First of all I'm looking into tips for modelling for animation, it is reccomended that I use either no triangle faces or as few as possible as it would cause difficulties with the topology and the deformation of the model.
(Topology is a term artists use to describe the construction of a 3D model, the arrangement of polygons and edges and how they flow together.)
Different Shapes In Modelling For Animation:
Triangle
When modelling triangles are, according to this guide, typically avoided as they can cause problems when deforming and animating a model as the faces don't really bend as you'd expect when for example bending an arm.
N-gon
N-gons, a five or more sided shape, should be avoided as it causes many problems when deforming and animating the faces could move in an unexpected way.
Quadrilaterals
When it comes to modelling, quads seem to be the way to go as they will ensure clean topology as well as ensure the proper deformation when animating.
Edge loops
Edge loops are, most of the time, a continuous loop which has no starting or ending point. E.g. if you were to select and highlight one vertex and follow the loop round eventually you'd arrive at the same vertex. Edge loops help when adding detail such as wrinkles or folds as well as they can help define how sharp an edge is. If an edge loop runs into a triangle, it ends the loop and breaks the flow of the line leaving it as no longer a loop.
Subdividing
When subdividing quads, your results are fairly predictable. You have rows and columns made up of four sided polygons and it is easy to see where those polygons will be split in half once it is subdivided.
When you subdivide triangles, things tend to get messy. There really isn’t a visual flow to the model.
Smoothing
When it comes to smoothing the geometry or using a quick smooth preview feature, triangles will cause anomalies across the surface of the model. Because of the uneven amount of vertices, the triangle can cause strange bumps or pinches in the geometry. This similar thing can happen to geometry created with n-gons but in an even more unexpected way.
When it comes to ensuring that the model will be able to deform and animate properly in places such as joints e.g. elbows knees and wrists the modeller will be benefitted by using quads. The reason quads will help is because the modeller can insert extra edge loops but if the modeller were to be using triangles it is harder to add or remove loops that would help the animator. With Triangles, it is also harder to see a clean flow of geometry and they tend to produce sharp angles that can harm the model’s appeal. When it comes to animation appeal is important both to the model and the artist who provided the model for animation.
My Thoughts
According to this guide using quads makes the model much easier to navigate, edit as well as it will make it much easier for someone else other than the creator to view and understand the model. Sticking to quads will greatly benefit me when it comes to rigging an animating and it will hopefully prevent the model from deforming incorrectly when animating. Using quads will be far better when it comes to improving the model's topology.
3D Modelling Standards
There are no set standards for 3D models but there are standards set by employers when it comes to the models you will be making for them. Standards from employers may differ and often conflict with other employers causing confusion.
Facial Shapes, Facial Rigging and Optimizing Rigs
Facial Rigging and Facial Shapes
When it comes to rigging faces as well as rigging as a whole, there is a lot that goes into it, but the basic idea of it is you put bones into your model which act as controllers for the body parts of the model. A character's facial rig is usually altogether separate from the main motion controls. It's inefficient and incredibly difficult to create a satisfactory facial rig using a traditional joint/bone structure, so morph targets (or blend shapes) are usually seen as a far better way of doing it.
How facial rigging works using joints/bones is very similar in theory except instead of putting long bones along each section of say, an arm, you put small bones in things like a character's lips and eyebrows etc.. Which then allow you to move the character's lips and eyebrows without just pulling the polygons around.
The way I've explained rigging makes it sound a lot easier than it really is but this is just my understanding of it, I know rigging is a difficult process so I will be gathering information on it to help me.
Facial shapes and expressions are expressed by a combination of the different parts of the face, when rigging a face to be capable of these facial expressions you need to have control over certain parts of the face.
Here is an example of where the facial "bones" go:
As you can see there are "bones" all around the eyes, nose and mouth so they can move in a variety of different ways rather than just a simple opening and closing or up and down motion. The eyes are controlled by the blue and red rings and as you move them around the eye will follow so you can have the character look at different things.
In this video, skip to around 20 seconds and you can see the faces of the character's face moving around in different and fluid ways which could not be done if the face could only open and close eyes and mouth and move thing up and down.
So as you can see rigging a face is no easy task so I will need constant information which I can use to help me while I rig my character's face if I choose to do it using bones.
As for morph targets/blend shapes, which I think I will be far more comfortable with, it's a very different process. In 3DS Max there is a morph modifier which you use to morph your character's face with a controller. The idea of it is you copy the polys and vertexes, in my case the head without the eyes and mouth interior, and create copies for each of the separate parts of the face you want to morph. For example, the bottom and top eyelids will need four separate copies, one for each eye lid. That goes for the nose, both lips and both corners of the mouth. I will also need one for the jaw if I wanted to move the jaw up and down, which I believe I do. That's around 10 copies in total as I may want to morph more of the model's face.
With each of these copies, I'll name them things such as "right top eye lid" and then I'll edit the copy so the top eye lid so the eye lid moved down to about where it would meet the bottom eye lid. Once I've done this, I'll go back over to the original head, to the morpher target and select the "right top eye lid" from the menu and once it's selected it'll move over into the morpher list and I'll be able to drag the value which starts at 0% to 100% which will move the eyelid all the way down to the point on the copy has the eye lid on. I can also drag it to around 50% and the eye lid will go half of the distance.
I'll then continue to do this for the rest of the model's facial features I want to move and thinking about it, this is a much better way of animating a face in my eyes and I think this is the way I'll choose to animate my face.
Optimizing Rigs
When rigging you basically set up a skeleton inside your character this is arguably the easiest part but in order for your rig to work properly, the bones and joints must follow a logical hierarchy. When setting up the skeleton, the first bone you place is called the root joint/bone. The root joint/bone is normally the pelvis as it works well as the center of the character. Every joint/bone after that will be connected to the root either directly or through another joint/bone e.g. the forearm bone will be connected to the upper arm bone, by a joint, which is then connected to the shoulder bone, by a joint, which is connected to the spine segments and eventually this will lead to the pelvis, the root.
Forward kinematics is one of the two ways to calculate the joint movement of a rigged character. Any given join can only affect parts of the skeleton that fall below it on the joint hierarchy. For example, rotating the shoulder of your character changes the position of the whole arm like if you were to stretch your arm out straight then bend the elbow which would in turn make the hand move. When animating with forward kinematics, the animator usually needs to set the rotation and position of each join individually to achieve a desired pose the animator would work through the hierarchy sequentially: root > spine > shoulder > elbow etc.. The final position of a terminating joint such as the end of your finger is calculated as a function of the joint angles of every joint above it in the hierarchy.
Inverse kinematics work in the opposite way, they work regardless of a hierarchy so if you were to move the hand which is fairly low on the hierarchy the rest of the body would follow the hand so the model doesn't break.
Setting constraints is important to make sure the rig maintains realistic proportions so by setting constraints you lock the position of each joint in relation to the rest of the rig.
Bibliography:
Modeling with quads or triangles + images: http://blog.digitaltutors.com/modeling-with-quads-or-triangles/
Facial Rig Picture: http://www.creativecrash.com/maya/script/speed-facial-rig-enhance-your-character
Youtube Video: https://www.youtube.com/watch?v=XULwjjk0DLo
No comments:
Post a Comment