LogInElena Maradei shared the workflow behind the realistic 3D model of Ganondorf from The Legend of Zelda, showing how she captured his age and worked with textures.
Introduction
Hello! My name is Elena Maradei. I am an Italian 3D artist specializing in modeling, grooming, and texturing.
Influenced by drawing and art, I studied New Technologies Applied to Multimedia at the Brera Academy of Fine Arts, receiving an introduction to 3D modeling and animation.
After graduation, I decided to learn more about computer graphics by taking some courses.
I had an opportunity to work at many studios, including MPC and UPP, contributing to projects such as Transformers: Rise of the Beasts, Gran Turismo, and Captain America: Brave New World. Experience in visual effects production has allowed me to refine my skills and apply new techniques in my own work as well.
Ganondorf
This personal project came out of the need to refresh my portfolio and brush up on some software and workflow. When I work alone, I like to focus on almost all technical aspects of modeling and surfacing configurations.
I was very inspired by the illustration of John Patrick Gañas, who recreated his personal interpretation of Ganondorf, the main villain of The Legend of Zelda franchise.
At a very early stage, I wanted to imitate the shapes of the drawing, but then I changed my mind and turned it into a styling exercise, in which I switched from a cartoon look to a realistic and more detailed one.
The new goal was to try to make him believable and recognizable without excluding his facial features.
I worked with some images taken from the video game, a lot of anatomical references, and real people with similar somatic characteristics, like the prominent nose.
However, Gañas' drawing served as a mood board for the pose, lighting, and colors.
Body & Groom
My workflow always starts with Maya because I can easily handle a 3D object with multiple perspective views at once and emulate the point of view of an image through the settings of the cameras.
In fact, after placing a camera and verifying that it simulated the perspective of the drawing, I sketched the most relevant shapes on a basic human mesh.
When the base mesh was done, I switched to ZBrush, where I better sculpted the volumes and proportions.
For skin details, I used Texturing XYZ's textures, brushes, and tools within the program, such as the Surface Noise modifier.
To emphasize the pores, I also extracted the cavity maps, always thanks to ZBrush's functionality.
For the eyes, I took many sources from the internet, including head X-rays, and modeled the eyeball with the cornea, iris, sclera, and pupil, adding a tear duct and a mesh that works as tear flow for additional detail.
Most of the accessories were created in Maya utilizing its deformers or procedural tools, while I used Marvelous Designer for the tunic piece.
Although such a simple cloth could be made in Maya or ZBrush, I preferred to take advantage of the program's physical simulation to have ready-made fabric folds and a clean and organized UV set.
As for the hairstyle, I divided it into 11 descriptions and dedicated one of them to the fabric hairs.
To make them, I used XGen and Houdini; I find the latter very convenient for generating and managing a large number of curves simultaneously.
Retopology
During retopology, I mainly used Quad Draw in Maya, which was especially helpful for the dress because I needed to control the position of the edge loops that would coincide with the embroidery.
I always worked on UVs within Maya and set them into different UDIMs with a 4K resolution. I also adjusted the texel density, making sure that all the pieces had more or less uniform values.
Texturing
To create the textures, I used Substance 3D Painter and Designer. With the latter, I made the basic fabric material and the wave decorations, while I used the first software to modify the texture manually.
As I mentioned in the beginning, I started with a texture preset for the skin, in which I altered the tones with the HSL saturation filter and added other effects. Gradually, I incorporated more and more details, such as skin spots, small wounds, the tattoo on the shoulder, and light and dark shades to bring out particular areas of the body, such as wrinkles and holes in the forehead.
I created roughness maps and coat maps for the pores and lips and divided the face into colored regions to extract masks. These were used to individually adjust any values within the shader.
The jewelry colors were made with procedural tools and hand-colored parts.
The final look I wanted to achieve for Ganondorf was that of a man in his early years of old age. As he is a sovereign warrior in the video game, I aimed to give him a regal but unadorned appearance with worn clothing.
I managed the lookdev with Arnold and often used its Layer RGBA node to connect multiple maps and let them interact with each other, such as cavity maps with displacement maps, or to connect other different nodes.
I found the aiNoise shader very useful, as with its grayscale patterns, you can generate values between 0 and 1 and produce various effects. For example, I applied it as variations for the melanin channel of aiStandardHair.
During the shading process, I referred to Arnold's documentation, which not only clarifies the various functionalities of a shader but also provides recommended parameters for realistically replicating certain organic and inorganic materials.
Lighting
The lighting is composed of a few different sets.
For the lookdev part, I took a lighting rig from CAVE Academy, which is helpful for managing colors in a neutral way.
To match Gañas's drawing, I created a simple lighting set composed of a key light, a rim light, and an HDRI with warm lights acting as a fill light. To accentuate the right eye, I inserted a small light that illuminates only that 3D object.
I did post-production in Photoshop, where I accentuated the wrinkles using an ambient occlusion AOV and utilized the curves adjustment tool to balance the shadows and colors.
Conclusion
I developed the project in 4 months, working in my spare time. I spent a lot of time on the hair and eyebrows, and because of their particular shape, I had to do a lot of back and forth, especially with the hair type and the colors of the shaders. In the first tests, it looked like the character had dyed hair.
This experience helped me to better manage anatomical volumes, resume using software I had set aside for lack of time, try out Arnold shaders I had never used – like the aiStandardVolume – and reorganize my personal XGen workflow.
For those who want to pursue a career as a character artist, I advise studying anatomy and not underestimating the importance of references. I also suggest choosing the subject carefully, ensuring you are motivated to see it through to completion. Give importance to the character's background, tell its story through details, and observe it from multiple angles to ensure the forms work.
Additionally, I recommend thoroughly exploring the programs you use to understand their limitations and potential without worrying about the time spent studying.
I thank 80 Level very much for the opportunity to share my work through this interview.