Ever thought of creating your own tilesets? Either for yourself or for sale? Why don’t you check out this very candid discussion of what it takes from one of our more accomplished tileset artists?
I’m Celianna, known mostly for the resources that I create for RPG Maker. To this date, I have created three full tilesets for RPG Maker. The first one was simply a collection of all the resources I’ve made over the course of three years that I spent making free resources for the community. The second one was my first real tileset, in the sense that everything was 100% made by me, and also included a TileE. The funny thing is, when I assembled my first tileset together, it took me about two weeks (since I did end up creating new resources for it, and I had to fix autotiles) to wrap it up. Not bad, I thought. I can easily do a tileset on my own in a month!
Or so I thought.
Creating your own tileset from scratch probably takes four times longer than you had originally estimated. And even then some more, just to be safe. When I started, days trickled by as if it was nothing, and before I knew it, I was only half-way done and already two months had passed. I eventually managed to finish it up after three months. That’s not to say I worked every single day (I did take breaks to prevent myself from going insane from drawing), nor did I work in a consistent schedule (some days I’d work 12 hours, other days it was 3 hours). All in all, I’d say I spent around 200 hours creating this tileset.
So I had finally managed to finish one tileset, I figured I’d be better at estimating how much work goes into creating them.
For my next tileset, which hasn’t yet been released, I wanted to create more. I wanted to create an exterior and interior tileset for the B-E tiles, not to mention different seasons for the exterior tiles. That, and I was going to create fitting icons, and bunch of character sets. All in all, it was more work than my previous tileset. I estimated I’d be done in four months.
But four months passed, and I wasn’t even half-way done. In fact, I even managed to scratch some things I had made because I wasn’t satisfied with them and restarted from the ground up. This set me back a lot. Not to mention real life bothered me quite a bit (not only did I go on vacation to another country for two months, I also spend a month and a half moving out into a new place), which made sure that my tileset’s progress dragged on forever.
To this date, at least as far as this article goes, I am only about 93% done with my newest tileset, and I’ve been working on this since March. It is now November. That’s almost 9 months. Honestly, either I’m terrible at estimating, or I’m severely underestimating the amount of effort and work that goes into creating a tileset. And that’s not just me, it’s what a lot of people do.
It’s not about skill, or finesse, or how good of an artist you are—though those things will certainly help you out—it’s all about managing your time, and realizing that completing these tiny images will eat up a huge chunk of it. More than you’d originally thought. A lot of budding artists, much like myself, think it’ll be easy to create their own tileset. After all, they can already draw, so it shouldn’t be hard making their own tileset, right?
This is why a lot of artists still end up relying on the default RTP, or resource packs already made available to the public for their tilesets, with maybe an edit here or there, but never almost 100% custom. Because it’s simply faster this way, even if they could, theoretically, make their own tileset. But no one wants to spend six months on only the graphics when they have to make the actual game as well. Because, well, creating a tileset takes a lot of effort! Way more than you’d think if you looked at the finished product. And that’s a mistake a lot of people make, and they end up underestimating the effort it took to finish it.
I’m still looking at my own tileset from time to time and think out loud; “How did I spend several months working on this!?”. But I really did. I’ve put in more than over 350 hours for this tileset (I use a handy dandy program called ManicTime that keeps track of this for me)—probably closer to 400 hours as the creation of a tileset doesn’t only include the use of an art program—and that’s all spread across those several months I’ve been working on it.
Other artists, who underestimated the effort it takes to create a tileset, get discouraged because it’s taking so long, and then they stall their work, or never complete it. Or people who are looking to hire artists to create a tileset for them, don’t realize they’re underpaying their artists because they’re under the impression that creating a tileset is easy work. It all looks so easy when you see the RTP, and you think: that’s not a lot of art at all! But it all takes its toll, it all takes a lot of work—tileset making isn’t easy.
First and foremost, it requires you to be able to draw, the better you are at art, the better the result. If it’ll speed things up? Not so much. A crappy artist might actually spend less time on a tileset than a very good artist, because the better you are, the more you’re able to see flaws in your work and the more time you spend on fixing them. But then you also need to factor in the time it takes to create an autotile. Obviously, you first need to know and understand how autotiles work. Fortunately I know perfectly well how they work, but it doesn’t make it any less time consuming. It’s easy to create a square bookcase in a few minutes, but animating a water autotile with corners? That’s going to take more than a couple of hours to get it right. Autotiles in general will always take longer, so be prepared to spend more than a month on finishing up those pesky TileA1-A5 tiles.
Well big deal—autotiles are only half of the actual tileset, and that is correct. But good luck planning out an entire sheet so that it fits in perfect 32×32 squares, and so that all your objects take up the appropriate amount of space, and that there are no stray pixels from one object bleeding into another object, and there’s no empty tiles left that you might have missed. Not to mention you have to set it up in such a way that people can use it without any instructions, so tiles have to look like actual objects, and not like jigsaw puzzle pieces. You’ll probably end up spending a lot of time playing tetris with your tiles to try and fit them into the tileset since it’s all so rigid. And forget about adding shadows to your objects! Those will fall out of the grid and get cut off, so you’ll probably end up with either a very wonky shadow, or no shadow at all.
And then, when your tileset is full, you realize you still have objects left—but there’s no space for them anymore! No amount of playing tetris can make these tiles fit in the image. So you have to decide; do you create an entirely new image that only takes up a few puny tiles while the rest is blank, and risk people complaining about how ‘empty’ your tileset is (or worse; they think it’s an error, or they’re missing tiles), or do you turn it into a characterset (which means doing math, and people might not ever think of looking inside that folder for tile objects)? Or do you throw your hands up in the air, and say; well I’m just not going to include this?
There are many tiles that don’t make the cut in my own tilesets. I tend to include these as extras for parallax mapping (but then I still risk getting complaints because people don’t know what parallax mapping is). But plenty of times, I simply do not add them. I’ve created a lot of autotiles that I end up editing, or fixing, or simply replacing with newer ones, and they’ll never see the light of day. Fortunately, I can always use them for something else, but it won’t end up in the tileset due to its rigid formats. I’d rather not have an entire image for only one autotile (since RPG Maker VX Ace now allows you to use multiple tilesets), so I cram everything I have in one image. There are a lot of artists out there who actually do this—keeping the space empty I mean—but they also get the backlash of the community who will complain about ‘missing parts of the tileset’ or ‘there’s an error and my tiles aren’t showing up’ or simply ‘wow what an empty tileset’, so I decide on full images. And what doesn’t fit, gets cut off.
All in all, tilesets require lots of planning, lots of correcting, lots of testing, lots of tetris, lots of idiot proofing—honestly, only about 60% is actually about drawing the tiles. The other 40% is getting it to actually work.
Making tilesets is a lot of hard work. So remember that artists, in case you’re thinking of creating your own tileset. Or for those commissioning artists; don’t underpay them. It takes far more effort than you could ever imagine.