WEEK3 : Full Body Walk Cycle | Constraints

Part 1 — Rotomation vs Matchmove Explanation

The process of matching CG characters / objects with live shooting materials is called “matching” (also known as “origin / match animation”), which is a hybrid name combining match moving, animation and some degree of rotocoping technology.

There are two main types of matchmove tracking, which are Camera Track and object track. Camera tracking needs to convert the camera information of the real shot lens into digital information and send it to the later software, so that the real shot and CG element positions include motion matching. For example, in the Avengers alliance, we need to replace the blue cloth in the camera shed with the street. In addition, there are also object tracking, which are relatively simple objects. Basically, there are only changes in distance and perspective, and there are not too many self deformations, such as license plates in driving vehicles and billboards in streets.

However, if you want to replace specific and complex deformation objects, such as characters, rotomation will be used. Rotomation is mainly used to replace the number of characters, including erasing (wire, original character and other elements) and role performance tracking.
In addition to replacing actors, rotomation is also used to make up a number of children. For example, in the film “the Irishman”, industrial light and Magic (ILM) makes Robert De Niro’s character 20 years younger.

Part 2 — Full Body Walk Cycle

1 . Animation Principles in Walk Cycle

• solid posing
• timing
• arcs
• slow-in & slow-down
• pose-to-pose
• anticipation
• follow-through and overlapping action
• square and stretch

2 . Animation Process

This time our task is to make a full body walking animation. Compared with the last half body walking, the whole body walking increases the process of arm swing and head shaking as well as the whole body center of gravity transfer. Details also include finger, props and subtle changes in expression.

This is my reference.

This time, I chose a model with exaggerated figure and more props, which is the cowboy. First of all, I want him to have a character of his own. He is a proud man, but he has an angry mood. I chose a bottle of wine as the prop. Secondly, I think that walking with personality is not ordinary walking. His arrogance will make him walk heavily, that is to say, he lifts his feet slowly and steps down quickly. So I did a normal walk first.

Later, I tried to make his toes have a certain rotation value, that is, walking with the toes outwards is more in line with the natural and unrestrained posture.

I try to make walking different. I use slow in and slow out. Lifting foot is a decelerating movement and landing is an accelerating movement. And it slowed down, but there was still slippage. But it looks better.

This is really a very long process, I even modified 15 versions, and my file name even began to become strange!!!

I first let the body twist up, that is, shaking left and right, so that the body’s center of gravity is always in a stable state. Then the body takes on a curved shape, with the head up. I adjusted the posture of my hand and noticed that the arm movement is a follow-up motion, that is, the upper arm moves first, then the arm follows slowly, so it presents a curved shape.

Finally, I added a little bit of detail. His scarf and hat need to be shaken up and down because they weigh a lot when they step on them. The rope on the body follows the movement. The face also needs to be squeezed and stretched. It is flattened when it is stepped down and elongated when it is lifted up. In the eye part, I added a line of sight shift, which was accompanied by a blink.

After adding props and wine bottles, I realized that the range of motion of my hands should be larger and have a sense of strength.

About the hat I only did rotation and displacement to make it fluctuate up and down, but it was too rigid. Later, I rotated the brim of the hat at wrong frames to create a shaking feeling.

Final version

After the teacher’s feedback, I realized that my swing was not enough and the body didn’t swing back and forth, making my body a little stiff. I adjusted it again according to the teacher’s amendment.

• 

• 

• 

• 

Part 3 — Constraints (Parent vs Parent Constraint)

Definition

With parent constraints, you can relate the position, translation, and rotation of one object to another so that they behave like a parent-child relationship with multiple target parents. The movement of objects is also affected by the average position of multiple objects.

When a parent constraint is applied to an object, the constrained object does not become part of the hierarchy or group of the constrained object, but remains independent and behaves like a child of the target. Constraint objects are also known as target objects.

Types

Point

Point is a point constraint. The center point of the constrained object will coincide with the constrained object, which is controlled by the mobile hand.

Aim

Aim is a target constraint. The constrained object will change its rotation according to the movement of the constrained object. The axial direction of the constrained object is always in the direction of the constrained object. A typical use of constraints includes aiming a light or camera at an object or group of objects. In character settings, a typical use of target constraints is to set a positioner for controlling eye rotation.

Orientation

Orientation is a rotation constraint, which only constrains the rotation of the object and does not change the displacement.

An orientation constraint matches the orientation of one object to one or more other objects. This constraint can be used to orient several objects simultaneously. For example, animating the head of one character and then constraining the heads of all other characters based on the head you just animated can make a group of characters face in the same direction at the same time

Scale

Scale is a scale constraint. Only scale is constrained, and the translation and rotation will not change.

A scale constraint can match one scaled object to one or more objects. This constraint is useful when scaling multiple objects at the same time. For example, you can set all of a set of characters to the same direction at the same time by animating the head of one character and then constraining the heads of all other characters to your animated head.

Parent

Parent is a parent-child constraint, which is often used in animation. If the constrained object has been constrained, it can no longer change its own movement and rotation. It completely follows the movement and rotation of the parent object.

Parenting is exactly what it sounds like. If I select this cube as like this ball and hit P. The cube and move it around on its own, nothing happens but if I move the ball. It follows the ball. Now, if I rotate the ball. It also moves around the ball. However, I can still move this cube around over there, and it will still move around.

Pole Vector

Applies to rigs only. The pole vector constraint causes the end of the pole vector to move to and follow the position of an object, or the average position of several objects.

In character setup, the pole vector of the IK rotation plane handle of the arm joint chain is often constrained to the locator placed behind the character. When using constrained pole vectors to manipulate IK rotation plane handles, the joint chain does not flip unexpectedly. Because flipping occurs when the handle vector approaches or intersects the pole vector, the pole vector should be constrained so that the handle vector does not cross it.

What do we use constraint for

The “parent” is a kind of subordination relationship. Selecting a parent object means that the related child objects of the object are selected at the same time. An object can have more than one child, but a child can only have one parent.In general, there are more “parent” in animation, while “parent constraint” is used more in binding and rigging.

How can they help in animation

Object B and Object A have the “parent”, then Object B’s movement and rotation control Object A’s movement and rotation (Object A’s attributes respond to changes). Object A was Parented on object B, the changes of three attributes of Object B will drive Object A correspondingly. The constrained object can adjust the size of other constraints, and can also release the constraint at any time by keying frame. The object constrained by the “parent” can not key the frame and can only adjust attributes through constraint body.

When we have finished the animation and need to add a prop, the “parent” will be useful. The “aim” constraint are also used to adjust eye orientation

Difference between Parenting and Parent Constraining

1. The “parent” can still move freely. The sub object in “parent constraint” cannot be moved. Even if it is moved, if the parent object is moved again, it will return to the constrained position, unless the offset value of the child object is adjusted separately
2. The “parent” works on the “displacement”, “rotation” and “Scaling” of the child object. The “parent constraint” only works on the “displacement” and “rotation” of the child object
3. The “parent” can be viewed by expanding its hierarchy in the outline view. The “parent constraint” cannot.

Solution (the right way)

However, you can have multiple constraints on one object so generally the rule of thumb is, it’s best to use constraints when it comes to rigs, because it means you can do other things. And one of the other things you can also do to that does tend to help is, if you’re parenting something to a rig, you can group it first. And do to do that you just go Ctrl G, and then go to modify center pivot, pivot back. The Cube can be moved around still, and that groups point is still there. So kind of is beneficial to constrain groups because the cube is still able to do the same thing.

Part 4 — Workflow in 3D Animation

3D animation pipeline in industry

Mid production

1. 3D model — in 3D software, the modeler makes the scene model of the story.
2. Texture material: according to the concept design and the comprehensive opinions of customers, producers and directors, setting the color, texture and texture of 3D model is an essential and important link in the animation production process.
3. Split shot animation — referring to the script and storyboard, the animator will make each shot animation for the lens or other objects that need to be moved according to the shot and time.
4. Lighting: according to the style orientation of the early conceptual design, the lighting engineer will illuminate the animation scene, describe it carefully, and adjust the material carefully to grasp the rendering atmosphere of each lens.

Post production

1. 3D special effects – according to the specific story, produced by the special effects artist. Several kinds of water, smoke, fog, fire, light effects in the actual production of three-dimensional software performance methods.
2. Layered rendering / compositing: after the animation and lighting production is completed, the rendering personnel will render each shot file in layers according to the opinions of the post compositor, and provide layers and channels for compositing.
3. Dubbing Music: according to the needs of script design, the professional dubbing engineer dubs according to the camera, and matches the appropriate background music and various sound effects according to the plot.
4. Editing output: the rendered images of each layer are synthesized into a complete film by later personnel, and edited into different versions according to the opinions of customers, producers and directors for different needs. In the process of 3D animation production, each link complements each other. If the connection is not good, it will lead to the failure of the whole film. Therefore, the staff of each link should not only know their own links, but also have a comprehensive understanding of the production of other links.

character animation workflow

We usually work with reference footage then get a storyboard made with a frame count, and then block out poses in Maya, and playblast, come back to, and spline pass it once we’re happy with the pot with the blocking. And then we playblast and come back to it and polish it. The difference between blocking and splining is main differences were blocking you’ll just get your base key poses in and then splining is where you start to come in and and smooth out the animation in between, and then polish all that final extra detail. And once you’ve finished it all and you’re happy with it, then render.