@Utsav singh, Yes. But a kind of lowpoly projection to have the best balance between poly count and alpha/quad overdraw.
Appreciate the shoutout!
your shader complexity is low because you used true polygon models instead of just a masked plane to prevent alpha overdraw?
Rebecca O’Neil discussed her approach to bright modular urban spaces with an abundance of flashy 2d ads.
Hello, my name is Rebecca O’Neil and I am a 2D/3D Artist from Scotland. I’m in my last year at the University of Abertay Dundee. My first role in the games industry was as an intern at 4J Studios, working on Minecraft, which was a lot of fun – pixel art is very addictive! But studying at Abertay has been great for opportunities as well; it was thanks to the university that I got to do a short internship at Ruffian Games as a concept artist, as well as a brief appearance as a tester at YoYo Games.
For this project, my aim was to create a modular environment with a special emphasis on colour, as well as designing the 2D art pieces such as the shop signs, but as my final project for university as well as my first real 3D project, my primary goal was to learn as much as I could about the environment art process.
I started to seriously consider learning 3D just under a year ago when I decided I wanted to create an environment for my honours year project. Up until then the only 3D modelling experience I had were small tasks for university and I had only basic, scattered knowledge of 3D in general – let alone game engines. At the time, I had no idea what I was getting myself into as there were still so many basic things I wasn’t fully aware of, but I felt that I really needed to fill the gaps in my knowledge and I wanted to push myself. Seeing fellow students go through their learning processes was a huge inspiration, particularly Kevin McKenna who was also pursuing environment art for his final year project, and Yannic Kawan, who was responsible for one of my ‘epiphany’ moments when I was first learning PBR.
When it came to choosing the theme for this project, I realised that urban environments have always inspired me, especially narrow walkways and cobbled streets, so I thought I’d go with my instincts and make what inspires me most – when you are learning something new it always helps if you love what you are making, especially for those dark moments when you’re pulling your hair out trying to get past a problem you can’t fix.
Working in Unreal 4, there were challenges getting the modular pieces to light seamlessly and ensuring you are modeling exactly to grid. Whole numbers and accurate lightmap UVs are definitely important here.
There is also the issue of making it look good, modularity can easily make your environment look bulky like it’s made from Lego. Continually testing and iterating on pieces is key.
Most of the building pieces are quite large, it’s not like other modular environments where the building facades are broken down into very small, one-sided pieces. My reason for keeping the pieces so large is because it was easy to test different level layouts and I could very quickly create a bank of pre-made buildings that I could simply clone and drag into the level.
I also decided to keep the bottom and top faces on some of the pieces, this was because I have some of them overhanging and protruding outward so you can occasionally see the bottom face. This was to help break up the surfaces and stop them looking uniform. Even though my method was more expensive, I knew the limits of this project so I figured I’d allow myself to make the level building process that little bit easier.
The 2D art pieces in the scene were a lot of fun to make. My inspiration for those were mainly vintage shop signs, but I wanted to have a mix of different aesthetics including some modern, neon signs.
I always felt that things like shop signs in game environments can really add an element of fiction, and this was something I always wanted to try. The 2D Art pieces were also helpful in emphasising the colour themes that run throughout the project.
The materials for the buildings are very basic, the colours are essentially just solid, with a simple grungy normal map and a roughness map. All of the wall surfaces have the same basic plaster material I painted on all the meshes in Substance Painter, but a few of the buildings, such as the building with the archway, have small details like cracked paint. I didn’t want the textures to take away from the strong colour palette, yet having them completely textureless looked wrong.
The plan for the colour palette was to have three distinct ‘zones’ with unique colour themes.
The road and pavement materials use Parallax Occlusion maps, to give them a fake 3D look. I started it as an experiment but I decided I quite liked how it looked.
For the skyscrapers, I used a simple, small tiling material and BSP brushes for the geometry. The good thing about BSPs is the ability to stretch, scale and rotate the material to create different shapes.
The environment is lit with a simple directional light, a sky light, a bit of volumetric fog and a few point lights, particularly in front of the neon signs. I added the fog when I added the skyscrapers to obscure them a little as they were originally too visually intrusive.
On top of the lights, some post processing volumes in each of the different zones help intensify the colours and darken/lighten the screen as you move through the level.
Using modular pieces made it incredibly easy to quickly generate complex layouts and it makes the iteration process much easier. It gives you a level of flexibility that you would not get if you made each asset bespoke, however once you are settled on a map layout, you could easily go in and delete unused faces and tweak the model to fit its designated space better.
For games, this is certainly an effective way to approach environment art, especially because things change all the time in game development and it’s important to not put all your eggs in one basket. This approach would give level designers more to work with and it means you can generate environments faster – and the more you test and iterate, the better the final result will be.
Work on this project has been almost full-time for around six months, and a lot of that time was spent contending with errors and gaps in my own knowledge. If I were to start this project all over again with the knowledge I have gained in the process I would definitely do things differently, but overall this experience has been invaluable to my development as a 3D artist.