In Art Spotlight, we invite Sketchfab artists to talk about one of their designs.
Hello everyone! My name is Murilo Kleine, I’m a Brazilian artist and graphic designer. I was kindly invited to write about my work on the game Sprout’s Tale, a 3D puzzle-platformer that will be released in 2016. We are a 3 man indie team scattered across the globe — Tom Murphy from the US (Game Director/Designer), Mihai Cozma from Romania (Programmer) and me, Murilo, from Brazil (2D/3D Artist). We work on the game in our free time between jobs, freelances, projects, family and dawn, well, c’est la vie!
A bit about me
Being an artist has been my goal since childhood. The first contact I had with 3D was around 1995, as a 6 years old little dude watching Jurassic Park – and all its epic 3D dinosaur-y glory – for the first time. Oh man, that was awesome!
Of course I was too young to even understand the concepts of 3D, but the movie inspired me to draw. As years went by, from pencil and paper drafts I moved on to inking, from inking to basic coloring, then shading, perspective drawing and so on. Much later I met Photoshop, pen tablets and digital drawing/painting. Having worked with traditional and digital illustration tools helped me a lot when I finally started learning 3D, not only when designing a character or a scene, but also for modeling, understanding of the 3D space, texturing or even to understand topology based on anatomy for better rigging and animation. But enough of the past, let’s talk about what is worth.
There are a few steps I usually go through when creating a character, scene or even the simplest of the models. I will talk about these steps below, but keep in mind that most of them are related and may influence one another.
Concept art and visual reference
Sprout’s Tale began as a 2D game, and because of that, when designing the characters, enemies and almost anything that would have to be frame by frame animated, we decided to follow a very minimalist approach.
Colors would have to be simple and yet meaningful, shapes would have to be of easy recognition. An enemy should look like an enemy even if it was just plain smoke or a simple ball.
At the same time the world of the game would have to change from a dreadful, dangerous and desolate land into a much brighter and full of life place as the player makes progress. Our main character has the unique ability of “growing” life (his name is Sprout after all, right?), so grass and trees would have to grow as he goes through the levels, not only as a decorative element, but also as a gameplay mechanic. So how do you, as an artist, transform all those ideas from the game designer into visuals? Well, concept art!
Drawing and loose painting is much faster than modeling and texturing, it’s also much easier to erase and redo parts of an image then to change your mind halfway through the modeling of an object and completely redo it because it’s not looking the way you first imagined, so I really like to spend some time designing what I’m about to model or at least spend some time observing and learning from images on the internet, books, etc.
I don’t usually draw or paint too much detail when I start creating these concept pieces, most of them are just drafts that nobody on their right mind would use in an art book, they work mostly as reference for myself, and can be easily discarded if not suitable for the game.
After the drafts are approved I usually redo some of them and add some detail, clean up the line work, add color and shading in Photoshop, just until I feel comfortable enough with the result.
Following the minimalist aesthetics, the models used in the game are usually simple, symmetric and considerably lowpoly. I use Blender as my main 3D software, so I normally start from cylinders or cubes, add a mirror modifier along the X axis, a surface subdivision modifier (depending on the model) and then use basic tools to change shapes, add loop cuts, extrude faces, join vertices, split edges and so on. My first objective when modeling is to get the object shape and proportions right and then, from there, add detail, tweak the topology to fit animation or visual needs and add or subtract polygons.
Once a model is ready, we test it along with other assets to make sure its scale and proportions won’t need any adjustments and then test it in-game to see if the bounding boxes are working the way they should. Being the only artist on an indie team often means modeling objects that require you to step out of your comfort zone, it’s a great experience but it also demands extra attention and testing.
After the model is tested and approved, it’s time to unwrap it. I really like to take some time to organize the UV coordinates in a way that makes sense when creating, painting and editing the textures in Photoshop. After I’m comfortable with the result, I export the UV layout and a simple AO (Ambient Occlusion) map.
The workflow is pretty straightforward, using the selection tools, I create masks or layers filled with the base colors, apply the AO texture map at a low opacity and use a pen tablet to add more shading, colors and details. Gradient layers of color set at different blending modes can be used to tweak the coloring, make warmer shadows and lighter or darker areas, edited stock textures and filters can also be used depending on the model.
Working with a lot of layers can get really confusing, so I try to keep it as clean as possible by giving names to important layers and by separating different parts of the model UVs into groups.
I don’t usually use sculpt tools for this game, the game visuals allow me to bake the normal maps from simple displacement or bump maps as most details are just painted into the diffuse textures.
Specular maps, when needed, are also created in Photoshop by adjusting saturation, grey scale values, inverting colors, adding or removing details and filters (Glowing Edges filter, I choose you!).
Rigging and Animating
I had never worked with 3D animation before Sprout’s Tale, so I’ve been trying to learn as much as possible since 2013-2014, when the project evolved from a 2D game into a 3D one. The rigs used for the animated assets and characters in the game are simple. In part because there’s no real need for very complex rigs for the animations that we use, but also because I still don’t know as much as I would like to know about rigging. For better or worse, the lack of better IK solvers don’t make a huge difference when creating our simple character animation cycles, the simplistic rigs can get the job done with no problem.
I assign the bones to parts of the mesh via weight painting and test its behaviors, if needed, I make changes to the topology of the model for better bending and tilting. A little knowledge and visual reference in human and animal anatomy can help a lot when it comes to rigging and topology, it makes it easier to animate and the meshes deform in a more natural way.
Just like designing a character, animating one without proper reference can be a pain. I watch as much material as I can find on the internet, study it and then transfer what I learned to the models considering how the animation should look in the game. Once I have the key frames right I add some in-betweens to correct the poses and the flow of the animation and try to consider the weight, balance and impact of each movement as much as I can.
Well, that’s all folks! I hope this post was helpful and interesting to some of you! It was hard to write about the different topics without making the text overly long, even harder to make it not too shallow! I’ve got a long way to go and a lot to learn when it comes to 3D and art in general, and Sketchfab seems just like the right place to do so.
Thanks a lot for the opportunity, Sketchfab! You guys are awesome!!!