AI applications are only beginning to materialize. It's so exciting what neural networks will do to art workflows in the coming years.
W T F!!! Why, really don't understand, EA = just a fucking looser company!
Amazing... Congratulations for the new way to show information.. I hope they could use this to teaching.
Melissa Buffaloe shared some insights on the production of the amazing open world game Middle-Earth: Shadow of War.
My name is Melissa Buffaloe and I’ve been an environment artist for just over a decade now. I grew up in North Carolina where I learned 3D and worked on PC mods until I got my first game job at Gray Matter in Los Angeles. I’ve helped ship quite a few games including several in the Call of Duty, Killzone, and Battlefield franchises, but I have recently moved away from first-person shooters and joined Monolith where we just finished up Shadow of War.
Shadow of War
My main task for Shadow of War ws building Minas Ithil and it was definitely a challenging beast! Luckily, our Lead Designer, Sofie Widyanto was an architect in a previous life and she was in charge of “blocking in” the entire zone. I use the phrase “blocking in” loosely because most of the time her blockouts were pretty detailed and they were already set up for movement meaning that the player and AI could climb and interact with the pieces quite well.
It was my job to then take her Minas Ithil blockout and start making it look like a city under siege. I began to model a bunch of destruction into her buildings and worked with our movement designers to make sure they remained climbable. We had to see how far we could push the damage without breaking player movement. Some pieces like the domed tower below even required implementing new types of player behaviors. Talion had never climbed over a dome before, so I worked with Advanced Designer, Scott Cummings who’s in charge of implementing new player behaviors to tackle issues like this.
Once I got all of the blockouts damaged and the movement over the pieces was solid, I took a few of those buildings to completion in collaboration with a texture artist, and in the case of Minas Ithil this was Josh Lynch. We worked together to develop texture sets to contrast the more upscale buildings of the higher tiers in the city against the lower class styles below where the peasants would live:
Most of his work was done procedurally in Substance Designer, or on occasion, I would model out a piece of geometry for a trim or an ornate window frame and give it to him to use as a height map. After I finished a few visual target pieces with his textures, we sent these buildings as reference to an outsource team who was to finish up the newly damaged blockouts in the same style. We had regular meetings to gauge the progress and give feedback while I moved on to working on the terrain and buttoning up all of the rest of the geometry that was more unique to the city. With these pieces the challenges were pretty much the same in that I had to figure out how to make everything look really damaged without breaking the movement. I definitely wanted to avoid situations like having a player getting hung up on odd rubble collision while trying to run into a battle. As we got closer to wrapping everything up I spent most of my time working with our Tech Art team to get Minas Ithil in budget and running smoothly on all platforms.
We have a bit of a mixed approach when it comes to building modularly and this was especially the case with Minas Ithil. We tried to keep it somewhat loose early on because there were a few major technical questions we had to answer before things got nailed down too tightly. This was the first large city in the Shadow games and that alone made it a bit of a technical challenge. Minas Ithil needed to feel like a densely packed battleground of a city that was all still climbable and we wanted to see how far we could push it. To add to the challenge, it was also laid out in four ascending tiers meaning that the higher you go, the more of the world you can see. We balanced this a bit by having the lower tiers of the city be the focus of most of the destruction and debris, and pulled it back a bit as you get into the upper tiers. At those higher points we could drop out some of the smaller props below. We had also considered bringing the drake into the zone but in the end it was too big of a risk just because of how much we had jammed into the world already. So in large part, because of these types of technical challenges, our approach to modularity changed a bit over the course of development.
Our Lead Designer’s blockouts were set up for quick iteration. A lot of them were large modular city blocks that we could reuse in clever ways to hide the repetition, so as a result of this setup the city started to come together really quickly as various assets in the level were being modeled out and textured. We could also see pretty early on whether or not more variety would be needed and get a good look at the big picture read of the shapes in the level.
As things got further into development I worked with our technical artists to figure out exactly what problems we were running into and where they were happening and optimize where I could to help fix those issues. I may need to rearrange how a few modular pieces are set up, or combine whole chunks of the level, or lose certain assets entirely. To complicate this just a bit, there are also various other considerations to keep in mind when making any changes to modular kits. For example, each railing piece in Minas Ithil was at one time an individual asset, and while this setup worked great for our movement designers so that they could move single railings in the level it could be pretty heavy on the draw calls when looking out over a long vista. Our Technical Art Director, Dan Thibadeau suggested that I combine some of the modular railings into a few unique assets to reduce draw calls in certain areas. Knowing that the movement designers liked having the freedom to adjust these railings, I waited to make the change until they were pretty happy with the placement. In the end it’s really just a balancing act, but basically in the early stages we keep things modular for the sake of speed and iteration and towards the end we adjust where needed for variety and performance. As long as we keep the level team in the loop, we’re able to adjust modular kits as needed for performance, and hopefully not create additional rework.
Telling a story with the environments
It’s very important to our World Art team that we tell a story with the environments, even down to the seemingly minor details. Any time I talked to the texture artist about his materials he always had a story about how the stones got so dirty, or why the grout was so thin/thick. He really liked to think about those little details that brought his textures to life. This approach also applies to how we detail out the environments with our actual object placement in the world. The assets need to be placed in ways that work for gameplay first, and the movement designers are incredibly familiar with how the action flows in each zone. For Minas Ithil this was done by Associate Designer Marshall Parsons who thought about the stories behind the city quite a bit as he began to detail the areas out. The Art Direction goal was to assign different narrative and visual themes in the spirit of creating a believable city. For example; we created a section where there was an open market for trade, another section had lush parks and government buildings for the wealthier residents, and so on. Marshall placed props mainly based on these themes and on the gameplay of each area. Once an area starts getting filled with set dressing we still want Talion to be able to move through it smoothly and this can get complicated. The movement designers start with broad strokes placing large props that ideally support both the theme of an area and create interesting silhouettes, and then they move onto placing smaller props that continue to reinforce the themes and tell little stories about the environment.
The scale of most everything in our game needs to feel as grounded in reality as possible. So we do aim for real world scale, but it tends to be dictated in large part by the movement metrics. We have a detailed guide that was created during Shadow of Mordor and was expanded during the sequel which defines various metrics like how far apart handholds need to be, or what range of angles they can be rotated to, how small playspaces can be before they don’t work for AI anymore, and so on. This guide has a pretty big impact on the scale of our environments.
Quite often we find that we need to make changes to the original concept to accommodate gameplay so it’s up to the artist to try to find creative solutions to marry the art and design goals. Early on when I began to work on the damage for one of the big domed towers in Minas Ithil I had to redesign the top platform to allow for the graug as we expanded the areas he could reach. In Shadow of War the graug can get to almost all of the areas Talion can, and it definitely had an impact on the scale as he is not always easy to fit. Here are a couple of early variants I made of the tower as I was trying to solve several movement problems:
The damaged dome on this piece was also quite a moment puzzle because Talion can’t climb round surfaces like this, so in the image on the far right you can see how I was starting to think about how to portray damage from warfare while still allowing Talion to still climb to the top.
For me personally the biggest challenge on this project was switching from working solely on first person shooters where I only had to worry about things like cover height and lines of sight, to a third person, open world environment where almost every object is climbable. Artists can’t get away with quite as much when the player can not only get up close to everything you make but can interact with the art as well. It’s been a fun and challenging transition for me, and I’ve been able to meet these challenges in large part thanks to an incredible team of experienced developers who have helped me learn the ropes of creating art that’s so interactive. I have to think about each piece in a different way because the gameplay has to be considered and really thought through so early on in individual asset creation, and that’s definitely new to me. It just adds a new set of rules that I have to now take into consideration when creating art, making it a bit tougher to create assets that work for both our visual and design goals. Luckily, Monolith is a pretty collaborative environment where no one is ever too busy to lend a hand and offer feedback or advice, or even create a new type of player behavior if needed.