Arm motion and deformation of a real-time character

Arm motion and deformation of a real-time

character

Johan Bergstrand

Computer Graphic Arts, bachelor's level 2018

Luleå University of Technology

Department of Arts, Communication and Education

Abstract

The anatomy of the human body is important in both the video game and the VFX industry.

Whenever a limb is bending or moving, there are several different muscles working for this to happen.

This work will look at the re-topology of the arm and the moving anatomy around the bending and twisting areas of the arm.

A method will be done were three different re-topologies of a muscular arm will be created.

These arms will be deformed with common arm movements/animations to find better or worse retopology methods.

Sammanfattning

Anatomin i den mänskliga kroppen är viktig i både tv spels och VFX branschen. När en lem böjer eller rör sig, arbetar flera olika muskler på sig samtidigt för att detta skall hända.

Detta arbete kommer undersöka den rörande anatomin omkring dom större böjnings och vridnings områdena i armen.

En metod kommer att genomföras där tre olika om-topologier av en muskulös arm kommer att skapas. Dessa armar kommer att deformeras med vanliga arm rörelser/animeringar för att hitta bättre eller sämre om-topologi metoder.

Table of Content

1 Introduction ... 3

1.1 Background ... 3

1.2 Purpose ... 3

1.3 Limitations ... 3

2 Glossary ... 4

2.1 Anatomical Planes. ... 4

2.2 Body Movements ... 4

2.3 Anatomical positions ... 5

3 Theory ... 6

3.1 The Synovial Joint Movements... 6

3.2 Bones of the Arm ... 6

3.3 The Muscle of the Arm ... 7

3.4 Topology & Modelling ... 10

4 Methodology ... 12

4.1 Softwares ... 12

4.2 Arm Creation ... 12

4.3 Method Criticism ... 15

5 Results ... 16

5.1 Arm Flex ... 16

5.2 Arm Supination and Pronation ... 17

6 Empiri ... 20

6.1 Critics Reviews ... 20

6.2 Critics Discussion ... 21

7 Discussion ... 22

7.1 Personal Discussion... 22

8 Conclusion ... 23

8.1 Future improvements ... 23

9 Plates... 24

10 References ... 28

11 Appendices ... 30

1 Introduction

1.1 Background

Anatomy, it´s been studied for thousands of years, since the earliest civilizations, dating as far back as Egypt 1600 BCE (1).

Since then, humanity has moved forward with the studies of the human body, be it in medicine or in art. Good examples are Leonardo da Vinci paintings based on the human body, most notably his "Vitruvian Man" painting (PLATE 1

Today it is very important for computer graphics(CG) characters to be as real and alive as possible. A failed or stiff animation based on bad anatomical studies of the body will give the player or the audience no satisfaction to keep on playing/watching (2).

For game artists, this is more of a challenge because of their need to lower the topology, in order for the game to function smoothly.

1.2 Purpose

The purpose of this research is to study the effects of different topological workflows based on deformations/movement of real muscles.

- How will certain muscles deform on different topology flows?

- Will there be better results, or worse?

The arm is a body part that can have full a range of motion. It is important to know the complexity of the muscles deformation around these areas, but for a CG artist it is also important to know a good topology workflow to make these muscles deformations look real on a low poly arm. A low poly arm means less information to act and behave on, as a normal arm would.

1.3 Limitations

• This is a research, based on the arms. The rest of the body’s muscles and joint movement system will be excluded.

• All the example will be taken from a male subject.

• There are several ways to work in muscle deformations. However, this research will only focus on the retopology part to find a good topology. That means no deeper research on 3D modelling, rig setup, skin weighting or animation will be given.

• The three different re-topology methods are based on my three years of studying.

There are more methods for re-topology, but I will only research these three.

• The arms of the methodology section will be performing flex, supination and pronation movements. Other types of movements will be discussed but not shown.

• The number of polys on the arm should be no more than that of a third person character.

2 Glossary

2.1 Anatomical Planes.

(Plate 3)

Used as a reference when describing the different moving directions of the body parts.

• Sagittal Plane divides the body in half from an Anterior/Posterior view.

• Coronal Plane divides the body in half from the Lateral Sides

• Transverse Plane divides the body from the middle in a horizontal state.

2.2 Body Movements

To understand the muscles behaviour, one needs to understand the basic movements of the body (3).

Flexion Decreases the angle between two segments Extension Increases the angle between two segments Abduction Pulls away body part from the sagittal plane Adduction Brings body part towards the sagittal plane

Medial Rotation Brings body part towards the centre of the body, following the transverse plane

Lateral Rotation Moves body part away from the centre of the body, following the transverse plane

Elevation Superior translation (up) of a body part, away from the body’s transverse plane

Depression Inferior translation (down) of a body part, away from the body´s transverse plane

Dorsiflexion When the front of the hand decreases the angle with the outside of the forearm.

Palmar flexion The antagonist of “Dorsi Flexion” of the arm. Decreasing the angle between the palm and the anterior forearm

Pronation Movement near the wrist. Pronation for the arm is when the palm of the hand faces inwards towards the body or downwards in a forward straightened position.

Supination The antagonist of the Pronation, having the palm facing away from the body or upwards in a forward straightened position. For the foot, it means having the sole turned inwards.

2.3 Anatomical positions

The view in which the anatomy is explained (V.L. Winslow, 2015, 20).

Anterior (in front of) A frontal view of the body or a body part.

Posterior (in back of) A view from the back of the body or a body part.

Lateral (side) A view of the side of the body or a body part.

Medial (middle) A view from the middle or referring to the middle structure of the body or a body part.

Superior (above) A view from above or referring to the upper structure of the body or a body part.

Inferior (below) A view from below or referring to the lower structure of the body or a body part.

Exterior (outside) Body parts facing outwards, away from the body Interior (inside) Body parts facing inwards, into the body

Dorsal Back of the hand.

3 Theory

The theory will start by naming the Synovial Joints, and later the bones, muscles and how they deform/react based on the synovial movements. Lastly, I will talk about topology and re- topology.

3.1 The Synovial Joint Movements

(Plate 4

)

Also called Diarthroses, the synovial joint movement stand for most of the body’s movement.

These joints exist where bones or muscles tendons end and begin. The basic arthrology of the joints is that of a joint capsule. This joint capsule consists of a fibrous tissue on the outside and a synovial membrane on its inside. Between the joint capsule and the bones are a soft coating called the articular cartilage that reduces friction when moving (Plate 5) (P.

Richer, 1986, 19-20)

All the synovial joints may consist of the same tissue/coating but are not all the same in shape. Some shapes give greater freedom of movements, while others are more limited (V.L Winslow, 2015, 42-44).

Ball-And-Socket Joint, just like the name says, this joint is shaped like a ball and connects with a capsule-shaped end. Of all the joints, this one provides with the greatest mobility.

Hinge Joint, this joint can only move back and forth in one direction. The joint is shaped like a cylinder in one end and has a concave surface on the other.

Pivot Joint gives certain bones the ability to rotate on its own axis. It´s shaped like a cylinder in one end and attaches to a concave surface.

Saddle Joint has two ends shaped like a saddle and provides more mobility than the hinge joint, but not as much as the ball-and-socket joint.

Plane Joints, two flat ends gliding against one another. The plane joints are the least moveable joints of the synovial joints.

Condyloid Joint, is an almost look-alike of the ball-and-socket joint, only this one is more egg-shaped than spherical. The sockets are differently placed as well. Movement is similar to the ball-and-socket joint but more limited, given its form.

3.2 Bones of the Arm

The humerus bone is the single bone of the upper arm. It connects with the Scapula (shoulder) from its superior side and the radius, ulna bones from its inferior side.

At the superior side is the head of the humerus. A sphere-shaped part that connects with the scapula. This means that the upper part of the humerus has a Ball-and-Socket joint, which lets the bone move around at almost every angle. On the other end, however, there is a hinge joint, which limits the inferior part (the elbow) to only flex and extends.

The Radius Bone (Plate 7)

The radius bone is one of the two bones of the inferior part of the arm and is also the smallest of the two. At its superior side, the radius bone is quite small and doesn’t stand for much of the hinge joint mobility. However, at its inferior side, it becomes larger and connects with the hand in a condyloid joint movement.

The Ulna Bone (Plate 7

The ulna bone is the neighbour of the radius bone and works as a sort of mirror from the radius. At the superior end, it connects with the humerus bone, creating the hinge joint. It is also here where the ulna bone gives shape to the olecranon (the elbow). But at the inferior side, the bone becomes smaller, making almost no connection to the hand.

(P. Richer, 1986, 36-40)

3.3 The Muscle of the Arm

(Plate 12

3.3.1 Upper Arm Muscles The Deltoid

The deltoid is the connecting muscle for the torso muscle family and the upper arm. Its primary function is the abduction of the arm. The deltoid is split into an anterior, posterior and a lateral portion that begins on various areas of the torso. They all merge and connect halfway down the humerus bone of the upper arm.

Important to know is that no matter how you move the arm, this attachment will always end on the exterior side of the upper arm.

When elevating the arm, the deltoid´s position (from a lateral view) will move back and end in the posterior view of the body. (P. Richer, 1986, 59-60)

The Biceps Brachii

The biceps brachii is the long muscle of the anterior part of the upper arm. The biceps job is to flex as well as supinate the forearm. At its superior side, it is split into two portions, the long head and the short head, both begin at the scapula, beneath the deltoid muscle. On the inferior part, it becomes one and disappears behind the Brachioradialis and the Pronator Teres muscles. (P. Richer, 1986, 64)

The most visible deformation of the biceps comes when you flex the arm. The biceps will then curl up to a ball shape in the upper arm. (U. Zarins, 2014, 156)

The Triceps Brachii

The dominant muscle of the posterior part of the upper arm. The triceps is the antagonist of the biceps. The action of the triceps is to straighten the lower arm. Doing so will flex the triceps and it will be visible on the posterior of the arm.

The triceps consist of three parts, the long head, the lateral head and the medial head. The long head´s origin is positioned at the scapula, the lateral head on the humerus, and the medial is also attached to the humerus, but further down. (P. Richer, 1986, 65)

The Brachialis

The brachialis muscle is the muscle beneath the biceps brachii. On a well-built arm, this muscle can be seen on the interior side of the upper arm. It starts at the on the humerus and ends on the ulna bone. Its main function is to assist the biceps brachii with flexing the

forearm. (V.L. Winslow, 2015, 117)

The Brachioradialis & The Extensor Carpi Radialis Longus

The “radial muscle group” duo, the brachioradialis and the ECRL are two important muscles for the artist because of their visible shape on the exterior part of the arm. They both have origins on the humerus bone, but the brachioradialis ends at the inferior part of the radius bone right before the thumb and the ECRL keeps going to the base of the index finger.

These two muscles are very used to the rotation when supinating or pronating. It is good to remember that the origins of the muscles never leave the exterior part of the arm but the attachments are always around the thumb. (V.L. Winslow, 2015, 125)

Coracobrachialis

A well-hidden muscle between the biceps brachii and the deltoid muscle in anterior view. It originates from the scapula and attaches to the posterior side of the humerus bone. It is most visible when elevating the arm close to the armpit. (P. Richer, 1986, 64)

3.3.2 The Forearm Pronator Teres

Part of the Flexor muscle group, the Pronator Teres is the only one of the four that does not attach somewhere near the hand. Its origins are at the end of the humerus and the start of the ulna. The attachments go around the arms lateral side of the anterior arm and lands halfway down the radius bone. Its job is to pronate and flex the forearm. (P. Richer, 1986, 66)

The Flexor Carpi Radialis(FCR), Palmaris Longus(PL) and Flexor Carpi Ulnaris(FCU) The rest of the forearms Flexor group, the FCR, PL and FCU start at the same position as the Pronator Teres at the end of the humerus. (V.L. Winslow, 2015, 120)

• FCR attaches around the metacarpal area (the thumb) and its job is to flex and abduct the hand. Of the three muscles, it is the one most anterior.

• PL is small in the beginning but fans out when passing the wrist, attaching on the palm of the hand (Palmar aponeurosis). Its job is to flex the hand and wrist towards the forearm. Of the three it is the one stationed in the middle

• FCU is the muscle on the exterior of the forearm and attaches to the side of the hand near the pinkie. Its job is to flex and adduct the hand.

Anconeus

This small muscle exists just under the olecranon and next to the radial muscle family. It starts at the end of the humerus bones lateral side and inserts on the ulna, on the posterior side. Its job is to help straighten the forearm. (V.L. Winslow, 2015, 118)

The Extensor Carpi Radialis Brevis (ECRB), Extensor Digitorum (ED), Extensor digiti minimi (EDM), and Extensor Carpi Ulnaris (ECU)

This large family (Also including the Extensor Carpi Radialis Longus) is on the exterior part of the forearm and they all have an origin (now excluding the Extensor Carpi Radialis Longus) on the end of the lateral side of the humerus, next to the olecranon (elbow).

(V.L Winslow, 2015, 122- 123)

• The ECRB is positioned almost under the ECRL and the ED but follows the ECRL and ends up on one of the metacarpal bones of the hand. Its job is to extend and abduct (radial) the hand around the wrist.

• ED is the main muscle in the middle of the forearm. Its job is to straight the hand but also the fingers. That’s why the ED fans out after it passes the wrist, as it inserts in all four fingers.

• EDM is a very small and lean muscle that follows the ED and helps it straighten the pinkie.

• ECU has two origins. One on the humerus and one at the beginning of the ulna. Its attachment is on the outer metacarpal as its job is to extend the hand but also adduct (ulna) the hand.

The Anatomical Snuffbox Muscles

These three muscles are deep muscles at origins, working beneath the extensor digitorum, but will stick out at the end around the wrist near the thumb. (V.L. Winslow, 2015, 123- 124)

• Abductor Pollicis Longus abducts the thumb

• Extensor Pollicis Brevis extends the thumb

• Extensor Pollicis Longus help in extending the thumb, barely noticeable.

3.4 Topology & Modelling

(Plate 13In the 3D world, there are different types of 3d models. The Polygon modelling is the most common one, representing both the film and the game industry. A polygon information consists of connecting edges and vertices. (W.Vaughan, 2018, p.08)(4) - Polygon/Face: A polygon/face is a platform consisting of several connecting edges.

- Edge: An edge is two vertices connecting, or a point where two faces meet.

- Vertex: A point in the 3D-space.

A computer works/thinks in vertices, but visually, the mesh will have tris. That means that a face consists of three vertices. For a human, it is possible, and sometimes preferable if the faces were quads. Quads are faces with four vertices, but in truth, the computer just hides the edge crossing the face. If a face consists of more than four vertices it is called an n-gon.

N-gons are to be avoided, always. (W. Vaughan, 2018, p.09) (T. Terävä, 2017, p.13)(5)

3.4.2 What is Topology?

The topology of a mesh is the flow and directions created by a number of vertices, edges and faces.

Having a good topology is crucial for believable results as it is the topology of the mesh that deforms when animating.

The areas surrounding the deformation should have more detailed topology than on areas where deformations aren’t happening.

Creating a re-topology means building a “clean” replica from a previous mesh. This re- topology is based on the artist's skill and knowledge of common topology flows. (W.

Vaughan, 2018, p.04)( T. Terävä 2017, p.17-18) (6) (7) (8) (9)

3.4.3 What is bad Topology?

Bad topology can be several things, not just for animations/ deformation purposes. A bad topology flow can be seen in a rendered result or complicate a UV Texture. Reasons for bad topology differs.

- I single badly placed tri-face in the middle of a quad topology - The lack of topology in areas where the mesh deforms the most

- A so-called polestar topology where more or less than four edges meet on a single vertex (this topology can be both good or bad for your flow depending on where it is placed).

Image 3.1 A polestar topology as shown in Maya

These are just to name a few. (W. Vaughan, 2018, p.05, p.12, p.16)

When limbs move closer to one another, they need to retain their shape as much as possible.

Image 3.2 bad deformation as shown in Blender. Image 3.3 Deformation of the biceps and the triceps

This is an example of a bad bending (image 3.2), because of the lack of polygon information, the limbs on both sides collapse with each other like a straw, creating a bad, unrealistic deformation in contrast to what they should deform like. (image 3.3) Is how an actual arm deformation should go, with no collapse yet a change in the muscle and skin. (8) (10)

4 Methodology

4.1 Softwares

The method was to create an arm with three different retopology workflows and test them inside a game engine. For this pipeline (9) (12), the following software’s where used as of April 2018:

- Zbrush 4r7 for the creation of the high poly details of the arm

- Maya 2017 for the retopology, UV mapping process, rigging, skin weight and animation process.

- Substance Painter 2018 for the texturing and painting of the muscle areas.

- Unreal Engine 4 for the final view inside a game engine.

4.2 Arm Creation

Starting off in ZBrush, sculpting the arm based on the theory’s muscle build. The arm is directed in an inward position (interior, facing the body).

Image 4.1 The high-poly arm as seen from Zbrush

Inside Maya, the arm was being re-topoed, based on three different approaches. The first two models were focused on where the limbs would bend and deform.

Image 4.2 arm 1 re-topoed 4.3 Arm 2 re-topoed

The third model where focused based on the directions of the superficial muscles, the muscles closest to the skin.

Image 4.4 The third arm in different views as seen from Maya

The arms vertex count where:

- 761 verts for Arm 1 - 1109 verts for Arm 2 - 1322 verts for Arm 3

These results are all acceptable for real-time rendering inside Unreal Engine. (12)

After the retopology, the arms were unwrapped inside the UV editor and sent to Substance Painter where I painted the muscles with a strong colour difference.

Image 4.5 From the top down, arm 1 to 3. To the right, the UV Map, everything seen from Substance Painter.

With the texture done in Substance Painter, a master rig was set up and animated, later to be used on all three arms.

Figure 4.6. Rig set up for arm 3 as seen from Maya

The rig was later banded together with each arm mesh. The weight paint was set in default mode and not changed.

Image 4.7 and 4.8 showing the weight paint result.

Two animations were created. One arm flex to test the movement of the upper arm. And a supination/ pronation of the hand to test the muscles of the lower arm. This was exported on each arm from Maya and imported into Unreal Engine 4, where it was also recorded. The recordings were sent to the different critics for reviews.

4.3 Method Criticism

The focus is to create arms with different topology and test them inside Unreal Engine. For the animation to take part, one had to create a rig and bind it with one of the arms.

The rig was an already finished plugin for Maya, and the skin weight was (as told) set to default. The rig creation and the skin weighting were not researched for this thesis but could aid for more believable results.

This arm was created, based on my studies from different sources. This does not mean that the arm is entirely accurate because all arms are created differently.

Also stating that on a full body character creation, the arms will be held in an abducted state, forming the T-pose (or halfway abduction, forming an A-pose). Since the body was not present, the arm was built in an idle state.

5 Results

The following are the results of all three arms inside Unreal Engine 4. A real arm was used to show and compare how real muscle and tendons move and deform.

5.1 Arm Flex

Arm 1

When flexing, the tendon of the biceps pulls itself closer to the hinge joint when it should be straight. The bulb that is the biceps when flexing, is barely visible. The radial muscle group (Brachioradialis and Extensor Carpi Radialis Longus) behaves in a normal manner as the flexing of the arm is none of their primary jobs. But the radial muscle group is also affected by the wrong deformation around the hinge joint. A small collapse occurs on the wireframe image.

Arm 2

The model presses closer to the hinge joint making the flexing of the biceps and the radial muscle group deform inaccurate, in this case, the problem is more visible than arm 1.

Looking at the wireframe, the collapse is also greater than arm 1.

Arm 3

The third arm provides the best visual appearance for the flex deformation around the hinge joint as it does not create a big gap between the biceps and the radial muscle group. But the bulb that should appear on the biceps is barely there. It is also on arm three that the triceps appear the least relaxed as it still has a bulb when it should be flatter. The wireframe collapse is, however, smaller on arm 3.

5.2 Arm Supination and Pronation

Arm 1

The three muscle families currently displayed on the real image (Extensor Group, Anatomical Snuffbox, Radial Group) is also present at the 3d mesh, though some minor cover differences.

In Pronation state, the major muscles families (Radial, Extensor, Flexor, Anconeus) is also visible on the 3d mesh.

Arm 2

Same as the first arm, arm 2 shows the three visible families on the 3d mesh.

However, in pronation state, the flexor family (orange) is not visible on the 3d mesh. It is mostly covered by the extensor family (purple).

Arm3

On the third arm, all three major families are present on the 3d mesh.

But just as the second arms pronation state. The flexor muscle family is barely visible on the

Differences

Flexing Differences

Supination Differences

Pronation Differences

6 Empiri

Three different people, one orthopaedic surgeon and two physiotherapists, who specialises in muscles behaviour, examined the different arms and gave reviews on the most realistic and most unrealistic muscle deformation. Critic 1 works as a physiotherapist, critic 2 as an orthopaedic surgeon and critic 3 also work as a physiotherapist. They were numbered, based on the time of their review.

6.1 Critics Reviews

Arm 1 Flex

Critic #1 states that Arm 1 was the least favourable deformation because it gave a very unrealistic feeling.

Critic #2 on the other hand favour arm 1 the most, because it gives the more realistic deformation of the biceps brachii.

Critic #3 is impartial, saying that there can be several reasons for this kind of deformations.

Arm 2 Flex

Critic #1 states that arm 2 had the more favourable deformation because of the bulb the biceps brachii would create.

Critic #2 favour arm 2 the least, because of the unnatural press on the bend of the arm.

However, the triceps of arm 2 was the most realistic one, as it extended the best when flexing

Critic #3 was neither in favour of arm 2, with the same reasons as with critic 2, that being, it was the least realistic deformation around the bend of the arm.

Arm 3 Flex

Critic #1 was impartial in arm 3: s deformation.

Critic #2 states that the biceps gave no major change around the bend of the arm as it should, and the triceps of arm 3 showed to much activity/tension for a relaxed arm.

Critic #3 favoured arm 3 the most, as it had the most natural deformation in terms of both the biceps flexion and the triceps extension. It was noted that the tendon of the biceps should be visible when flexed (for arm 3, this does not happen), but in some varied cases, this does not happen.

Arm 1 Supination & Pronation

Critic #1 could not find any major flaws in either pronation or supination state

Critic #2 states that there is no wrong with the deformation, except a very stiff animation movement.

Critic #3 could not find any major flaws in Arm 1.

Arm 2 Supination & Pronation

Critic #1 did not find any major flaws in this Arm either.

Critic #2 states the same as the previous arm. No flaws except stiff animation.

Critic #3 says that the flexor muscles are hard to detect in pronation states, but that this is normal in some cases.

Arm 3 Supination & Pronation

Critic #1 could not find any unusual deformations on this arm either.

Critic #2 gave the same results as the previous two arms

Critic #3 gave the same answer as the previous arm. The flexor muscle is hard to detect in pronation, but this can be normal and is not a big deal.

6.2 Critics Discussion

The criticism I received, was mixed when talking about the flexing part. Each critic had one favourite arm, but two critics agreed that arm 2 was the less realistic of the arms. The one critic that favoured arm 2 did so because it created the ball-shaped biceps. But on the other hand, it also created the wrong deformation gap, one of the main issues that needed resolving.

Another ingesting topic was that critic 2 disliked arm 2 the most but favoured its triceps movement the most. This could be that certain shaped topology movement around the elbow that created a stretch when the arm was flexing.

However, the supination and pronation movement was very similar according to all three critics.

It was pointed out by the critics, that small issues (such as the lack of flexor muscle when pronating) could be results of stiff muscles or damaged bone movement. Same with the biceps not creating the ball shape, it all depends on the person it comes from, what state or issues they have. Not everyone is created exactly the same.

7 Discussion

7.1 Personal Discussion

The purpose of this thesis was to try/ find out a retopology workflow best suited for realistic muscle movement/deformation. Though several game characters today need more than just a good retopology, finding a good retopology is the best start.

But the results became very similar to one another with some minor differences, yet neither one achieved a realistic deformation. It should be known that every person has a different set of muscle motion based on many different factors. The biggest one, in this case, would be the visibility of the superficial muscles. The arm that was sculpted would best suit that of a bodybuilder, whereas my own arm (reference arm) is not in the same shape.

The biggest difference of the deformations would be the flexing part. All three arms were similar with one another, yet not realistic in terms of the reference. A good example would be that of the biceps. Any normal biceps would curl up into a ball shape when flexed, yet on all three arms, it barely showed.

As for the Supination/Pronation movement, the biggest issue was that of the pronation. The position of the muscles was barely in place, as described in the “Results” section. This was due to the basic parameters of the skin weight. As mentioned in the “method critic” section, the skin weight was at default and therefore, the joints attached to the mesh closest to it.

Image 7.1 areas of skin weights

A realistic muscle, however, can have a different area of control. The brachioradialis muscle, for example, controls a section from the middle of the upper arm to the base of the thumb (see plate 12). To add this when skin weighting, could have improved results.

It was clear that the least realistic arm was Arm 2, showing clear results of the issue stated in the theory section when a limb deformation goes wrong. Yet, as critic 1 stated, arm 2 showed more sign of the biceps bulb than that of arm 1 or 3.

Yet, these results show a great way to start an organic pipeline. Arm 1 was the most basic one and therefore the quickest to create, it was also the arm with the smallest amount of topology. The results were acceptable and given more work, could create more realistic results.

Arm 3 also gave an acceptable result but at the cost of time. It was also the arm with the highest amount of topology. This shows that a retopology following the structures of the superficial muscles system is sometimes not optimal in terms of just a basic rig setup.

8 Conclusion

Spending time, learning the muscles and their functions have helped to understand a lot in terms of believable results that will go a long way if one is to create an organic mesh.

As for the construction of the arms. Looking back at the issue, how well the arms muscles deformations fit in a 3D arm, a retopology and rig animation alone is not enough for believable results, but it shows that you do not need the most advanced topology to make limb deformations work.

8.1 Future improvements

Muscle rig. As mentioned in the discussion section, A muscle rig would create a relative realistic body movement. It requires, however, a more complex pipeline than just a simple rig creation.

Skin Weight. A research and additional methods with skin weight would greatly improve results.

Motion Capture. Having a real-life animation of an actual arm would greatly improve the study of muscle deformation.

More Topology. There are more types of topology workflows that can be tested if given more time.

More muscles. This thesis was focusing on the arms. But to fully grasp the understanding of the muscles, one must know almost every muscle in the body. Where you make one certain move there are several muscles at play.

9 Plates

Plate 1¹ Plate 2²

Plate 3 ³ Plate 4⁴

1Public Domain, https://commons.wikimedia.org/w/index.php?curid=1170932 2https://commons.wikimedia.org/w/index.php?curid=6936349

3 https://commons.wikimedia.org/w/index.php?curid=17280382

4By OpenStax College - Anatomy & Physiology, Connexions Web site. http://cnx.org/content/col11496/1.6/, Jun 19, 2013., CC BY

Plate 5⁵

Plate 6 ⁶ Plate 7⁷

5https://commons.wikimedia.org/w/index.php?curid=10158703 6https://commons.wikimedia.org/w/index.php?curid=27796932

Plate 8⁸ Plate 9⁹

Plate 10¹⁰ Plate 11¹¹

8https://commons.wikimedia.org/w/index.php?curid=24764500 9https://commons.wikimedia.org/w/index.php?curid=27796927 10https://commons.wikimedia.org/w/index.php?curid=24693053

Plate 12¹²

Plate 13

12Created by own design based on images on this site: https://human-anatomy101.com/muscles-of-upper-limb-

10 References

Websites:

1. Ipfs.io. (2018). History of anatomy. [online] Available at:

https://ipfs.io/ipfs/QmXoypizjW3WknFiJnKLwHCnL72vedxjQkDDP1mXWo6uco/wiki/H istory_of_anatomy.html [Accessed 05 June 2018]

2. Polygon. 2015. The Reconstruction of Lara Croft. [online] Available from:

https://www.polygon.com/features/2015/7/10/8925285/the-reconstruction-of-lara-croft- rise-of-the-tomb-raider [Accessed 05 June 2018]

3. Opentextbc.ca. (2018). 9.5 Types of Body Movements – Anatomy and Physiology.

[online] Available at: https://opentextbc.ca/anatomyandphysiology/chapter/9-5-types- of-body-movements/ [Accessed 05 June 2018]

4. Justin Slick 2018 3D Model Components – Vertices, Edges, Polygons & More. [online]

Available from: https://www.lifewire.com/3d-model-components-1952 [Accessed 05 June 2018]

5. James Taylor 2015 Why Are Triangles Bad When Modelling? [online] Available from:

http://www.methodj.com/why-are-triangles-bad-when-modeling/ [Accessed 05 June 2018]

6. Jahirul Amin 2013 Maya Modeling: Polygonal Modeling Theory. [online]Available from: https://www.3dtotal.com/tutorial/1754-maya-modeling-polygonal-modeling- theory-by-jahirul-amin-character-face?page=2 [Accessed 05 June 2018]

7. Polycount. 2018. Topology. [online]Available from:

http://wiki.polycount.com/wiki/Topology [Accessed 05 June 2018]

8. Polycount. 2018. Limb Topology. [online]Available from:

http://wiki.polycount.com/wiki/Limb_Topology [Accessed 05 June 2018]

9. Polycount 2018 ReTopologyModeling. [online]Available from:

http://wiki.polycount.com/wiki/ReTopologyModeling [Accessed 05 June 2018]

10. Ben Mathis 2016 Limb Deformation Tips. [online]Available from:

http://www.poopinmymouth.com/tutorials/limb-deformations-tip.html [Accessed 05 June 2018]

11. Youtube. 2018. 3DCharacter Workflow For Beginners Tutorial. [online]Available from:

https://www.youtube.com/watch?v=cn0z9yJkR2s [Accessed 05 June 2018]

12. Polycount 2018 Polycounts in next gen games thread! [online]Available at:

http://polycount.com/discussion/141061/polycounts-in-next-gen-games-thread [Accessed 05 June 2018]

Articles:

13. Tapio Terävä 2017 Workflows for Creating 3D Game Characters. UAS., KAMK University of Applied Sciences

Books:

14. Valerie L. Winslow, (2015), Classic Human Anatomy in Motion, 3 editions, United States: Watson-Guptill Publications.

15. Richer, Paul (1986). Artistic Anatomy. France: Watson-Guptill.

16. Zarins Uldis (2014). Anatomy for Sculptors. USA: Exonicus LLC. (e-book)

17. William Vaughan (2018) Topology Workbook: Volume 1.Clermont, FL: Hickory Nut Publishing (e-book)

11 Appendices

Arm animations:

https://drive.google.com/open?id=1QTK3Lku22dr90soCfgbtf1HEsL-kXOvw Zbrush:

https://pixologic.com/

Maya:

https://www.autodesk.eu/products/maya/overview Substance Painter:

https://www.allegorithmic.com/products/substance-painter Unreal Engine:

https://www.unrealengine.com/en-US/what-is-unreal-engine-4 Sculpting:

https://www.youtube.com/watch?v=i6KKp8UpjYI