Ancient Temple Entrance: ZBrush to 3ds Max Workflow

Introduction

My name is Vincent Moubeche, I'm a 3D Environment Artist living in Montreal for the past 3 years. I have been working in the 3D field for almost 9 years after my graduation from a 3D animation school in France.

For 3 years, I worked for an architecture company in Bordeaux where I learned a lot on how to work in a team. At the same time in 2014, I decided to work on personal projects with UDK and learn my way to the video games (with forums like polycount, YouTube videos, and UDK wiki).

January 2015 was the year I joined Eko Software, an indie company where I was an Environment Artist on How to Survive 2. From there, I could grow as a 3D artist and managed to find my way to Ubisoft Annecy where I worked as a Level Artist on Steep. After Annecy, I ended up in Ubisoft Montreal a year later where I worked on Watch Dogs: Legion as a 3D Modeler with lots of talented people.

Since then, I would say I’ve been an environment artist, half level artist, half 3D modeler with an ounce of level design in my veins.

Ancient Temple Entrance: Concept & Goals

It all started with a dare from a friend; I did not make any environments for a long time and I needed the motivation to do so. At the same time, another friend of mine went to Santa Monica Studio and so making a God of War inspired project was an easy choice.

I spent a lot of time choosing the right concept (it takes motivation to make 3D stuff after 10 hours of doing it at work even if you love it) and that is how I ended up with Florent Lebrun’s piece. It was far from what I had in my portfolio and different enough from my comfort zone, so it was perfect.

The goal was simple: have fun, show something different, and work on something I am not especially familiar with (natural environment with giant man-made stuff) in a short period of time (and all that with ZBrush). Florent’s concept was perfect for that purpose, it had clean repetitive shapes, details on facades, and nice gigantic proportions.

Preparing for the Adventure

With the concept chosen, as most artists usually do, I gathered a bunch of references from video games and photos. When it came to gorgeous environments, the main inspirations were Uncharted and God of War. They manage to create that feeling of immensity where the player is just a tiny piece in a gigantic environment and I wanted to capture that feeling!

First Steps

First of all, I did not invent anything new. My workflow is quite common for the industry, and I just wanted to focus on art.

With that being said, one of the things we don’t stress enough is to keep everything clean. When it comes to projects, I keep my folders in order, everything is named and classified logically. Of course, it takes some time but when you have to find something later, you will be happy to have everything in order.

At first, I took some time to define the repetitive shapes: pillars, columns, trims around the building area, - and made some placeholders for the whole scene without any modular intentions. These blockout pieces would be used in the first pass for the proportions and a quick preview in Unreal.

From those meshes, I did a quick mockup in Unreal to get the basic level art of the scene. For the terrain and rock formations, I simply used cubes and the landscape from Unreal. That tool allowed me to work quickly on the slopes and get a good sense of proportions.

Of course, mistakes happen sometimes and the picture below shows why that phase of blocking is important. The proportions were off but a 50% scale on the Z-axis corrected everything! It is easier to fix those stuff on placeholders instead of final assets.

With everything in place, I could start the real asset production.

In ZBrush

From the generic placeholders, I split the building into modular pieces. Those same pieces would be used as base meshes in ZBrush later on.

In such environments, you have plenty of choices on how to deal with it. Sometimes, it is defined in the production requirements but here, I was free to do whatever I wanted!

Instead of having just one trim for everything and using it in overlay on more generic meshes, I decided to sculpt everything mesh by mesh. This way, I had more control over the meshes and could add details on certain pieces easily.

Once in ZBrush, I used the classic Dynamesh\Trim combo :

• I dynameshed my newly imported mesh to get a nice even topology.
• I simply used Trims and Clay brushes: Clay to remove big chunks of polygons and TrimSmoothBorder with a Square Alpha to get those hard cuts.
• I also used DamStandard to make some cracks when needed.

I kept in mind that the building pieces would be seen from afar so I had to exaggerate some silhouette details. For the smaller details, I trusted the normal details of my material to do the job for me.

Also, since everything would be used a lot around the scene, I avoided very recognizable details like big cracks or damages. In the end, it is just a good balance between adding details and making those meshes generic enough.

Final Result

With my tileables and masks ready, I just needed to import everything in UE4.

Following the Moss Material tutorial from Lukas’ Ramblings, I created my base material used on almost all of my meshes:

In today's productions, it is common to use 2k texture for 4m so I kept that value. It allowed me to keep a constant ratio on all of my assets.

Work on Materials

For all the assets, I used at least 2 software solutions:

• Substance Painter for baking and texturing small props
• Quixel Mixer to create my tileable textures

In Substance Painter

I had two categories of assets: gigantic ones with two UVs and all the others with only one UV (plants and props; I’ll come back to them later). In the case of gigantic props, Substance Painter was really useful for creating the mask that I would use to blend my tileables.

The easiest way to get that mask was to create an RGB8 Channel and then work on layers for each (Red, Green, and Blue). This way, I could directly do an export from Substance Painter to UE4 without recomposing my RGB mask by hand. To make that process even easier, I made this RGB Mask a smart material that I could use on all the pieces in the future.

Mixer to the Rescue

I heard about Quixel Mixer from my friend and decided to give it a go for a simple reason: for me, texture creation has become a job on its own and it would have taken me way more time in Substance Designer. Also, I wanted to focus on the modeling part more. Mixer allowed me to blend, rework, and adjust my tileables very quickly.

Being fairly new to Mixer, I used it the same way I would use Substance Painter. First, I blended some rock textures from Megascans to break the tiling and add some interesting details. Later, I created some solid layers for the color and roughness with mask stacks using underlayer information.

It didn’t take me long to create my first and most used tileable. The software, though still in beta, was already very powerful and allowed creating lots of variations of textures for the rocks, grass, or even decals.

Final Result

With my tileables and masks ready, I just needed to import everything in UE4.

Following the Moss Material tutorial from Lukas’ Ramblings, I created my base material used on almost all of my meshes:

The same process was repeated for all gigantic meshes and I made most of the scene this way.

Plant and Props

The plants and props followed a way simpler process:

For the plants, I created an atlas for leaves in ZBrush, baked it in Substance, and recomposed all branches in Max manually. To help me with the trees, I used GrowFX, a powerful tool for that kind of stuff. It allowed me to make quick variations of trees and I especially loved it for the branch placement and its magnet tool.

The props, on the other hand, went through the usual ZBrush to Max to Substance Painter workflow and were finished in UE4 with a unique texture.

From a Concept to a Personal Piece

After 4 weeks, the project took a radical turn in terms of art direction. I wanted to add more details to the vegetation and change the colors to add more contrast to the scene.

From the beginning, I knew the vegetation would be a problem since the concept didn’t have that much detail except for the grass and small ground plants.

So from there, I developed my own idea of the place, forgetting the Indiana Jones theme and leaning towards an abandoned oasis type of story.

The colors changed, the water area got smaller and the vegetation was not so present anymore. The lighting, in the end, was there just to frame that gigantic door even more. The setup was fairly simple since I only used one directional light and an HDR map. I just needed to work on the big mountain you don't see to have a nice shadow to frame the door and that’s it.

When It Is Time to Stop

Knowing when to stop is hard when you are working on personal projects: you don't have any deadline, only the one you set for yourself.

This project was meant to be short and that is why I stopped after 5 weeks; it was enough and I liked the result.

After looking at it 3 weeks later, I saw some stuff I'd change of course, but that’s why we start new projects: to make something better than the previous work, or at least different.

If you're reading this, thank you for your time, and thank you for enjoying my work. I hope I can show you more in the future.

Vincent Moubeche, 3D Environment Artist

Interview conducted by Arti Sergeev