Long life to Embark studio and its fabulous procedural artists dream team !
Tobias Forsling talked about the way he created a very interesting submission for the Beyond Human competition. It’s a nice study of lighting and prop creation for 3d environments.
I grew up in Ystad, Sweden but about 4 years ago I moved to Skövde to study 3D art at university. During my time at university I met fellow artists that had the same goals as myself. We started working together in school projects which then grew into a side project. This is what came to be the first full scale game I worked on, Thunder League.
It was during the development of this game that we nailed down our style that would come to be seen way more in Lance A Lot (our next game).
Thunder League ended up not getting published on Steam (as intended) after members of the team left. This is where some of us moved on to become a new studio named Rocket Hammer. At Rocket Hammer I worked on Lance A Lot where I along with my fellow artists Anton Ehn and Johannes Palmblad polished up the visual style that became somewhat iconic for Lance A Lot. It was also during this project that I decided to focus solely on environment art. Up until then I had had a more general role of doing a little of everything when it came to creating 3d assets. But after working at Rocket Hammer for about a year and a half I decided to leave for a job at Tarhead Studio where currently work.
I’ve always felt a sense of pressure that I wouldn’t improve enough only working in one style of art at a time. Therefore, I have always countered the art style at work with the opposite in personal projects. If I do stylized at work, which has been the case for my time working on Lance A Lot, I try to move towards realism in my personal pieces.
3D Scene in Sweden
Right now Sweden is a place where game development and new tech is flourishing in both indies and AAA. I think indies are flourishing from all the support they get in starting up. There are incubators all over the country helping new studios to get established and grow out of infancy. At the same time, I think AAA has it great since Sweden has quite an attractive workforce with good healthcare and work/life balance, making it attractive for lots of overseas talent. Also, there is a lot of money and effort put into the education of game development in Sweden so every year loads of people get the chance to step into the industry via these schools.
Yeah for this project I wanted to move more towards a closed-off interior space. I felt that most of my previous work had been outdoor open spaces and doing interior was a way to tread new ground and further expand my skills. A big influence was that a very little part of the environment would be visible at each given time since walls would be covering other parts, this gave me a lot more performance to play within that closed of space. It also forced me to light things differently. Working in an exterior scene I would use directional lights but that didn’t work very well since the walls weren’t supposed to cast shadows from outside the world. These new factors proved to be interesting new challenges that I had lots of fun exploring.
When it came to the modeling I started to break down the environment asking myself what parts can be repeated both for the sake of saving time and performance. I broke down the walls and floors in sections building it in modular sets to easily be able to adjust the length of the hallway.
I then built simple blockouts of all the parts to nail the scale down from 2D to 3D space, which was the first weeks of my work. The biggest challenge in the modeling phase was the bodies in the tanks since I hadn’t made a character in quite some time I had to dust off my character topology skills a bit.
For the thicker cords going along the floor, I made a few vertically tiling materials. I made different materials at first since I didn’t quite know what I wanted, however, I ended up only using one of them in the end. Having cables that big I had quite a problem figuring out how I should texture them without getting a too high texel density. So by unwrapping the cables UVs so the width of the cable took up my entire UV-space horizontally.
Now the real power in getting them to tile, of course, comes from the textures. In Substance, I played around quite a bit with different ways to get a good tiling spiral pattern and in the end, I found that using the Fibers1 node worked the best. By rotating it and scaling it in along one way until I got the thickness I wanted of the lines as well as when the lines matched up in the viewport, making it tile correctly.
For the texturing of the rest of the scene, I only applied tiling materials in the beginning to all the parts of the environment without a finalized UV-map to try out the textures in the scene. It was then quite easy to just go in and give the models proper more optimized UV-maps. Some meshes were then textured in Substance Painter while some remained to use tiling textures. Working like this gives me a feel for the colors and materials in the scene quite early without having to spend a lot of time on UV-mapping.
Hah I have gotten a lot of questions over the past year to talk about the lighting in my Unity scenes. As I mentioned before I had to change my workflow a bit in this project for the lighting. I used spot lights to more easily control where light and shadows were cast in the scene. Later during the project, I also found a solution for volumetric lighting in Unity which have it a bit more punch. Now up until the end I stuck with the lighting similar to the concept but as Mathew O´Hallowran (one of the challenge hosts) said towards the end, I had focal points all over the place. So, towards the end I focused the light more on specific parts like the tanks and at the end of the hallway.
One thing I always do in my scenes which I think a lot of people miss out on is using color correction. I personally use LUT (look up textures) to tweak the colors of my scene, and it does so much for the feeling of the scene.
I worked on the project since the challenge started in the middle of July making the time about 2 months. I encountered several problems along the way, one major was getting the lighting to give me a good tonal range. I often ended up with areas either too dark or too bright for my liking. In the beginning, I solved this by adding eye adaption but I later turned it off since it limited the control I had with the lights.
Another step that I almost overlooked was to get all the dirt that is shown in the concept. Every part of the environment feels dirty and old which I wanted to preserve while not killing performance. I wanted the environment to feel old and almost abandoned or neglected. I ended up adding that dirt using deferred decals, now they had some issues with rendering the albedo in shadows but for me, the important part was that I got that breakup in the micro surface.
Thanks a lot for your help. Looking forward to hearing from you soon.