprite, and 216 colors overall.
  • Design sprites and other graphic elements similarly to Super Mario World, an SNES game that kept some NES sensibilities like having a dark outline around characters and relatively simple background graphics, as shown below.

I’ll have full screenshots up as soon as a basic, working engine is up and running but for now, I’ll focus on the player character of Solid Light, who I’ve named Erika. To the left, you can see the color palette I’m working with for her.

Now, given that I just gave Mario as my primary graphical inspiration, having a red shirt, blue pants, and brown boots might seem a little bit suspicious, but there are actually good reasons for each step in the design. For example, I knew ahead of time that I wanted Erika to be holding a torch, which sets us up for a lot of warm color usage already. The torch uses 8 colors by itself, and with another 2 dedicated to skin tone, we’re going to hit our 15 color limit quickly if we don’t reuse reds, oranges, and browns in the rest of our design. The red shirt, orange bandana/hair tie, yellow gloves, and brown boots all stem from this desire to reuse colors. However, we can’t use only warm colors in our design, and having Erika just wear blue jeans both adds a nice break of cool colors and helps further characterize her as a pretty regular, relatable person.

Come back next week for a discussion of how to keep characters from falling through the floor, and stay tuned to find out how my decision to only use 216 colors will backfire on me. Because it will.

Leave a Reply

Your email address will not be published. Required fields are marked *