For quite a while now it’s been the practice of many smaller/indie developers to feature low resolution pixel graphics. This is both an aesthetic choice and a matter of resources as a lot of developers can’t afford to create the visuals of big budget titles. So they go the pixel-art route instead.
In general, I really like the idea. I grew up with the 8-Bit and 16-Bit consoles and I get a lot of enjoyment out of this style. IF it’s done right.
Now, I assume that pixel graphics are chosen to emulate a certain level of technology, like the SNES, to speak to people like me, who like the style. I’m also talking specifically about games that use lower resolution graphics and not 2D in general.
Another disclaimer: Nobody’s perfect and while this, and probably the next, article talk about “mistakes” when making these games, I’m just speaking for myself and my personal preference. As a developer you can do whatever you like and I’m sure most people don’t mind the things I might complain about or maybe they even like them. A lot of developers might also have legitimate reasons to forego a more accurate style.
Here we go!
Pixel Sin No1: Using different resolutions
There are a couple of different variations for this and usually all of them look “not quite right” to me. Graphics back in the day were pixelated, because the system didn’t have a higher resolution. Those “blocky” pixels, were the highest resolution possible and some people seem to think that any kind of pixel-ish style is okay and they use sprites with completely different pixel sizes. This looks really bad in my opinion.
Another, better looking, but still not authentic way of this is the interface. Often, games have low-resolution pixel graphics, but to improve usability, the developers make a more high-res user interface. This is good, especially for the end user, but it still doesn’t feel right for me. Also the challenge back then was to make a functioning user interface with a very limited resolution.
Now, there were a few rare exceptions on the original 16Bit systems (and some of the few later 32Bit 2D games) that had a similar effect when upscaling a 2D sprite (not zooming in). Not a lot of games used it, so it was never a style used widely.
Where I can look the other way is not using system specific display resolution, for example not using the 256 × 224 resolution of the SNES. While I prefer an overall low resolution, especially with modern displays, I don’t care if you choose a 4:3, 16:10, 16:9 aspect ratio. As long as the resolution isn’t too high (at least below the 640×480 VGA resolution) go nuts.
Pixel Sin No2: Modern effects
Things like post processing, high-resolution explosions, procedural effects generated with a more modern engine might seem cool, but again, these effects don’t fit the idea of retro graphics. The above screenshot is a very subtle example as it only has a small blur effect, which isn’t even that high-res, it’s just using a 2x higher resolution than the actual ingame graphics. On first glance it looks okay, but there’s also games that shove very high resolution lens-flare effects in your face, which looks quite bad.
This also applies to lighting. Actually lighting a scene by hand with the appropriate pixel resolution is much more time consuming than just putting in a generic light source. The lighting is connected to another modern “effect”, if you can call it that, and that is the color depth.
The systems back in the day were quite limited in terms of color depth. Not only was the overall color palette limited, usually the system could only choose an even more limited number of colors to show on screen at the same time. With the VGA standard, DOS PCs had quite the advantage over consoles like the SNES or computers like the Amiga, because instead of choosing for example 32 colors from a total 4096, the palette could be defined as 256 colors from an 18-bit (262,144-color) RGB table, allowing for much better visuals.
But many developers choose to utilize the full color palette available in their graphics software, which is of course easier than sticking to a limited number of colors, but often has results that aren’t helping the overall consistency of the visuals. Even games that look like they don’t use a whole lot of colors, might still be guilty of this.
It’s a fun exercise: Take a screenshot in a game and use your software of choice to count the colors used. Then reduce the color depth of the image to 256 colors or even less. You’d be surprised how little is lost with an optimized color palette. And even with a heavily limited number of colors, you can achieve good results…if you can actually make good graphics using pixel-art.
Which brings me to my next point, but I’ll continue with that in part II, so stay tuned.
What’s your opinion on modern games that use pixel art for their graphics and are you feeling the same about my personal issues with the matter? Let me know in the comments!