In Art Spotlight, we invite Sketchfab artists to talk about one of their designs.
Hi there! I am David Junghanns, 18 years old, living in Germany and recently finished the German equivalent of high school. Right now, I am taking a year off from school and studying to earn a bit of money, so I can see some parts of the rest of the world.
I started using 3D software about 5 years ago because I was enjoying games a lot (still do), so I got into Source SDK, a level editor for the games “Counter Strike: Source” and “Team Fortress 2” by Valve. Of course, this software is very limited to a certain use, which is creating maps for these two games.
I discovered that creating 3D environments was a lot of fun, but the software was so restricted that I couldn’t create anything interesting or even remotely realistic. After looking around for a bit, 13 year old me decided to use Blender because of its surprising amount of realism (If you can manage to set all the parameters right) and of course because it was completely free, as I couldn’t quite afford the many softwares costing thousands of dollars with my pocket money.
After watching way too many YouTube tutorials, I proudly created my first little picture. From there, my projects got more and more detailed and ambitious. At some point I realized that the amount of work that you put into a scene isn’t equal to the beauty of the result in the end. Of course, a minimum of work is required to make a scene look like something at all, but I made the mistake of working way too much on details that no one would notice in the final render.
For example, I worked several weeks on this pretty simple, pure modeling task. I had a lot more detail worked into it at first, but I didn’t like the result, so I toned down all the details and left it as easy and with the most easy shapes as possible.
Following this mindset, I got hooked onto low poly art. This kind of art has a special feel to it, as every low poly model is stylized, but you still can see what the model was based on in the real world. I made some low poly models and got encouraged by the positive feedback online, especially sketchfab and the steam community, so I continued making them.
The inspiration to this model is, as you would expect, “Monument Valley”. It is a mobile game by ustwo, based around a little girl with a pointy white hat. The concept of the game is to lead the girl around various different levels, all designed to mess with your head, as every single level has some sort of optical illusion to it, which you have to exploit to get to the exit of the level.
I played the game and was absolutely amazed by it. I wanted to create something like these in blender, but had no clue how to do it at first. How do you create a physically impossible thing in a software which is based around imitating things out of the real world ?
Actually, the whole game was made in blender, as David Fernández Huerta, a member of the team who created Monument Valley mentions in this interview.
The solution is that you have to trick the human brain into thinking certain areas of the models are connected when really, they aren’t.
How? The first step is an isometric camera. This enables you to have all objects be the same size no matter how far away from the camera they are. The second step is the rotation of the camera, in my case being 54.726°/0°/45° (X/Y/Z), which is about the same rotation they use in the game.
I started by creating the waterway, as this is the most striking and mind bending part of the model. From there I modeled the usual way, big chunks first, then more and more details. I tried to work in some less noticeable illusions, such as the horizontal staircase leading “downwards” behind the pillar with the ladder on it, or the waterfall not hitting the ground where it should besides seemingly falling down straight. These illusions are most noticeable when you simply turn the camera, as the second condition to get this illusion working (the cameras rotation) is not met any more.
What is also a bit special are the plants all over the model. The all look not very 3 dimensional and out of place. This further creates a feeling of an artificial, lonely world this girl is living in which the game provides. The odd 2 dimensionality is achieved by -who would have thought- creating 2-dimensional planes for the plants and tilting them by the same values as the camera. This way the plants face the camera exactly and feel like they were glued on or photoshopped in afterwards. The technical term for this way of placing textures makes the textures “billboard textures”.
After I did all the modeling, I went on to the texturing part. As I am not particularly good with texturing, I did it the noobish way: By coloring in the different faces. Everywhere I wanted a different color, I created a new face. This sounds like it would make the model unnecessary big, and it does with bigger models, but here it’s not that big of a problem as i only created more faces at the waterway and in the back at the target-like painting.
After I painted in all the faces, I baked it all to a big texture, excluding the waterfall as this should be a little bit transparent afterwards. This makes the part of uploading it to Sketchfab later easier, but you don’t have a lot of control over the textures afterwards, as you can only change values for the whole texture.
With that, I was finished with this model:
If you didn’t quite get my written explanation on how you create the illusion of an impossible object in 3D, here’s a model that explains it further by simplifying:
I also wanted to thank you guys for all the positive feedback, especially on this model, and of course, the crew of Sketchfab for making this all possible in the first place. Without Sketchfab, I would have never shared any of my work on other platforms than on steam, which is pretty meh for 3D art.