Facial modeling and topology is something that has always caught my attention. I like to focus on the artistic point of view, including the interpretation of shapes & volume, but also work hard to describe a concrete volume with the kind of topology that will make it possible to maximize the deformation and flexibility that the animators may desire.
This side of our work is something that has lost relevance in the learning curves of new modelers, since the massive implementation of sculpting software has allowed them to focus on the artistic side and avoid all of the headaches that topology might bring to volumes and, ultimately, to model deformations. Of course, focusing on shapes and volumes allows us to quickly improve our model and to obtain quality results quickly, but we have to assume that, if it isn’t a production model, we will probably not retopologize it, so we’re always going to avoid topology.
Those are the reasons that pushed me to start this kind of topological library that I hope will be useful for those modelers who are looking for facial topology references and who are also interested in taking a general look at the entire process of completing a cartoon head ready for animation.
I usually work with Maya, but all of the steps and methods showed on these timelapses are the same ones you’ll find in other packages, such as split, insertEdgeLoop, extrude, merge, sculpt, etc., so you shouldn’t have any trouble following the videos. So let’s get started!.
1.- I like to start modeling from a cube instead of from any other kind of primitive to avoid unnecessary poles and extra divisions at the beginning. With a 2x2x2-resolution cube, it is very easy to separate the upper & lower; front & back parts of the head, and with just a few extrusions and divisions, to get all of the main areas of the face, such as the jaw, neck, nasolabial area, brow and eye socket areas, ears, etc.
I always model my characters with their lips and eyelids closed, as this will let me keep the same number of divisions coming from the upper lip/eyelid to the lower. Later it will make the labor of shape modeling or facial rigging easier because each part will begin its movement from the same initial point.
Blocking Topology: The goals at this point are to define the character’s overall volume and to make the main topological areas ready to add more topology in the next step or to send it to our sculpting software.
During this first modeling pass, I like to leave the eyelid area not 100% defined, as far as topology. The main goal in this area at this point is to collect all the loops coming from the surrounding areas, letting them flow to their homologues. This method guarantees the continuity of the topology through the area, and will let me decide in the future what kind of solution I will use to finish this area, a standard eyelid solution or a Pixar eyelid solution. In the future I will try to cover this topic in another article from the point of view of shape modeling.
Finally, I like to colorize the different topological main areas of the face to improve their visualization for future work. Also, you may notice that I have tried to put all of the pole vertices (vertices which connect more than 4 edges) on the same edge ring surrounding each main area. This topology organization allows me, if necessary, to reuse these areas coming from the different characters that I have modeled with the same topology organization.
2.- It’s time to move from the rough to the final topology model with all of the information required for future facial deformations, such as Blend Shapes or facial rig. As I mentioned in the beginning, this process could be done using any digital sculpting application and leaving retopo for later. However, with this example I prefer to add all of this topology by hand and, at the same time, to keep the focus on volume so as to make it more obvious where more detail is needed. In another article at some point in the future, I would like to talk more in-depth about topology, the main areas of facial topology, and the relationship of these with facial deformation, so I think it could be useful to add all of this detail showing the color scheme that differentiates the distinct main areas of topology.
Final Topology: Added all required topology to achieve nice, fleshy, sliding skin, as well as wrinkles & folds for the character’s future expressions and facial deformations.
Very often I work on the eyelid area as a separate low-resolution piece due to the amount of density required. I usually work with half of the final resolution required in this area, so before I attach this piece to the rest of the head, I apply a mesh smooth operation to get the full density required on the mesh to match with the loops coming from the surrounding areas.
One of the areas to which I apply the most changes during this episode is the neck area, where topology varies from a simple cylinder to a reorganized mesh in order to achieve rich and believable neck deformations where the sternocleidomastoid and trapezius muscles should appear. This work may not be necessary with stylized characters, but in this case, given the proportions and details of the design, all this new neck topology is going to be very useful to give the character added appeal.
3.- As you know, this is a kind of mesh created for animation which needs to have all necessary topology to get nice facial deformations, so the high density of mesh on the model could be a headache when attempting to maintain clean and smooth shapes. Due to all of this, the final modeling is probably the step that is going to take the most time, by far, when compared with the other two steps, as this last step involves the artistic side of the work while at the same time requiring that you continue to pay attention to the flow of the final mesh.
During this process I usually do a lot of work with deformers such as lattice, soft selections, and Sculpt Geometry Tool, as well as deformations on the mesh using direct vertex manipulation in normal mode. All of these tools let me focus on the volumes and shapes, and if necessary, they allow me to modify the mesh at the vertex level to fix or improve the flow of the topology.
Final Modeling: Wax on, right hand. Wax off, left hand. Wax on, wax off. Breathe in through the nose, and out through mouth. Wax on, wax off…
All of the elements of the face, such as eyelashes or brows, should have the same topology structure as the surface behind them in order to get nice deformations coming from the face rig without any extra work. So a good time to make them is when you have finished all of the rest of the work on the face. This means that you have an easy way to extract those elements from the mesh of the head, keeping the same amount of divisions. Finally, hair modeling is something that is hugely dependent on the production workflow, because it could be as easy as polygon hair, flat cards with hair painted with color and opacity, or just real hair. My choice in this case was a technique that will let me decide in the future if I prefer to paint textures or to use all of those curves as a guide for a real hair solution. With this technique, I’ll only need to create different NURBS surface levels over the whole head. All of these layers are made using three NURBS curves groups to create a loft surface between them. This method allows me direct control over the hair surfaces with just manipulation of the curve CVs.
That’s all for now, but as I mentioned in the beginning, all of those methods can be done using any 3D software and using other topological solutions. Those are just the kinds of solutions that I feel comfortable with and which give me the kind of deformation results I’m looking for.
I hope you’ve found it useful, and if you have any questions or comments, please feel free to leave a reply. Thanks!