I asked Twitter if anyone had questions about the art or art process of Firewatch and we got quite a few!
Here is part 1 of the Q&A.
What art tools do you use? (@SofyK_)
We are building the game using the Unity engine. We use Perforce as our version control and it mostly talks to Unity well.
I build 3D assets in Maya and export them as .fbx files, which are very easily imported into Unity. Textures are made in Photoshop, and normal maps are generated using NDO2. Some organic 3D assets, like rocks or fallen logs, are sculpted in Zbrush. Sculpted assets would have their normal maps baked via Xnormal using a high res sculpt and game res model. A few landscape props that are not traversable are made using World Machine.
We also use the following Unity extensions: SECTR, Marmoset Skyshop, Amplify Color, NGUI, Playmaker. The first 3 relate to my work.
What is your workflow from concept to final asset? (@Xocari)
We have a very straightforward process for props. Olly creates a rad concept, posts it in our company chat, we all like/dislike/comment, and when there are enough thumbs ups I model and texture it and put it in game. Olly sometimes paints over the texture for a more painterly pass or if the asset requires specific design, such as the book covers.
In terms of the broader environment, we have a very different process. It is always evolving but this is about what we did for the PAX demo:
- Olly creates concepts for specific locations and or story moments, conveying mostly mood and general spatial vibe, tightly looping in Sean and Jake for story and creative direction throughout.
- Jake and Nels create a “greybox” pass with just a grey landscape and cubes with game pacing and Olly’s loose concepts in mind.
- We take screenshots of the greybox and Olly does paintovers on the screenshots.
- Given all the above, I go ahead and start layering in first-pass art, placing rocks and trees, painting in proper terrain materials and setting up general lighting, trying to match Olly’s paintovers as much as possible.
- The team starts putting in design elements like triggers, dialogue volumes, etc., while art is being brought up.
- We talk about what we like and dislike in weekly playthroughs and address them. Occasional panic ensues.
- Repeat 3-6 times with more paintovers, art details, and design adjustments until roughly shippable.
- We place atmospheric volumes all over the space to fine tune lighting and color grading changes.
- Move on to another area because there is nothing over there.
- Eventually come back for polish passes when the rest of the game exists.
What is your process for trees? (@BryanRenno)
Up til now the trees have been built using Unity’s tree creator. Once the shapes are pretty okay with temp textures, I convert them to meshes so Olly can create final foliage textures easily while seeing his changes in game. Using Olly’s textures I go back and plug them back into the Tree Creator for our trees. We are in the process of moving away from Unity terrain trees, instead only using mesh trees, so Paolo has better control over their shading.
Do you have a favorite tool/technique? (@OIninyo)
Being able to see changes in Maya and Photoshop instantly in Unity. My favorite way to work is to see how everything looks inside the game, so the seamless hot reloading of changes is great. A beautiful looking flat texture might not look great in game. The only way to know is to look at it in game, all the time. As for tool, I really love Xnormal. I find it amazing that it is free software and it just does an amazing job baking out maps.
Has working to the restrictions of the tools negatively affected your creativity? (@Akelaa)
Quite the contrary in my opinion, I do my best work when I’m given constraints. Constraints force me to really understand what I’m doing and prioritize what is most important, both in terms of process and visual result. If I have infinite time and infinite tools to make whatever I want, I’ll probably end up staring at a wall and not create anything at all. That’s why I am no concept artist. ;)
For such a large project with relatively few artists, what sort of practices do you do to increase efficiency in production? (@Atomander)
We have a theee-person art team: Olly does production design, James does animation, and I do environment art. To ensure maximum efficiency, ideally all dependencies are identified and work arranged in the correct order so no one is ever blocked by someone else. Having a clear plan and set goals before you start any work is key to avoid any rework. For example, Olly is usually “ahead of me” in my tasks because my work depends on his concepts. If the concepts aren’t ready, I don’t start. I find something else to do. It is way worse to start on half-baked ideas and then realize later you have no idea what the goal is and then have to redo a large area. Change is easy during concept and greybox phases, so constantly play your game to do course adjustments instead of rework.
How do you hand off to developers to put into game? Maintain documents, verbal, something else? (@simnon13)
I implement environment art directly into the game. For assets that require animation, I will check the fbx into Perforce and tell James it’s ready (also adding a task in Pivotal, our task tracker), he would do his work on it and then implement the animation in game afterwards. For assets that require more engineering hookup, like our mantling obstacles, usually we hook up everything to temp art like a cube. I then go in and replace the cube with better art. We maintain a “wiki” channel on our company chat for any complicated “how-tos.”
How has working on a brand new game from scratch outside of a studio system with legacy tools/assets been? Does the Unity Asset Store help mitigate? (@Siromatic)
Legacy assets only really help if you are making very similar games. There is always the temptation to use old assets to save time, but you are also missing out on the opportunity to make better assets now that you are a better artist. The Unity Asset Store is absolutely great, more for tech add-ons than assets for us. Want a color grading package? $40 and you have it. We are not constrained by what tech we used to have. That said, we are spending a considerable amount to time to build up our own proprietary set of tools specifically for Firewatch.
Firewatch’s gameplay seems tied to the exploration of environments. Do designers ask for sets or do they fit story into your rough sets? (@LareVar)
I think this is mostly answered above regarding our workflow. I will add that art and design is really quite a fluid feedback loop. The primary, critical path of the game environment is driven mostly by story/design, but the secondary areas are mostly free expression by me. Most of the time I just play the game and then go, huh, I want to go over here, and won’t it be cool if this and that is like this. Then I just make it. This in turns get the team excited and sometimes inspire a secondary layer of design and story elements. A lot of dialogue in the game is also inspired by how the art is in the game. For example, I added a vista point off a path and told Sean I want to be able to talk to Delilah about the view, and now Henry can.
Does seeing an object placed in the environment ever lead you to want to make changes so it fits better? (@Rtroock)
Oh definitely. This is why it is so important to make your art assets while constantly viewing it in the game, preferably with different lighting conditions.
Do you find yourself wanting to put small details in that might not be seen in game but make it feel more “real”? (@Rtroock)
I used to do that quite a bit, when I was a younger artist. I would get carried away making an asset because the process is so fun, and then you view it in game and you realize the details you spent 3 hours making amount to maybe 3 pixels on the screen. I don’t do that anymore, because I’m older and wiser, but mostly just lazier.
Are there memes in the game? (@gormanate)
More questions answered tomorrow!