@Tristan: I studied computergrafics for 5 years. I'm making 3D art now since about half a year fulltime, but I had some experience before that. Its hard to focus on one thing, it took me half a year to understand most of the vegetation creation pipelines. For speeding up your workflow maybe spend a bit time with the megascans library. Making 3D vegetation starts from going outside for photoscanns to profiling your assets. Start with one thing and master this. @Maxime: The difference between my technique and Z-passing on distant objects is quiet the same. (- the higher vertex count) I would start using this at about 10-15m+. In this inner radius you are using (mostly high) cascaded shadows, the less the shader complexety in this areas, the less the shader instructions. When I started this project, the polycount was a bit to high. Now I found the best balance between a "lowpoly" mesh and the less possible overdraw. The conclusion of this technique is easily using a slightly higher vertex count on the mesh for reducing the quad overdraw and shader complexity. In matters visual quality a "high poly" plant will allways look better than a blade of grass on a plane.
Is this not like gear VR or anything else
Joel Westman talked about his incredible City Back Alley initially started in CryEngine but later moved to UE4 and his experience with Substance Designer.
My first real industry gig was as a QA Tester at Crytek working on Crysis 3 back in 2012 which was a super exciting and eye-opening experience. I ended up spending 5 years at the Studio in Frankfurt where I also transitioned into a 3D Artist position and moved through different projects before making the move back home.
Goals of the Project
This Project started with having three main goals:
- Make a bigger environment for my portfolio.
- Start learning Substance Designer.
- Start learning Unreal Engine 4.
Short rundown as to why these three goals. At the time I didn’t really have an environment of a decent size showing off my ability to build structures covering a big area using only a few tiling textures and structures. I also had someone ask about this one time when I did an interview for a company so that also reinforced my belief that this was something I really needed to demonstrate with the next thing I did.
Substance Designer was quickly becoming an industry standard in the hay days of this project and I saw material spheres popping up here and there so I felt like that was something I needed to delve into as well.
The project started out in CryEngine because it was the game engine I was most familiar with from before. But as I transitioned to a new studio that uses UE4 as one of their game engines I thought I might as well start playing around with it at home and hence I decided to move my project over to UE4 instead. On top of that I also just saw all these cool projects being done with UE4 and with all the nice updates Epic keep rolling out there were plenty of reasons to get into it.
Early days, CryEngine:
Since I was venturing into a lot of unknowns when it came to pipeline stuff I wasn’t too worried about researching and coming up with a super original concept and instead went with something pretty generic, a City Back Alley. Which doesn’t necessarily have to be a bad thing, sure it might not stand out on the art communities but on the other hand a lot of games take place in cities so to show that you can make those sort of assets might be helpful for potential employers and also if you maybe want to sell your assets on UE4 Marketplace, Gumroad or whatever you’re probably more likely to create something of value to the community with that rather than if you make some sort of super specific Sci-Fi style environment that will probably have limited use to most users. However, in general, I think you should just create whatever you want without really thinking about that stuff but it might be something worth considering depending on what you’re after.
I started out with making the obligatory red brick material just like most other newbs when they first decide to dip their feet in Substance Designer. It’s basically the 2018 version of the standard portfolio crate/barrel prop.
Once the material looked decent I started putting some simple buildings together with the brick texture applied. At this point, I’m mainly working on the composition, mood and focal points.
As far as the blockout went I only blocked out the main buildings in the beginning in one go and then I progressively added blockout bits for whatever props and architectural elements I wanted in the scene. I don’t really have a full whitebox level that I then go over and make art for, it happens more gradually, which is nice because it means I’ll get to jump around a bit between tasks.
For most of the architectural pieces, I made modular sets in 3dsMax and used them to construct a few different versions of buildings that I then exported as combined meshes over to Unreal.
If I were to do a similar type environment today I’d push the architecture further, add more architectural details and a few more distinctly different versions of buildings.
I had two different versions of the wall, one that you can’t see through and another that has fake interiors. I threw together some simple rooms in UE4 that I took screenshots of and then plugged into a material I made that uses bump offset to make the image move creating this layering effect. It makes it behave sort of like it would in real life.
If I’ll do something similar again where this technique would be useful I think I’d play around with rendering out a cubemap as a render target. I’ll feed that into a material using the interior cubemap node instead to get a more accurate looking parallax effect from all angles.
In regards to the metals in the scene, there’s nothing crazy behind it really, I feel like it’s the same as with anything 3D art related, each step matters and shouldn’t be too lopsided in either way. So, in this case, the model needs a decent amount of detail, same with the textures, they need little nuances, slight roughness variations, and small imperfections etc to make them look interesting and “rich”. Then that’s just half the battle, you need to spend some time on the lighting and making sure you’re cube/sphere -maps are doing all that good texture work justice in your final scene.
Carefully placed Cube/Sphere-maps can have a big impact on your metals:
And no matter what texturing software I use (for this project SD/SP) I always make sure to send stuff over to the game engine as soon as I have a base texture pass done, like really early and then I keep jumping back and forth into insanity. Just to make sure it really looks like I want it to in the final renderer, which is all that really matters in the end anyway. For a lot of people that might just sound like common sense but I’ve seen people who spend a lot of time in their texturing-app only to later hit export and say “it looks much better in Painter/Designer”. For the most part, they match pretty close, but there will always be small differences. You might have a very bright environment map in Painter but you’re exporting to a night scene in UE4 and the roughness might not be coming through as much in one software vs the other etc. So it’s just better to keep re-checking when you’re doing adjustments until you’re happy with the final result.
Everything is dynamic. I went with fully dynamic lighting simply because it’s what I’ve been used to and I wanted to test out how well it works inside Unreal with Distance Field AO and all that jazz.
I want to do some experimenting with baked lighting in the future though, I see so many awesomely baked-lit environments out there.
When I think about creating a city alley environment, to me it’s all about capturing this sort of forgotten backside of a city where you pretty much only go if you have to. So all those details had to serve that vision, these are dirty run down places that no one cares that much about. So naturally, there will be a lot of dirt and trash, streets that might not necessarily get the love it needs. Working a lot with decals to dress up areas helped convey this. Adding little cracks in the road, graffiti on walls, trash around the dumpsters etc. Volumetric Fog in combination with different lights helped to make it a bit eerier as well.
It took me about a year to finish the scene and since this was my first project in UE4 there was a lot of trial and error and a lot of experimentation involved. I think the most challenging thing was to create all the different shaders/materials from scratch, coming from CryEngine which has a set of premade shaders to work with, I’d say that was probably the biggest challenge.
Joel Westman, Environment Artist at Starbreeze Studios
Interview conducted by Kirill Tokarev