We’ve talked with Jon Arellano about his approach to the production of rich and colorful 3d environments. Currently, Jon is working on the huge MMOG from Intrepid Studios as the senior environment artist.
Hello everyone, my name is Jon Arellano. I am a 3D environment artist from San Diego, CA. I graduated from the art institute of San Diego with a bachelor’s in Game Art and Design. While I was in school I was a part of an Sony Online Entertainment (SOE) mentorship project. Our team worked on creating a node in the style of Everquest. This was my first big project, and where I first began to take on a role, as an environment artist. That eventually lead to an internship at SOE. Since graduating, I’ve mostly worked on small projects. After a few months I got an amazing opportunity to work with Intrepid Studios. I’ve been there for a little over a year now and we are currently developing our first project, Ashes of Creation.
I believe it actually started for me back in College. I was in the last few years of my schooling, and I needed to focus on what I wanted to do as an artist in the industry. I wasn’t a particularly strong modeler or texture artist at the time, but I quickly learned I could create a mood with my environments using fairly simple pieces. This is a similar technique used in digital painting. I wanted to take that into my 3D scenes. I always loved art that was moody and made you feel something when you saw it. I tried to replicate that in my own work by focusing on the overall feeling, composition, and color of the scene. I’ve learned so much since then, both technically and artistically and my passion for creating environments this way has only grown.
I usually begin by looking for an idea or theme for what I would like to create. Usually through concept images. Once I have those, I begin looking for different compositions and gathering more reference. I then try and merge 2 or more together, usually to create something unique. I like this method because it gives me the freedom to change things along the way and not feel so tied to one concept or idea. Through photobashing or concepting I find something that I can go off of. Once I have that I move onto planning: pieces/models, texture sheets, and personal milestones. I will then work in those milestones usually in this order:
- Blockout scene
- Create all models
- Create all texture/trim sheets
- UV unwrap
- Create props
- Add effects
After every milestone I will usually take a pass at lighting and composition adjustments, in order to ensure the scene is still working as a whole. These are also great times for feedback and adjustments. As my progress shots show I like to change things around, sometimes drastically. Throughout the entire project, I am never afraid to make big changes to improve the overall piece.
The Ancient Civilization Environment
For this challenge I saw it as a great way to practice merging ideas together. I had an architecture type I wanted to use and I began by constructing parts such as pillars and arches. I also new I wanted to incorporate gears and mechanisms showing technological advancements. On top of that it needed to feel like an ancient place. I had already created some large basic shapes for blockout that I had used to get the basic composition. I began by using my constructing parts to piece together structures that resembled the shapes of the blockouts. I knew how many pieces I needed for this environment based on my rough concept and planning. The few pieces that needed to be broken, I simply deconstructed from the models I made. A lot of the gears I placed within the engine in order to have more control over the composition. Lastly I used material overlays and material blending to make sure all the elements tied together.
There are about 20 or so different meshes. I did work with ZBrush, but only on the main props. Everything else was done using Maya, Substance Painter, and Photoshop. The plants were all done with photo references mainly using substance painter to create normals and roughness. All the meshes for plants were done in maya. The metal gears were done in Maya and then using Substance Painter to create height detail that I then baked and textured. The buildings and base were all done in maya and used trim and tileable texture sheets. The main clock in the center was done like the gears. All rock like structures were done in ZBrush and textured in Substance Painter.
I didn’t use Megascans for this project. Most everything was either done from scratch or built off of older models. The shader for the grass is a modified version to one that Unreal has in their open world demo. The plants I built the old fashioned way using photo references and the 1 rock I used I sculpted for this project. The shader for the moss I learned from a tutorial on material layer blends that I found on youtube.
When I begin to texture I will typically create alphas that I will use in the theme of my environment. I will then create trims designs using those alphas. I also save those to use as stamps in Substance Painter or ZBrush if necessary. I try to go about creating texture/trim sheets and tilables first. My process is to design in grayscale using photoshop to create a heightmap. I then plug that into a layer within Substance using only the height info. I then export my height info and normal map and bring them into substance designer. From there I use the node converters to produce an AO map and a Curvature that is based on my design.. I’ll then export and plug those back into substance painter’s material slots. So I have a normal map, AO map, and a Curve map all from the height I created in within photoshop. From here I simply bake out the other maps such as the world space, thickness, and position. This allows me to utilize the generators Substance Painter offers and allows for a quick way to creating some really complex patterns. I will use a similar method for any props I do not wish to create a high poly for, but still require complex designs. Almost all materials were made using this or a similar method, aside from my Zbrush sculpts.
I tend to like my environments to have a mood. I feel it helps create a tone and overall feeling for the environment. I will most likely establish that in the concept or reference gathering process. As I continue to create the scene I will always do paintovers to see what could make the scene better. I will also ask for critiques. If the scene is getting to monochromatic I will paint color blotches to see what helps the overall environment. If I like what i’ve painted over then I think to myself “Okay, what can I put here that can not only satisfy my color, but also makes sense.” I will try to reuse models and if that doesn’t work I will create something brand new.
I am a sucker for having a powerful light source that dominates the scene. I feel like it contributes to the composition by helping the viewers know where to focus. I usually have 1 directional to create the main source of light. My fill light is usually my main color and is typically a cooler color. Once I have those I will then use my fog as a color as well. Depending on what I feel will make the scene look better. I will then add other lights if necessary to act as additional bounce lighting or if I believe it will help the overall scene. My colors are usually pretty saturated because I like to error on the side of color. I think about really great painters and how they are not afraid to use colors. I try to bring some of that into my scenes.
Aside from fog I don’t usually mess with atmospheric effects early on. When doing paintovers and composition adjustments I tend to experiment with them in 2D. If it works in the paintover I will try to translate it into 3d. They are the cherry on top so it is best to try and save them for last, but sometimes they play a big part in my composition. So whenever I need to push that I will use them sooner.