The Art of Even the Ocean

Chapter 1: From Pixels to Painting

Marina Kittaka
9 min readSep 20, 2017


Hi, my name is Marina Kittaka, and I made Even the Ocean with Sean Han Tani. Even the Ocean is a narrative action platforming video game about balancing the light and dark energies that hold the world together. I’ve wanted to do things like share my process and feel a part of some kind of (idealized) vibrant scene for a long time, but I haven’t been able to. It’s so easy for everything to seem like a dead-end… I’m too frustrated by my own work… I’ll never be aware or astute enough… I don’t give enough interest or attention to the work of my peers… Sometimes life just feels like a series of depressions, and I can understand the previous ones but not the current one (that’s what therapy is for, Marina).

Recently, I released a big game, I ended a relationship of 5 years, I moved halfway across the country, and now I’m gonna try to make a new start and be braver and more open and more forgiving towards myself. I think I’m entering a season where I want to share as much as I can and just see what happens.

Paul Arden once said “The problem with hoarding is you end up living off your reserves. Eventually you’ll become stale. If you give away everything you have, you are left with nothing. This forces you to look, to be aware, to replenish. … Somehow the more you give away, the more comes back to you.” I don’t really know who Paul Arden is, but writer Austin Kleon shared this quote in his book, Share Your Work!. That book partly inspired me to start a new website/blog/newsletter, and that quote is comforting to an artist like me, who sometimes feels like their life is small and boring and poor sustenance for a lifetime of artistic work.

So let’s go on a journey together through Even the Ocean’s art! You’ll find that one of the recurring themes here is how I developed a workflow to create visuals for a large fantasy world while managing the costs in terms of time and motivation. Unclear what — if anything — here will be transferable to other workflows, since a lot of it has to do with my personal quirks, limitations, and interests as an artist. But anyway, let’s have some fun!

Limited Color Palette vs. Curve Tweaking

Limited color palette for Even the Ocean. And the palette applied to Anodyne screenshots for rough testing.
Concept art using the limited palette.

Initially the game was going to be completely pixel art, and also two intertwining but separate games (Even, a modern-day life story, and The Ocean, an Anodyne-esque fantasy world). I tested the idea of creating two distinct 256 color palettes to create cohesion and contrast. However, I soon realized how heavily I relied on tweaking color curves throughout my general art process — and tweaking curves is somewhat incompatible with having a strictly limited set of colors.

Bonus Demo: Gamut Masking and Curves Adjustment

As a side-note, I realized that tweaking curves is sort of like the reverse of (Creator of Dinotopia) James Gurney’s incredible gamut masking method, in which you create the mood and palette of a painting by roping off a limited section of the color wheel to work with. One method limits your colors before you paint, the other limits your color after you paint. A mix of both methods sure is nice for digital painting! Warping the curves will often look cheesy if push it too far, but you can gamut mask the heck out of the color wheel and it’ll basically be cool.

The reason I bring this up right now is that I learned about gamut masking too late to really use it for Even the Ocean, but it’s rad enough that I thought it would be fun to do a little demo. It may also help explain what “adjusting curves” is and why I rely on it, for those who are curious.

Psst… Here’s a handy tool for gamut masking.

Image 1A uses the full gamut. Blues are Blue, greens are Green, reds are Red. It’s pleasant, but what if we want something more distinctive and dramatic? Image 2A was made with the color wheel restricted to a triangular gamut — the greens and blues are basically gray, red is a bit stronger, and yellow is the strongest. By restricting ourselves to this chunk of colors, we created a stronger mood, but we still intuitively know the trees are green because they’re the greenest possible thing we can depict in this gamut.

Color curves can be adjusted in some image-editing software (I used Photoshop here). By messing with the different lines, you can, for instance, make the dark areas of your image contain more red (as is the case in 1B — look at the graph, and the way the dark tree trunks sort of glow red). The easiest way to understand is to play with it!

Image 1B is 1A with a color curves adjustment. We’re able to mess with the mood a lot retroactively with this method, but the colors also get more wonky and unreal the further we push. Image 2B is 2A with a pretty subtle curves adjustment. We started from an already strong place, but perhaps took it in an even more cohesive and dramatic direction! In particular, I regretted how weak the reds were while working on 2A, so I bumped those up a bit. Of course, you can also do adjustments then continue painting — it doesn’t have to be only at the end! Now back to Even the Ocean!

Steps and Stages, Illustrated

Let’s take a trip down memory lane, shall we? Or as you like to call it, “perception/encoding lane” (I’m assuming a lot of people have not seen this imagery before). Over the years of development, we tried a series of different methods which slowly guided me towards the final art style. Here’s an overview of that process:

As mentioned above, I began with tile-based, limited-color-palette pixel art.

The parts that the player stands on and interacts with are made of 16 x 16 pixel tiles, while most background and foreground elements are made of larger images.

Next, I ditched the limited color palette and used any color I could get my grubby mitts on.

Note the subtler use of color and reduced dependence on dithering. Most blatantly, the background gives a much airier sense of atmosphere.

Then I started working in what we called “setpieces”, building areas out of bigger chunks of art like rocks and trees that would repeat throughout the area (instead of only building with square tiles). Many high resolution 2D games are made this way. But doing it in pixel-art as an extensive part of the level geometry is iffy because you can’t easily rotate and scale to break up the repetition (computers have a hard time scaling and rotating pixel art because there is so little visual information to work with!).

The rocks, grass clusters, tree trunk, and sand dune were large chunks created to be reused throughout the area.

So then we figured out that it would technically be doable to just make the areas out of giant single images, rather than composed of small parts (whether tiles or “setpieces”). I felt quite freed up at this point, because I was able to more naturally create large architectural and geological forms such as the bridge pictured below. But I soon ran into the difficulty of creating these enormous images in pixel art (duh).

The amount of time it takes to go from “almost done” to “done” in pixel art at this resolution is huge if you want it to feel traditionally polished.

Finally I started digitally painting the areas. In case that terminology difference is unclear — I stopped drawing the art by manipulating individual pixels and began to use brush tools which manipulate many pixels at once — a process that more closely resembles traditional painting.

The Graveyard of Unused Pixel Art

The following is most of the unused pixel art I created for Even the Ocean’s outdoor/nature areas. Most of these images represent limited tests and were not applied to create entire finished areas.

Pixel Art and the Big Picture

Pixel art is easy to fetishize. There’s a perfectionistic allure to handcrafting each little dot on a screen, to chunking those elegantly arranged dots together to create a huge and wonderful world. In some of the many many hours I spent pushing pixels, I would self-indulgently envision an excited player piecing together whole maps in image-editing software and seeing the larger shapes and structural ideas underlying our area design.

So I understood why some people liked aspects of these earlier versions better than our new direction, and maybe even better than the finished art (from screenshots anyway). But aside from practical workload concerns, the problem was mainly that it felt wrong in motion, and interaction. The density of detail and the limitations on shape felt stiff and ill-fitting, everything felt like it was made of the same dreary material. After Anodyne’s small screen size and top-down perspective, I was excited for the luxury of open skies and hills rolling in parallax, but it came out feeling dead. I couldn’t sense the humidity of the spaces; they didn’t envelop me in their aura. I gotta be able to sense those dang humidity levels!!

Screenshots from the finished game.

So for all these reasons and more, I switched to a non-pixel-art, painterly style. The new style had its own quirks and struggles, but I knew quite quickly that I was headed in the right direction! Could someone else have evoked the proper humidity for this game while continuing to use pixel art? Perhaps! But when you want to finish projects, it’s great to have your interest/enjoyment of a task be proportional to how heavily that task is required in your work as a whole.

On May 16th, 2015, I posted about switching to a new, painted art style on our TIGSource development forum. And apparently someone had been rude to me about it on twitter… ha ha.

When you look at the time that I announced switching over to the painterly method — it’s actually past the half-way point of my entire work on the game! (~1 year, 9 months after starting. ~1 year, 6 months before release.) Relatively speaking, the work began to fly by — not just because it was inherently more efficient, but also because I found the process more enjoyable and attuned to my interests!

Thanks for reading! If you enjoyed this chapter, consider subscribing to my newsletter or following me here on medium or on twitter! In the next chapter, I’ll be talking about the actual painting process for Even the Ocean areas!