32×32 Characters – Creating the Sprites for Wonton 51
Oli tells us about how he's 'terrible at art' and then goes on to say how he did all of the art for THE ENTIRE GAME. Jeez, PR nightmare. Astonishingly, it turned out pretty sweet. Read how he decided on the character styles here.
Creating the art for Wonton 51 was terrifying. I'm no artist, I'm a designer and I have always been terrible at art.
The art lessons I had at school were worthless to me, the teachers spent all their time with the talented students and gave me no instruction. Game artists that I work with are horrified when I bring them a sketch of how I think something should look in game. In the games industry we often joke about "programmer art" as coders will put in horrible placeholder images to for prototypes, hoping that the artists will replace it with something pretty. Let me tell you, the one thing worse than programmer art is designer art as we do the same but also get it into our head that our sketches are awesome and we're just like real artists.
This was hurdle number one for Wonton 51.
When we decided to make Wonton I knew straight away that the art would be a problem. I really wanted to put my 'designer art' into the game but knew I might not be able to take on such a big project when I can't draw. Instead of passing the role onto someone else, I immediately decided to use pixel art for the game.
That's not to say pixel art is easy or a lesser form of art, I adore it. I grew up with it and it has always enthralled me, without a doubt Street Fighter 3 is the best looking game ever and the beauty of the character sprites is astonishing. The fact is, I knew I could could at least attempt pixel art because designers are problem solvers and I compare pixel art to problem solving. You have a super small amount of pixels you can use to create a character and the skill comes from placing individual pixels in just the right place. People need to look at a 16x16 sprite and see a character they can identify with. This constrained, problem solving aspect seemed like something I could do, whereas drawing something freehand with no rules terrified me. So I dived in.
I took the bull by the horns and drew the first Jonny sprite one Saturday morning. I started off thinking that Jonny would be one colour and the enemies would be another as Jonny needed to pop from the background and clearly be seen so I chose to do him in white. I wanted him to have large shoulder pads and a rounded head/chest, just like in the classic Japanese mech designs I love. He would also have a long fin coming out of his back and a jet engine style jet pack.
All of these elements needed be in contrast to the enemies but when you design game characters you must also make them look like they inhabit the same world. I wanted Jonny to look like he pilots an older model of mech than the enemies, so the enemies ended up having pointed chests and multiple sleek fins on their back. They also ended up being red because RED EQUALS DANGER.
Jonny and Enemy 1 ended up changing very little once they were drawn. In the first Jonny image Jonny had a smaller fin and some strange looking jet pack exhausts. He was also missing a much needed splash of colour. I revisited him some time later and spruced him up a bit. My first sprite was done and other than some small tweaks he was pretty much good to go.
Initially the collision box for Jonny (an invisible box that checks to see if Jonny has collided with a bullet or a building) was the same size as Jonny. We decided to scale the box down as this meant Jonny was more manoeuvrable and the player could do Wicked Shit by weaving in and out of bullet patterns and out of the way of buildings. To communicate this to the player, I added the blue core on his back that you can see in the image above, allowing the player to learn what will and won't collide with Jonny.
Enemy 1 received very few tweaks, I got in the zone and blitzed him. He defined the look for the rest of the enemies and provided the key to making the rest of the enemy sprites. I spent a while going through a bunch of different designs for enemy 2 and 3.
I knew 2 would be smaller and faster and 3 would be bigger and slower and I tried to reflect that in the way they looked. I tried loads of designs from scratch and they all looked terrible. It seemed I had struck gold on the enemy 1 and that wasn't going to happen again any time soon.
Luckily, I eventually realised I needed to kit bash them. Modellers often use a term called kit bashing to make game character models. This is where you take an existing model and scale, move and reshape parts of it to make a new model. This provides consistency and makes the models look like they are all part of the same group. For example, in the first Gears of War they created the standard Drone and then scaled him down in lots of different ways to make the Wretch and then up to make the Boomer. Gears is a great example of consistent enemy faction design. They all look like they are part of the same group. It also makes life a bit easier for a new artist.
I pulled enemy 1 apart and scaled him up and down. Once I had the size in place I redrew bits of him to make the new enemies. Enemy two moves quickly so is more pointed and has extra fins. Enemy 3 is the heavy guy, so he is much bulkier and has a huge gun.
The next enemy to do was the boss. He started out life as a big grey block drawn on Boxing Day (no rest for the wicked) to act as a prototype to test sizing. It turned out the size was perfect and I launched into doing the proper art.
The boss is a big flying fortress rather than a mech so I looked at a lot of aircraft references and the beautiful boss sprites in the Raiden games. The theme across pixel art representation of aircraft tended to be all about the panels and vents. I found that they helped to break up the large flat areas and made it look more like a thing than a block.
As with enemy 1, I got in the zone and over a couple of nights did the final art which looked amazing (if I do say so myself) when it first appeared in game. I'm used to seeing my design being implemented in games, but seeing something you've drawn being put in felt much less abstract. It was a strange but amazing feeling.
The hardest art to do was the Cookie and Jonny Mech portraits. Cookie was done first and being inexperienced I did her in a painful, convoluted way and I've never been happy with her. Jonny Mech was really hard as he needed a body and shading, something I don't have a lot of experience with, so that took a lot of time. Rachel did the bulk of the Jonny portrait and fortunately I could pass on some of my previous experience and she ended up doing a much cleaner and nicer image.
Baron Wonton is a mess and I don't want to talk about him.
Maybe now I could do a better job as I am more experienced but one of the hardest things about making a game you've invested in is that sometimes you have to stop fiddling and just get something in the game to meet your deadlines, even though you desperately want everything to be perfect.
Fun fact time. Wonton 51 was once a side on platformer with a running character but we knew that there was no way someone new to pixel art was going to make fully animating character sprites. The complexity of the art had to give way in order for me to get to grips with doing simpler pixel art and this in turn necessitated a design change.
The game is still spiritually in line with the original concept but a lot has changed since I looked at a platformer sprite sheet and said "Oh fuuuuuuuuuuuuck..."
My advice to anyone starting art when they haven't done it before is to just start doing it. You'll only get better. Know your limits but don't be afraid to kick the door down.
P.S. There is also one other sprite in here I haven't talked about. Have you managed to find it in the game?