3D Creation Pipeline Workflow: Mexican Wrestler

For my 3D Creation Pipeline Module, I had to create two hero assets: a hard-surface SubD workflow and an organic character asset. The objective of this project is to show my acquired knowledge on the computer-generated creation pipeline, and all of its stages: modelling, sculpting, retopology, UV mapping, baking, texturing, lighting, and rendering.

On this blog post, I’ll focus on the character asset: a Mexican Wrestler based on a concept art by Tim Burton.

Project Concept and Inspiration

Since I was a child, I’ve had an affinity to movies, especially if they were animated. One of my favorite movies ever is The Nightmare Before Christmas by Tim Burton. He is a director which I find fascinating. His way of creating magical and profound worlds inspire me to keep being curious and keep that child-like imagination. In October, I found out that The World of Tim Burton exposition would open in the Design Museum in Kensington. As a fan, I immediately bought a ticket.

While in the exposition, not only I found props and costumes from Burton’s movies, but also many concept art sketches made by himself. One of those sketches was of a Mexican Wrestler.

Being Mexican myself, I instantly knew this would be my character. I had a few ideas besides this, but my gut told me to choose this.

After asking Richard and Mollie for their approval, I started to collect reference pictures. I could not find more sketches of this character, but a few years ago someone made a statue for another Tim Burton exposition. Those two pictures would be the most resourceful references I had. I compiled these references in a PureRef file along some others.

Planning, Research and Management

Before starting with the modeling stage, I created a Gantt chart to visualize my project’s timeline, track the various stages deadlines. I found this the most helpful way of organizing myself and creating goals that were achievable so I could have my project completed for the delivery date.

During the course, to prepare to achieve my projects, I used various of the internal tutorials in Canvas to familiarize myself with the tools and software, specially Zbrush, which I had no experience working with.

After doing various exercises and getting to know the UI and the tools, I felt confident to start modeling my character.

This asset would be used for a game engine, so the workflow would vary from my hard-surface asset. As it was made for a game running in Unreal Engine 5, I would had a limit of 50,000 vertices for my model, as I researched that that amount is one of the industry standard practices.

Production: Blocking and Sculpting

I used Zbursh as my software of choice. I began by blocking out the main shapes of the body. I used as reference George Zaky’s character blocking examples. I found one that had more of a muscular stylized silhouette, which I found most similar to my concept art. I began by blocking the face shape, as well as the arms, legs and torso shapes, getting the proportions right.

After having the main shapes mostly defined, I began to model the secondary details of my character. By using the Dynamesh tool, I could create enough topology on my blocks to starting the details of the muscles. I used as reference the book Anatomy for 3D Artists by George Zaky as it is full of information of the human muscles, tendons and anatomical structures of the body. Even if my character would be designed in a stylized manner, I could use the references poses and pictures in this book as a guideline for my sculpting sessions.

As the only image I had to base of my design was a frontal view, I had to make some design related decisions: First, I took the executive decision, after speaking with Mollie about it to have the face modeled in a realistic manner, having lips and a nose in contrast to the concept art, but still keeping the eye sockets empty, inspired by characters like Batman where their eyes sockets illuminate.

Also, the fingers of the concept art appear as little nubs extruding from the palm. I decided to model fingers which are a little longer, so that I could have control over the different parts of the hand model and with the intent of having the possibility of animating them in the long run.

After being happy with my shapes, I started to merge, DynaMeshing them together and using the smooth brush to blend in the shapes until I had one subtool only. Then, I moved around the silhouette of my character so it resembled more to the concept art I had.

Production: Retopology and UV Mapping

When my character base mesh was ready, I decimated it so I could import it into Maya for the Retopology stage of the pipeline. Before the project, I had worked on some exercises and tutorials on retopologizing that I found on the Canvas library to get comfortable enough with the Quad Draw tool, as well as the retopologizing process. Using reference I found of the main face and body edge loops for a correct distortion of the body, as well as the reference photos in the Anatomy for 3D Artists book, I started to retopologize the mesh.

I started with the face, first drawing in the eyes, mouth, nose and face loops which are the most important. Then I began to fill out the spaces between them.

Then, I began by marking the body loops of the arms, shoulders, knees, and elbows. Because my character was very muscular, it was a little difficult to get into the crevices in the armpits and between the legs to retopologize it, but in the end I achieved it to the best of my ability.

Later, I began working on the fingers. I began by retopologizing the index finger, and the separating that part of the mesh to duplicate it and moving it in the position of the other fingers. By using the relax feature in the Quad Draw tool, the retopologized mesh conformed to the base mesh and then I manually move the vertices where I needed them to be.

Lastly, I worked on the shoes, which were the easier parts of the process. When my retopologizing stage was done, the final part was to implement a technique I learned from watching Abe Leal 3D’s video on retopologizing organic assets. I smoothed my mesh one level and then, with the Decimated mesh live, I used the Conform feature on Maya. This created a mesh that had almost the same silhouette as my zBrush original mesh.

Having this retopologized mesh, I used it to extract the face mask and the pants from it, retaining the correct topology on it and deleting from the body the legs, as they would not be visible in the final asset.

Next, I began to UV unwrap my mesh, being mindful of where to place my seams around the body and the newly created mask and clothes meshes. This was not a lengthy process as it had not that many of seams. I packed it all up in one UV tile, as this asset could not use a UDIMs workflow, as I would be using it for a game engine.

Finally, I exported my meshes and got them into Zbrush for my final sculpting phase.

Production: Projecting Details, Baking and Texturing

Having my retopologized meshed in zBrush, along my original undecimated meshes, I used the Project feature in zBrush to have the secondary details in my new mesh. Then I imported the pants and mask meshes into zBrush so I could sculpt the last details onto them, mainly the folds in the fabric.

With the final Sculpting stage of the pipeline done, I started working on the Texturing stage. I used Substance Painter to create the texture maps.

I started by importing my Low Poly mesh into Substance and baking the maps using the the High Poly mesh. This gave me all the maps I would use during the texturing stage, as well as the Normal Map containing all the details of the High Poly mesh in it.

Then, I created a folder for each type of material. Like my hard surface asset, the reason was to have organization in my workflow. This was an easier asset to texture, compared to the Leica Camera. I started with the skin, adding a base layer with my color, and adding layers of color variation to create a stylized skin appearance. Finally, I hand painted the lines in the chest, face and hands, as well as the makeup in the face, so that later it would merge with the mask textures.

Moving to the clothes and mask, I started with the color layers, then applying a fabric texture to it. Then I made the lines pattern in the upper area using a lines map I found on the Substance Library. Finally, I hand painted all the other details, like the lines in the pants and the shapes on the mask.

Production: Lighting and Rendering

After I finished the texturing stage, I went onto Maya to create my lighting setup. I used the three point light technique using Arnold’s area lights to illuminate my scene, as well as using the Studio Tomoco HDRI from Substance Painter for its neutral lighting in a Skylight Dome, just like my hard surface asset setup. I also created a backdrop made of a subdivided cube made into a half sphere with an emissive material.

I also created a few fill lights for the parts that were not receiving a lot of lighting from the main lights, mainly the shoes front and back. When the lighting was ready, I moved on to rendering my scene. Here are a few detail beauty shots that I took:

Analysis and Conclusion

The greatest strength of this project was that it involved the successful designing and modeling of the two hero assets. The character asset indicated my first trip into organic modeling, witch was a quite challenging venture, as I had experience in hard surface modeling prior. By learning various abilities addressed within the modeling process, I am now proud to show how far I've come. Nonetheless, there were some considerable drawbacks. Time management became a problem several times during the project, which meant less time to execute the overall process and less opportunity to fix it if things went wrong. I still feel my skills in organic modeling do need time to improve.

The decisions I made during this project were heavily guided by some clear objectives from the assignment brief and the input from my teachers during the support sessions. Each decision made through out the pipeline was deliberate and focused around the goals and limitations of the project. Some decisions were disorganized along the way, but it was a steady progress with the right approach that led to success. The areas of opportunity that I found after finishing my work are the following:

• Skill development: This whole journey has made me realize how important skill training is. I am keen to spend more time doing sculpting practice for confidence and proficiency, as well as refining my hard surface techniques.

• Adaptability: Having faced and overcome unforeseen challenges has highlighted the importance in me of being adaptive and knowing how to solve problems in creative workflows.

These lessons learned here will help me to efficiently manage and run things in my future projects.

The journey of this project has been quite challenging yet very rewarding. From a firm basis beginning with hard surface modeling, I stepped into uncharted territory by entering organic modeling, having been an enriching learning experience.

This project has been a transformative experience, both personal and professionally. I have gained confidence in my ability to learn and apply new skills, even under challenging circumstances. Looking ahead, I aspire to further refine my modeling expertise and explore advanced techniques in character creation and hard surface design. My ultimate goal is to contribute to innovative and immersive projects so I can enter the animation industry. Moving forward, I am excited to build upon this foundation and continue evolving my skills and creativity.