The Crimson Path

Goals

For this project, the main goals I set out for myself were to make a highly modular environment with a unique shape language that resembles Middle Eastern motifs.

I knew from the start I wanted to create a long-lost place that was overrun by sand, as this creates a very interesting composition in my opinion.

I did not rely on a specific concept, which, in hindsight, I regret a bit, because working full-time, I was only able to work on this at weekends, which meant figuring out any uncertainty took a lot of creative energy.

Another main goal for this project was to transition to Blender from Maya as my main tool for modelling. I’ve been wanting to do this for a while and this project was a perfect opportunity to make the jump.

References

For references, I used both real-life references and game references.

I often hear people say it’s not good to use games as a reference, but I always found it useful to keep some as a quality bar for high-fidelity work.

Blockout

I wanted a workflow for my blockout that allows me to make very quick iterations, as I didn’t have an exact concept I was matching, so I knew I’d make a lot of iterations and it could be time-consuming.

I ended up using a USD workflow, which was developed by Pixar, I believe, to transport entire scenes between different software. I’m not entirely sure, but it worked great. With a minimal setup, you basically have a dynamic preview of your Blender scene inside Unreal.

For the blockout, my main goal was to find an interesting shape language for my architecture and to lock down my composition.

I also did some quick sculpt tests on some pieces because it takes no effort in Blender to get some very basic surface and edge breakup, and I also relied on modifiers as much as possible to speed up my iterations.

Sculpting

Once I was happy with my pieces and my composition, I moved forward to sculpt all the assets. I ended up having quite a few pieces, so going through and hand sculpting them seemed like a never-ending task and I also wanted to find a rather non-destructive workflow.

I looked into procedural techniques like creating a modifier stack in Blender and using displaced skinned edges as booleans, or setting up something similar in geometry nodes, but I found most of these procedural solutions either work only on simple meshes or they break modularity, so I came up with a different idea inspired by the boolean workflow showcased by the great artists at Machine Games in this video.

Since this workflow is also better for breaking up large rather than simple surfaces and can become tedious with small corners and details, I decided to turn this into an IMM brush workflow inside ZBrush.

The idea is to get a highly detailed rock, which will be turned into an IMM brush, and this allows you to drag out rocks on the surface of a mesh. You then set it to a boolean and decrease the depth of your IMM brush so that it will drag out from a bit under the surface – and there you go.

You can very quickly chip edges and change the scale of the chipping based on your needs, and it’s highly art-directable as it is all live until you make your boolean mesh, and you can go back any time to adjust your cutters. All the pieces have their own polygroup as well, so it is also easy to isolate them and fine-tune them.

This allowed me to very quickly sculpt all my meshes that otherwise would’ve taken a very long time. The main downside to this workflow is that, depending on the size of your cutter, you can have different scale detail on your final mesh, but once textured it is quite hard to see it.

You can also turn all the chipped edges into their own polygroup in the end, so you will already have your mask for a broken stone material.

Texturing & Materials

For my texturing process, I chose tileable textures and RGB masks as they are very easy to tweak and create a cohesive environment. It is a bit of a hassle to set up with material layering in Unreal, but once you have it, it is easy to reuse. You can also hardwire the logic in a master shader as well, but that can be a spaghetti fest, so I recommend doing it through the layering system.

I also added a top-down projected sand option in the material and added a few parameters for tweaking textures.

I made a list of materials I’d need and created them in Substance Designer. Here I did not do anything fancy, just some simple noise mixing and usual Designer tricks.

After I had the materials, I went to Substance Painter to create my RGB masks. For this, I have a smart material setup that automatically creates the layers, but if you need something similar, I recommend picking up this package by the awesome Dylan Abernethy.

Here, I used the R channel for dirt, G for edge damage and B for surface wear, and used smart masks and a bit of manual tweaking where needed to not break my modular pieces.

Terrain & Scattering

My terrain was first just a static mesh I quickly blocked out in Blender with sculpting, but later in Unreal, I turned this into a proper terrain and set up a terrain material.

To get some of the wavy breakup in the sand, I ended up making a wavy version and a non-wavy version. I used the plain version as a base and simply painted in the wavy version where I wanted some breakup. I also enabled displacement, which helped sell the effect quite well.

For scattering the rocks, I just used foliage painting and painted them around with different settings until I was happy with the results.

Effects

For my main effects I had the sand blowing in from the large opening on the right and the falling rocks.

For the blowing sand I used the fog cards from Unreal and tweaked the shader to change the direction, as by default it is locked to the Z axis.

To help sell this effect I also added a noise texture panning with a slightly brighter base color to the terrain material to create the illusion of sand blowing close to the surface. It’s a pretty hacky solution but works surprisingly well.

Lighting & Composition

The lighting for this environment was quite simple overall. I had a warm directional light with a bit of increased indirect lighting to get a bit more bounce of warm tones on the underside of the structures, and a slightly colder sky light with a cubemap to balance it out.

I went through a couple of iterations of light sculpting, but most of the attempts just made the composition less focused, so the only thing I changed in the end was the angle of my directional light to highlight the entrance more.

I wanted to keep the colour palette simple to make the shapes and the overall image easy to read. I followed the 60-30-10 rule with the stone material being 60% of the image, the red sand 30% and the sky 10%.

For my composition, I focused on the thirds. I kept the sand to the lower horizontal third, my main highlight to the centre and less detail and light to the upper third to not attract as much attention.

I also placed my main focus point on the left vertical third and isolated my source of light to the opposite vertical third, but I made sure to cut it straight through with the arches to create a straight line pointing towards my main focus point. I kept looking at screenshots scaled down and far away to see if it all reads well

Conclusion

Overall, this project was a bit more exhausting than I initially expected it to be. Managing my time was a bit tricky working full time and having only the weekends and some evenings to chip away at it.

I would’ve definitely benefitted from a clearer idea of what I wanted to make from the start and from sticking to decisions I made, as I kept going back to the drawing board, which drains motivation quickly.

In the end, however, I did manage to make something I’m happy with and the workflows I used were the right choices to push through to the finish line.

For now, I would like to thank Games Artist for the opportunity and thank you for reading this article. If you have any questions about the project or the workflows used, feel free to reach out to me on ArtStation or any of my socials.

https://www.artstation.com/davidszieber

https://www.linkedin.com/in/davidszieber/

https://www.instagram.com/david_szieber/