I think the last post nicely wraps up my current thoughts on the cluster of topics related to writing, creativity, and AI, and I think we are all ready for something completely different. Therefore, today I will expound upon the concept of pixel art, what makes it aesthetically pleasing, and my favorite example of the medium.
If you play video games or spend much time on the internet, you’ve definitely seen some form of pixel art. The only criterion needed to qualify a piece as “pixel art” is that one must be able to discern the individual uniform points of color (pixels) that make up the image without too much effort. Interestingly, many other media could be considered pixel art or at least pixel-art-like by this standard. I thought of cross-stitching myself but Wikipedia also lists mosaics and beadwork. You can read the Wikipedia article for the rest of the history if you like, I won’t bother reproducing it all here. Suffice it to say, there is great variety and depth since the term encompasses everything from very low-res images with few colors to ones where you can barely even see the individual pixels.
Back when artistic media were much more limited, it is easy to see the appeal of threading beads, laying tiles, stitching stitches, painting pixels, or arranging other small uniform pieces in a uniformly-spaced way to create an abstracted 2D picture. Due to the uniform mapping, it becomes easier to do a kind of quasi-mathematical transformation from the 3D space onto this new 2D one, enabling the representation of pretty much anything you’d like. But now that we have advanced graphics, video, animation, and other technologies, why do so many people still go back to pixels? It’s not truly unique to digital media, but there is a strong association with the early days of digital technology so maybe there is a nostalgia factor.
I think a large part of its appeal comes from not having to get every little detail right and instead focus on the “essence” of what you are trying to convey. If you use a medium that has more fine-grained detail than a limited pixel grid, there are lots of small things that can make the image feel “off” somehow. This not only includes minor technical details like shading or proportions but larger stylistic elements such as the look of an animated movie or TV show. Just going for a purely photorealistic look can be boring, but as soon as you stray from photorealism it can be interesting but also alienating, leading many animation styles to simplify things much in the same way pixel art does.
By limiting the way you can represent things, pixel art naturally leads to emphasis on other aspects of the medium like color, animation, and interactivity. My favorite examples of this are from an artist named Mark Ferrari. You can see examples of his work here and here. Not only do I love beautiful landscapes and nicely-done pixel art, but his use of animations and day/night cycles is genius. The images are rich with detail: the water flows, sunlight changes, rain falls, all by changing the colors of the individual pixels.
I wanted to use some of his art on my website if possible, but unfortunately all of it is tied up by various copyright protections. I tried using AI to create pixel art in a similar style but so far haven’t been very successful. Hopefully there will be a follow-up to this post when I have figured out how to create my own art like Ferrari’s landscapes but that could be a tall order. Regardless of whether I have to create the entire piece myself or can somehow rig an end-to-end workflow to mass produce them and simply select ones I like, I really want to make works like his. Would something be lost if I could just summon as many of these dynamic pixel landscapes as I wanted in an instant? Perhaps they could be made even better, more interactive, but isn’t the main point of pixel art its simplicity?
Wait a minute. How did I end up on the topic of creativity and AI again? I was trying to change the subject! I’d better end things here before I bring up the writing process somehow.