Mob Boss

For this piece, I first started with “The Office of the Casino Boss” by Tian Gan as my concept.

I had wanted to make an interior scene with a lot of intricate props, and his wonderful concept seemed perfect for what I wanted to make.

After picking a concept, my first step in starting any scene is to plan and organize, so I started by outlining and color-coding the concept to figure out all the props I wanted to make.

Rather than just making a list, I find this process to be a lot more visual and helpful as it can help me figure out which props take up the most space and thus are essential to finish early. At this point in the process, I also gather material and object reference for the scene.

For this project, I specifically collected a lot of reference for the machine gun, grenades, Versailles wood floor, wallpaper, and chairs.

Next, before I even start modeling any assets, I start to block out my scene in Unreal as well as get some basic lighting done.

This allows me to get a sense of scale and lighting for everything early on so when I do start to slowly add in props and textures, they have context.

Next, I generally export these grey boxes out of Unreal and into Maya for each prop and start modeling from there. I was able to model certain props entirely in Maya like the chandelier and tea table, but assets like the chairs whose cushions I had to sculpt, or the dresser that required a custom curve brush, ended up being made in Zbrush.

Afterwards, I would bake in Marmoset Toolbag and texture in Substance Painter.

Although this is my usual process, I decided to break it for the light machine gun, grenades, and cases. I had been wanting to learn how to hard surface model in Zbrush for quite some time now, so I ended up completely skipping Maya altogether for these props.

It might have been a pain to learn, but Zbrush can be extremely powerful for hard surface in certain situations. Arraymesh for instance is an amazing tool as it allows you to instance subtools which can be really useful for repetitive parts that you want quick iteration on.

I used this specifically on the repeating bolts and holes in my light machine gun as well as the stun grenade. It’s simple, but I also really enjoy not always having to select faces in order to do something to them. With Zbrushes Zmodeler you can instead have something done (like insetting, extruding, etc) by polygroup or polyloop instead of always having to select the faces individually, which ended up saving me a lot of time.

In terms of texturing, except for the wooden floor and wallpaper (which were made in Designer), almost every prop was textured in Substance Painter. Many of the props were too unique to just have a tiling Designer texture, and I wanted to be able to get some unique weathering effects on my props, so Painter was the best way to go.

I find one’s strength in texturing relies heavily on their reference. Guesswork often ends up with questionable results, so I always make sure I have a bunch of references open for whatever I am texturing. In Painter, I tend to texture “chronologically.” By this, I mean I start by making what an object would look like coming right out of the workshop/ factory.

Then, I ask myself what environment it has been in and for how long. Once I know that I can start to add worn edges, dust, water stains, fingerprints, etc., but I always try to add these wear and tear details where they make sense environmentally.

Finally, there is lighting.

For most of the project I just stick with whatever lighting I set up during my blockout phase, however, once all the props are in, I go through with a final lighting pass.

I had a problem with the original concept, however, as it has light streaming in from just one of the windows while the other is blocked by curtains. Although it may have looked nice in the concept, that setup left the room extremely dark, so I ended up deciding to leave both curtains open in my scene.

I also ended up including several lights in the room with cast shadows turned off to give more ambient light and reflections, otherwise, anything not directly touched by the light coming out of the windows would have been black. For the god rays, I ended up making a plane with the vertex color of the edges set to black in Maya, then used that vertex color along with a panning black and white “beam” texture I made to drive opacity in Unreal.

Finally, I added some emissive colour to make it brighter and more yellow. Also, although it may look like light, the orange “light” that appears to be visible behind the curtains was actually an emissive texture driven by vertex paint. I attempted to get subsurface to work on the curtain, but I found that this solution ended up being more optimized and easier to art direct.

There were certain parts where the light wouldn’t have scattered through the curtains, but with this method, I could paint it where it looked best artistically, rather than where it should have been realistically. Moments like these are what make us artists, not some machines simulating realism.

Overall, my main challenge for this scene was just dealing with the sheer amount of props.

To finish on time, I had averaged at least five completed props a week, most of them being extremely intricate and detailed. Regardless, this project certainly made clear (yet again) the importance of planning. If I had not outlined what props I had to make and by when I would have been completely lost and even more behind schedule. If I can give any advice to future artists, it is to plan well and plan early.

I hope you learned something from reading about my process in this article. If you are interested, you can also find the rest of my work on ArtStation(https://www.artstation.com/rhiannonremo).

Thank you!

Rhiannon Remo