Environment Production Secrets of David Schultz

David Schultz talked to 80.lv about his personal projects and some of the work he did for The Division. David was kind enough to talk about his environment production process and how he works with Substance tools. Amazing production details.

Introduction

I am an Environment Artist at Red Storm Entertainment (Ubisoft) currently working on The Division. I went to undergraduate at the University of North Carolina at Chapel Hill where I studied French and Philosophy. I really enjoyed my time at school, but after getting a little taste of academia, I knew I wanted to get back to making art. In 2004, I started working as a freelance artist on local commercial work in Asheville, NC, where I was learning on the job, working as a 3d/motion graphics artist. After seeing some of the later Unreal Tournaments, Gears of War, and Half Life 2, I realized that video games were starting to look really good and I wanted to try and work in the video game industry. It took me a little while to get to a level where I was hireable (“My high poly work isn’t game ready?”). I started off at Vicious Cycle Software and worked on Earth Defense Force:Insect Armageddon,Ben 10: Omniverse,and then at Funcom, where I worked on The Secret World.

I had the pleasure of working with Joseph Drust and Ty Shelton early on in my career at VCS. They were doing hard surface work in Zbrush, and I knew I wanted to get to a point where I could make models of similar quality. I knew I wasn’t anywhere close to their level, but after we shipped E at Lead ( my first shipped title) we had some down time, and I really went after those guys (Ty especially) to try and get them to teach me as much as they could during that time. I made some really ugly stuff, but it got me to started on making personal projects and pushing myself to learn as much as I can and be a better artist. I am always looking for new software and new methods to improve what I do. I moved around a lot with software early on, but I primarily use 3ds Max, Modo, Zbrush, Substance Painter and Designer, UV Layout, KeyShot, and most recently I’ve been pushing Fusion 360.

Learning

It’s easy when you’re starting out to fall in love with the idea that a new software will make you a better artist and waste time learning new programs, but in the beginning it’s probably better to a pick a few things and get really comfortable with them. If you are learning 3ds Max, don’t switch to Modo until you feel like you’ve found the limits of that program and vice­versa. At the same time, not learning the right programs is a death knell to progress. It’s very important to always be on the lookout for ways to improve your pipeline. If you find yourself doing something that is overly repetitive or something that seems unnecessarily complicated in a particular piece of software, take the time to do some research and figure out how you can make it faster and better.

A lot of the work I’ve done in my personal work is usually more about testing a certain process or design idea. As a result, a lot of these projects take me longer to finish, because there’s a lot more trial and error than I allow myself in my professional work. I really enjoy testing out different ways to do the same thing just to see what produces the best results for the minimal amount of time in production. Sometimes, I do let myself go a little bit further on pieces and just let myself go crazy on detail. I’m not actually OCD, but I get a weird sense of enjoyment from pushing detail as far as I can in certain personal projects. This doesn’t work in a production environment, but it’s fun for me to just do my own thing on the side at times and feed that part of myself.

Gameready Assets

As for making game­ready assets in production, I don’t really have any easy way to retopologize high poly assets. I primarily use Topogun and Zremesher in Zbrush. Topogun hasn’t been updated in a long time, but it’s still better than anything else I’ve used. Zremesher is pretty great for retopologizing more organic shapes. I found out that if you press “alt” before you use Zremesher it uses a different algorithm, and it seems like it does a better job of following the topology of your mesh. Then, I can bring it into 3ds Max and use PolyDraw Optimize to get rid of edge loops, collapse edge rings, and collapse verts. It’s a really amazing tool for optimizing.

I will say though, if you have the tool active, and you click in the viewport on your geo, and you don’t see anything happen, click undo. You probably just collapsed faces on the back of your mesh. It can’t operate on backfaces, and if your camera is too close to your geo, it decides it is occluded and starts working on the part of your mesh that is actually occluded from your view as the user. It can be confusing at first, but once you get the hang of it and just “undo” if you don’t see anything happen after you click, it is hard to beat for game optimization.

Using Substance

I started using Substance Designer a few years ago when I was at Funcom. I was primarily using it in conjunction with Zbrush to make tiling textures, and even though there was a bit of a learning curve (and not nearly as much training material as there is now), I could see that it could be very beneficial in a team environment. Substance Painter came out about the same time I started making textures following the PBR guidelines, and it made that transition so much easier for me than it would have been otherwise. I think the big thing making PBR textures (and I still get hung up on this at times) is making sure your albedos aren’t too dark. As one of my coworkers, Jon Lauf, reminds me from time to time, your albedo should be the color at 100% white light. Setting up your roughness and metalness with Painter is pretty easy. Just look at your guides and their material presets and don’t deviate from those values. There are many good references out there, but if you haven’t looked at his b log, Sebastien Lagarde has tons of valuable info.

In my current workflow being able to make Smart Materials and simple substances for layering textures in Painter that I can save out and re­use any time I have a prop with similar properties is a huge time saver. I see a lot of people throw on a couple of generators and call it a day (and sometimes that is all you need), but you can really build up levels of subtlety setting up a generator on the mask and being able to add in fills and various noises on the actual layers and multiplying them and adding them together. You can set the first layer to multiply, duplicate it, and set the new layer to color or color burn (something pushing saturation), and add a paint on the mask side to this layer and paint back in some areas of color. It’s a bit hard to describe, but it’s all to say Painter makes it so easy to get an amazing base, and you can still build up levels of complexity and subtlety like game artists used to do in Photoshop.

Workflow

Here’s an example of a workflow I’ve used on a few different pieces recently. One of my tasks was to make five or six burned bodies in body bags, and this would have taken me forever in an older workflow to resculpt and repaint all of those details. But, I was able to use a single Marvelous Designer pattern as a base for each of the body bags, some noise generators in Zbrush to get the base deformations of the bodies, the deformities of the skin, and the burned sections of the bags. Then, I made a Smart Material out of some of the materials and generators that come with Painter for the burned skin, and I came up with another one for the burned bags. After all of this was set up for the first one, I was able to produce one new body in roughly a day.

Building Environments

As for thoughts on building out environments, I think the biggest thing you can do is to look at references and begin to build your “visual library,” as Feng Zhu put it. It’s important to continue to build that library by immersing yourself in as much reference as you can, grounding yourself by gaining a general sense of how things work and are put together and familiarizing yourself with the history and practices of design in general. Learn as much as you can about composition and color theory, and not just in terms of lighting, but learning how colors can be used to push elements to the foreground and background, etc. You can get a lot of this from looking at various forms of fine art, photography, movies, etc. I would also say that watching tutorials by concept artists like Feng Zhu, James Paick, and Maciej Kuciara (to name a few) can really help you figure out what they are looking at when they design environments and what you can do when you start constructing to adhere to and build on these same principles and ideas.

Tools are getting better, teams are growing in size and we are getting to the point where environments (for more realistic games, at least) don’t have to be made in a way that is feels as gamey, or where the re­use is as obvious as it used to be. So, really looking at reference and figuring out how the space would actually be set up is very important. Honestly, if you look at most buildings that were designed by an architect worth anything, they should be designed in a way that communicates how traffic should flow and move your eyes through the space in terms of composition. The other thing, I don’t think more inexperienced people do enough is figuring out where and how the camera will move through the space in a game and how that effects how large a space should be and where your detail needs to be. It seems obvious, but I’ve seen it so many times, where someone is spending so much time worried about a certain vista or something up toward the ceiling of an area, when 90% of the player’s time in their space they will be looking at the ground and they have next to nothing of interest to look at. Don’t get me wrong, I’m not saying that vistas aren’t important, because they can be, but you have to spend your time and scene budget on what is seen the most by the player.

Another thing to consider (and in this way games are different from a spectator in a real location), is that the player will experience your environment more than likely while moving through the space quickly and possibly while being shot at or something else that is demanding their attention. So, in that way, we still have to do some things to exaggerate or emphasize where the player should go and what they need to look at to indicate that. A lot of that can be done with repeating forms and/or signs that become familiar to a player and indicate a desired action at a certain point, or, using lighting and contrast to indicate points of interest. In keeping with this, it is also important to remember when looking at your references to know when you can’t add in certain elements or props because they obscure these indicators or confuse the player as to points of interest and desired paths. If you don’t have experience doing this in games, look at other games, and pay attention to the things that help orient you in the space and guide you through a level.

I would also like to say, I see countless forum posts where people are complaining of a lack of inspiration, and it may sound cliche to say, but there should be sources of inspiration for you everywhere. If you find yourself getting bogged down and uninspired you’re either doing it wrong or there’s too much going in your life to allow focus (this can absolutely be beyond your control, I know). If it’s the first, I think a lot of people want to do “x”, but they get started on it, and realize it is going to be an overwhelming amount of work, or they don’t have the skills yet to do it, or they picked something that is actually really boring to them, but they think it will show off certain skills, and so they get frustrated and quit. Try not to let the frustration beat you. Refocus by taking a small prop, a bust, a vignette that has elements you find interesting and use it to try out some new techniques and learn some new software. It’s not only about developing certain skills, but it’s also about finding what you want do. You will get better the more you do something, but if you find something you really like to do, I think your improvement and, hopefully, your satisfaction will be exponential.

Don’t be impatient. Everything takes time and work. Learn from all of the people out there who know more than you and stay open to what they have to teach.

David Schultz, Environment Artist – Red Storm (Ubisoft)

INTERESTING LINKS

BUILDING CATHEDRALS WITH 3DS MAX

ACHIEVING A PERFECT LOOK OF A 3D SCENE

3D MODELING FOR GAMES & FOR SHOW