I started this project in March of 2018 in an attempt to push my 3D art abilities as far as they could go, making something over the course of several weeks. I had just joined an amazing online collective of 3D artists who offered me some technical and stylistic advice (thank you 3DINTERNET) over the course of my work.
This project was my first foray into several tools, including Zbrush by Pixelogic, Quixel’s texture tools, and Octane Render. My workflow generally consisted of: sculpting in Zbrush, then exporting the mesh to Autodesk Maya and unwrapping the UVs (the coordinate system that allows the computer to tell where to place the texture, or the image that wraps around the surface of the 3D model). I would also use the Mental Ray (by NVIDIA) plugin for Maya to bake an ambient occlusion map, when necessary. If the mesh needed cleaning up, or if the mesh was a manufactured shape, rather than something organic (Zbrush is great for organic shapes but I find Maya easier to use for simpler, inorganic meshes), I would use Maya’s sculpting tools as well. Next, I would import the mesh with the ambient occlusion map to Quixel Suite, creating normal maps and color ID maps as necessary. Then I would go about painting the textures, and create final maps for all the necessary nodes in an Octane material. All that was left once all this was done with each component of the Forest scene was to import the model and textures to Cinema 4D, and, if necessary, rig the model with a skeleton so that it could be posed easily. I did this with all the plants and insects in the scene.
I started with basic forms that I already knew how to make in Zbrush and Maya — a rock, a disk for the CD — and moved slowly into more complex modeling and texturing. Examples include mushrooms and plants, and eventually insects, trees, and more. If I ever had a question about how to do something, like texture the shiny side of a CD or model a thin insect wing or scatter leaves on branches, I would consult a tutorial or ask other artists. More process pictures and renders follow:
After the rock came vegetation; fungus, an old stump, plants. Later came leaves, a tree, the fully textured CD, and insects. Each step of the way I learned something new about sculpting, texturing, or rendering. Eventually, I also started experimenting with HDRI environments for realistic lighting and the Octane camera, which allowed me to simulate physical properties of a real camera, even film.
Above: the original sketch of the scene with planned elements labelled | the wireframes of all the models in their semi-final arrangement. Below: A test render of the arrangement using Octane with Directlighting.
Next steps included creating fog, an HDRI (high dynamic range image; in this case a 360 degree capture of a space and its light to be used both for lighting the scene and as a backdrop), and some post-processing in Photoshop. I also edited the arrangement further. At this point, I was reminded that my curiosity regarding this medium is only partially related to the technical aspects; I am also struck by the unique and beautiful quality of the images one can create with 3D rendering — the idyllic, almost hyper-real (“plastic”) nature of carefully textured and posed CGI.
Above: several test renders of the final posed scene with different lighting from HDRIs. The last one was my favorite; I liked the soft misty light and how clean the textures looked. The final step was to generate a fullsize (24″ x 36″ @ 300dpi) render and apply some-post processing in Photoshop. While pathtracing would produce slightly higher quality in terms of the lighting and reflections in the scene, a pathtraced render at this scale would take so long on my laptop (over a week) that I opted for directlighting. which took about an hour (Pathtracing and Directlighting are two of the render kernels available for Octane).
I learned a lot from this project: technical aspects of creating a scene like this with so many organic forms, artistic challenges of creating a visually pleasing scene that draws the eye where I originally envisioned it would be drawn, and pieces in-between — what it means to make a model, to make something recognizable, and what elements to cut when there is too much noise. Ultimately, I’m not convinced the final product fits my original vision, but it rarely does, and I’m happy with this. During my next project I will have a much better sense of what elements need more work and which ones are complete.