essential techniques igitald ARACTER MATION 2 I i Maestri George leu Riders | www.newriders.com[ @ ig i CHARA ANIMA Volume I: Essential Techniques tad] CTER TION 2, Animating characters is a life-long journey—this book will help you take the first few steps! Animation is the art, computers are only the medium, Whether you animate with pencils, clay, or pixels, you still need to know the basics @ You n Whether you use pencils or pixels, the fundamentals of animation are the same. Computer-generated 3D animation requires many of the same character animation techniques that traditional animators learn, Ciarrer 1: Basics of Design Citarter 2: Modeling Basies Chartres 3: Modeling Bodiess for Animation CHavTeR 4: Modeling Hands for Animation CHAPTER 5: Modeling Heads for Animation CHANTER 6: Skeletons and Mesh Deformation Charter 7: Posing Characters Cuarren 8: Basics of Character Animation Cuarrer 9: Walking and Locomotion CHAPTER 10: Facial Animation and Dialogue APPENDIX A: Suggested Readingsdigital character animation 2, volume I: essential techniques george maestriPublisher David Dwyer Executive Editor Steve Weiss ‘Acquisitions Editor Laura Frey Development Editor Audrey Doyle Technical Editor Angie Jones Managing Editor Sarsh Kearns Copy Editor Naney Albright Indexers Cheryl Landis Craig Small Software Development Specialist Jason Haines Proofreader Linda Seifert ‘wil Cruz digital character animation 2 Digital Character Animation 2, Volume I: Essential Techniques By Geomge Maes te by Riker Pabing 201 Wet (de Steet, Inanspols IN 44290 USA ‘Copyrigh © 1999 by New Riders Publishing [A eights reserved. No putt of tis book sal be repxhice ore i retrieval ster, ot ‘mans by any means, eleeonic, mechanncal, photocopying. recordin. of otherwis,wthout \wntten permission font the publisher No putes sity sumed with rspcet 9 the eof the information contained herein. Alkhough evety precaution has bee taken the prepara ofthis ‘book: the pulser ad authors) ssame no expat fr erors or omisions. Neither ie ny liabileyasuined for dantages resulting fm the we of dhe formation contained her, International Standard hook Number: 56205-9800 Library of Congres Catalog Cand Number: 9-512 Printed inthe Unite States of Ameria Fist Peng: Aug 198) 3 76543 Invcrpreation ofthe printing codeThe righunon doublet mabe nthe year ofthe book’ Brinting: the rightiost single-digit number is the mmber ofthe books printing For example, the Printing code 99-1 shows that she fint printing ofthe book oscureed in 199%About the Author Dedication Acknowledgments Publisher's Note acknowledgments George Maestri has worked as a writer, director. and producer in both traditional and computer animation for such companies as Nickelodeos Warner Brothers, Disney, Comedy Central, Film Roman, MGM, ABC, CBS, and Fox, among others, He was the original animation producer on the Comedy Central series, "South Park” and was nominated for a Cable Ace award for writing on the Nickelodeon series," Rocko's Modern Life:” He has written several books including, /digital/ Character Animation for New Riders Publishing, and numerous articles on computer animation for magazines such as Digital Magic, Computer Graphics World, DY, and New Media, among others. This book is dedicated to Moyet and Preston Many thanks to Angie Jones for being a wonderful technical editor as well as my conscience. Albo thanks to her for many of the animations on the CD, as well as for modeling most of the facial animation targets used in Chapter 10. Thanks to Tuong Nguyen for the great sculptures wed in Chapter 1. Thanks to the people on the CG-CHAR list for all the great advice. ‘Thanks to Audrey Doyle for being a terrific editor. Thanks co Laura Frey, Steve Weiss, Alicia Buckley, David Dwyer, and everyone else at New Riders for letting me create this book as [ envisioned it New Riders would like to congratulate author George Maestri for creating this [digital] masterpiece. A special debe is owed to Jennifer Eberhardt, Sarah Kea Seifert, and Wil Cruz for sharing their amazing talents for this project. s, Linda,Trademark Acknowledgments Warning and Disclaimer digital character animation 2 All terms ntioned in this book that are known to be erademarks or service marks have been appropriately capitalized. New Riders Publishing cannot attest to the accuracy of this information. Use of a term in this book should not be regarded as affecting the validity of any trademark or service mark. Every effort has been made to make this book as complete and as accurate as possible, but no warranty of fitness is implied. The informa- tion provided is on an “as is” basis. The authors and the publisher shall have neither liability nor responsibility to any person or entity with respect to any loss or damages arising trom the information contained in this book or from the use of the CD or programs accompanying it.acknowledgments =z | Us What You Think! ‘As the reader of this book, you are our most important critic and com- mentator. We value your opinion and wane to know what we're doing, right, what we could do better, what areas you'd like to see us publish in, and any other words of wisdom you're willing to pass our way. As Executive Editor at New Riders Publishing, | welcome your comments. You can fax, email, of write me directly to let me know what you did or didn't like about this book—as well as what we can do to make our books stronger. Please note that I cannot help you with technical problems related to the topic of this book, and that due to the high volume of email | receive, I might not be able to reply to every message. ‘When you write, please be sure to include this book’s title and author, as well as your name and phone or fax number. I will carefully review your comments and share them with the author and editors who worked on the book. Fas: 317-581-4663 Email: [email protected] Mail: Steve Weiss Executive Editor New Riders Publishing, 201 West 103rd Street Indianapolis, IN 46290 USAdigital character animation Contents at a Glance Introdtetion ....0- Basics of Design Modeling Basis Modeling Bodies for Animation Modeling Hands for Animation Modeliag Heads for Animation Skeletons and Mesh Deformations Posing Characters Basics of Character Animation Walking and Locomotion Facial Animation and Dialogue Suggested Readings: Index 68. 88 116 4 164 208 236table of contents Table of Contents 1 Basics of Design Approaching Design as an Artist Knowing Your Character .. Making Decisions Categories of Design... Realistic Designs Stylized Designs Designing a Character Head and Body Proportions w...snnnnunnenesnn Hands ... 2 Clothing and Other Accessories. Designing Your Own Charactets «... Designing on Paper... Using Images as Reference Designing in Clay Getting Sculpture into the Computer 2. Modeling Basics Surface Types Polygonal Surfaces Patch Surfaces eos NURBS Surfaces Modeling Strategies Keep It Light ... . Mixing and Matching Surface Types .. Modeling Basics ... Basics of Polygonal Modeling Polygonal Modeling Techniques «jn Subdivision Surfaces... Booleans Are Evil. Basics of Patch Modeling Creating Patches from Curves Stitching Patches Together ‘Three-Sided Patches .. Modeling with NURBS Curves on Surface... Joining NURBS Surfacesdigital character animation 2 Metaballs. 44 The Trouble with Metaballs 46 Animating Metaballs 46 Conclusion 47 3 Modeling Bodies for Animation ‘The Structure of the Body The Skeleton Muscles Construction Methods Sogmenting Bodies 0 usnatene Seamless C ts . : Modeling a Seamless NURBS Body 64 Conelusion 4 Modeling Hands for Animation Examining the Hand’s Structure .. Flexibility of the Hand and Fingers The Fingers .. The Thumb . Modeling Methods for Hands Modeling Segmented Hands Modeling Hands from Polygons .... Modeling a Hand with NURBS Modeling Metaballs Hands. Adding Realism Fingernail. Knuckles ‘Tendons Webbing Conelusion os 7 5 Modeling Heads for Animation 88 ‘The Structure of the Head and Face si 9 The Muscles of the Face semen on Basic Head Modeling Methods 93 Paste-On Features .. 3 A Face with Flexible Skin cstnesnntneeeee A Topology and Facial Modeling 1098 The Pole at the Mouth 97 The Pole at the Ears 98 Which Topology to Use ” More Complex Heads 102 Modeling a Polygonal Head ~ 103,table of contents Eyes... Internal External .. Asymmetrical Eyes... Eyebrows. Teeth Tongues . ‘Connecting a Head and a Body Neck on the Body Neck on the Head . Head, Neck, and Body as One Object Conclusion peste 6 Skeletons and Mesh Deformation Hierarchies and Character Animation. Manipulating Hierarchies Pivots and Joint Rotatio Skeletons and Hierarchies. Manipulating Hierarchies Forward Kinematics Inverse Kinematics ‘ Which Kinematics 10 Use? Building Skeletons, . Building a Body Skeleton Hand Skeleton . Mesh Deformation .... Direct Assignment... Weighted Assignment/ Envelopes Advanced Deformation Tools Deformation and Surface Type . Skeletons and Segmented Characters Other Deformation Methods Lattices . Clusters Spline Deformation .. Conchasion sone . 7 Posing Characters 144 Posing 146 Animating with Poses 146 Pose-to- Pose Animation 146 Straight-Ahead Animation 147 ‘Combining the Two Methods .. 47 Posing a Flour Sack 7 a lAB, Setting Up a Flour Sack 149 Posing the Sack «ors 149digital character animation 2 Posing Characters Balance Symmetry .. Weight Staging Creating Strong Silhouettes Line of Action .. Staging Multiple Characters. Staging and Camera Placement Moving the Camera Conclusion 8 The Basics of Character Animation The Importance of Timing ‘The Process of Timing .. How Much Tim ‘One Thing at a Time Developing Good Timing Timing and Software Curves... Dope Sheets Paths/ Trajectory. Using Timing to Suggest Weight. The Language of Movement Arcs Slow-ins and Slow-outs. Force and Drag.. Squash and Stretch . How to Squash and Stretch a Character « Anticipation... Overshoor Secondary Action Overlap and Follow-Throu Cycles Moving Holds... Purting It All Together Conclusion . 9 Walking and Locomotion ‘Understanding the Mechanics of Walki Feet and Legs ‘The Hips, Spine, and Shoulders . The Arms ‘The Head and Spine from the Sidetable of contents Avimating a Walk ‘Timing the Walk Keeping Your Feet on the Ground To Cycle or Not to Beyond Walking ‘The Run The Skip .. ‘The Sneak Animating Four-Legged Walks Conclusion .. 10. Facial Animation and Dialogue Methods for Animating the Face Morphing ..... = ngle-Target Morphing, Mulkiple-Target Morphing .. Modeling Morph Targets for Animation Lower-Face Poses. Upper-Face Poses Upper-Face Poses for Stylized Characters Other Facial Poses... Assigning Targets a Manipulating the Face .... Creating Facial Expression. Basic Expressions The Brows and Expression ... Symmetry and Expression Complex Expressions se Animating the Head and Face . Amimating the Head... Eye Direction, Blinks ..... (Overlap in Facial Animation Animating Lip-Syne Recording Dialogue . ‘The Fight Basic Mouth Pos Reading the Track... Animating Dialogue Mouth or Body First? Animating the Mouth Eyes and Dialogue Head Motion and Dialogue Body Language and Dialogue Conelusiondigital A Suggested Readings Index character animation 2 268 272oduction king with a computer is very similar to writing with a word ot. Both are creative tasks that just happen to involve technology. ‘cases, creativity, not technology, is what matters. People—not are creative. Technology is simply a tool that helps things processor may make your documents look stunningly beautiful, of fonts in full color, but it does not make you a better Writing is a creative task that comes mostly from the right side ns brain. The word processor doesn’t read the classes, it idy poetry, and it can't create a brilliant novel, script, or ‘paper—it takes a creative person to do that. The word processor Hmakes it easier to get the words out of your head and onto a page so people can read them goes for computer animation. You may have the best software in the world, and it may render each frame of tion beautifully and quickly, but the content of those frames by your knowledge and creativity. Character animation is a f motion, timing, and acting, and is just as creative an art as, drawing, or painting. Like the word processor, the er is simply a tool that makes it easier to get the animation out d and onto film or videotape so that people can see it. you animate with a CPU doesn't mean you're some kind zer in the world of animation. Digital animation may be new, but people have been animating since motion pictures ented. The skills and knowledge that animators discovered long foundation upon which computer animators now stand. The does make the creation of animation easier and quicker, but it easier and quicker to make garbage. In order to produce tion, you need a firm grounding in the basies of anima- ch have not changed since the golden age of animation over “The same principles and techniques that went into the toons of the 19408 are principles and techniques every should know today, regardless of the medium. n is the art, computers are the medium. Whether you animate cil, clay, or pixels, you need to know the basics of animation.digital character animation 2 You need to understand anatomy foundations of animation. Some motion, weight, jimators may dav, but all stand on the same foundation id timing—the ome may sculpt, What You Can Expect from This Book For this second edition, the book has been broken up into two volumes. This volume is meant to ¢ ve you the basics of designing, building, and animating characters. The second volume goes deeper into the technical issues of animation and character building, as well as the art of making films. What You Should Know Before you begin this book, you should have a fundamental understand ing of computers, computer graphics, and 3D software. This book will not bore you with a dissertation on CPUs, RAM, and hard disks; nor by expl ou don't know these terms, then you need to study will it bother y and rendering. If the fundamentals of computer graphics. You should also be familiar with your chosen 3D software. Work through your software’s manuals and tutorials to understand its features ning such things as pixels, alpha channels, What You Will Need All you need to animate characters digitally is a computer and a 3D package. An extremely basic package will only handle simple characters, however, and some packages lend themselves to character animation more readily than others. Look for packages that support features such as shape animation, skeletal deformation, multiple target morphing, inverse kinematics, among others, Beg! before buying any hardware. Each hardware plattoray has its ig animators should decide on the software they want to use is points, but its the software that will animate your characters. When buying hardware, always try to buy the fastest machine with the most memory. Systems become obsolete so quickly, so an investment in a fast machine will last much longer than an investment in a slower one. Of cit is still entirely possible to produce terrific ind used equipment can be the basement animator’s secret weapon. Quality is not a function of processor speed. animation on almost any budg¢introduction ‘eal lure of digital animation is interactivity, or the ability to play ‘your animations on the fly and make changes immediately. The fer your machine, the higher the level of interactivity and the her the flow of the cteative process. To help speed things along, accelerator card helps you play back shaded animation tests ctively a8 you create, Not all 31D cards are created equal, so swith your software vendor for a list of supported 3D cards you buy: important issue is output, Almost any computer can play back [sized animations directly on your computer screen. For your this type of playback is perfectly acceprable. you want to videotape, however, you will need to include a full motion card for final output. These cards come in many flavors—some analog video, some only digital video, some both. This will depend on your choice for final output. Ifyou own only a eck, then analog will have to do. Once printed to tape, anybody ‘your creations. it This Book's Approach to ware sid have been much easier to write this book using specific A tutorial using Brand X software would prove to be hat useless co the person using Brand ¥ software. Added to that et that as soon as Brand X got a slew of new features or a new: this book would be headed for the trash, Besides, there’s a book about the sofeware you own; it’s called the manual. ide to anyone with a es presented here are done ge-neutral standpoint. Hopefully, this book will be applica~ | broad range of platforms. To this effect, the book focuses on es of these packages that are similar, ‘you work with different 3D packages, the move you will similar they really are to each other. Polygonal modeling ge is very similar to polygonal modeling in another. Some logy may be a bit different between packages, but the try and placement of detail on the model is identical.digital character animation 2 This book does have to jump through a few hoops when it comes to terminology: The exact same feature on one package may be called something completely different on another package. When a conflict existed, we tried to choose the term most commonly used among all packages A Final Note Animating characters isa life-long journey, and this book will help you take only the first few steps. Animation can be incredibly fin, but it ako a very difficult art co master. After finishing this book, it will take many years of practice to become a true animator. Hopefully, you will rely on this reference throughout your journey.\ ubisaq fo soisegBasics of Design his book is about creating and bringing characters to life. Before you model and animate a character, however, you need to think about its design.A character design can be as simple as a sketch, or as omplex as fully rendered sculpture. The design serves as a simple road map. a plan of action for creating a character digitally Design means making decisions about your characters. You need to decide how tall or shore they'll be, the textures you'll use for their skin, and what type of clothes they'll wear.You also need to decide just how realistic or stylized you want to make your characters. A realistic~ looking character means it should be animated realistically as well. Designing your characters properly makes their personalities jump off the screen Your audience will know who your characters are immedi- ately—and like them. A well-designed character is also a character that will be easy for you to animate. If you can animate your character with out fuss, it makes the animation process easier and far more creative. Designing characters that animate well requires a knowledge of anato- iy as well as a good understanding of the strengths and weaknesses of your software tools.ital character animation 2 Approaching Design as an Artist Just like animation, design is an art. When great design is coupled with ‘great animation, the results are much larger than the whole. There are many ways to approach the design of your characters. Design is a ere- ative process, so everyone approaches the task using his or her own artistic strengths and weaknesses. Some people work out their designs with pencil and paper, others prefer clay, some prefer pixels, All these methods can inspire new and innovative designs. Knowing Your Character When designing a character, you first need ta know a bit about him or her. Is the character young or old, short or tall? Is the character the hero of the villain? Is the character the star of the film oF on Aggressive or meek? All these criteria, plus many more, will factor into the designs of your characters Many times, the script provides clues to the character, and perhaps even a written description. If you are designing a character for your own pro- ject, you may want to write down some of the major character traits before you begin the design. A short biography of the character might help as well. Making Design Decisions “Those who want realistic characters may simply place a human in a laser scanner and snap a digital picture of the real thing. Deciding who steps into the scanner, however is a design decision, [eS important that you consider your character from a design standpoint before you ever touch hand to mouse. First and foremost, a well-designed character conveys personality. Your characters should abo be well-proportioned and appealing to the eye. ight fully gruesome way in addition to your character’ outward appearance, you ako need to design with animation in mind. This meazs under standing your sofware and what it needs to animate a character successfully Even the villain needs to be appealing—particalay if iis in a del Some of your design decisions will be dictated by the limitations of your software, If your sofewa instance, you might want to avoid animating an Old English sheepdog, © doesn't allow you to create hair, forchapter basics of design You should use the strong points of your software to your advantage and design around the weak spots. No one will know that you didn't Ihave that whizbang new hair plug-in. Instead, the audience simply sees a character on the screen. If it looks good, and has character, it doesn’t ‘matter how you created it, even if it’s bald. Simplicity is the ke amazing character turns out to be the simplest to construct. You should alvays think about your character's construction and try to keep your tnodels simple and light. Simple characters also animate faster and easier. to designing good characters. Many times, the most Ifyour character has too many parts or more detail than he needs, you will have more problems to keep track of when animating, Problems that slow you or your system down will detract fro processes and cause your animation to suffer. your creative Good design means knowing your software. This dog has Without the custom hair software, the dog becomes, Joi5 of hair, which requires custom software to create. a different character e vq Tis character is very simple, but still can be animated Even a character with no body, like this sun, can ite effectively. have lots of character.Audiences expect realistic characters such as this to move realistically digital character animation 2 Categories of Design Designs designs try 10 mimic reality. If this is the fall into two broad categories: realistic and styhzed. Realistic sou need 10 d n your ized designs are caricatures of character in accordance with nature. St reality and give you far more design choices. Realistic Designs Digital character animation has been very successful in the area of special effects for f ure films, Computer animators and special effects, teams have been able to make characters, such as dinosaurs, look cont= pletely the larger studios are working toward creating digital human sure they will succeed. Digital stuntmen have already been employ land integeate them into live-action environments, Many of complete stunts that were too expensive o impossible for a real stunt- nan to do. Ifyou want to create characters that closely mimic reality, digital anima- tion is the medium you should choose. I must warn you that reality is a hard thing to simulate, regardless of medium. Creat animation of a realistic character requires a thorough knowledge of an effective anatomy and motion,ou want to design and animate a realistic character, reference of the thing is essential. In traditional animation, many of the large studios, been known to bring animals to the studio for the animators to Ifyou can’t afford to hire an ostrich or an elephant for the after- ion, trip to the 200 or a videotape of a nature documentary can vide equally good reference. If you want to create a realistic human, tip to the 200 could also be in order. Of course, for humans, our jes are our z00s. In a big city, his means stepping, out your front door d watching the passersby. lized Designs a caricatured world is usually more fin and gives you much freedom in your designs. Computer animation can mimic reality, imation is always at its best when it goes far beyond reality: Think classic cartoons from the 1940s. Dafly Duck was a wild and character, but he was a duck. In real life, ducks swim, eat, and eed, but thac’s about it. Ducks certainly don’t talk, and they don’t hold idges against smart-aleck rabbits. I'S way beyond the realm of reality, ut a talking duck is still entertaining. The audience will accept the an red universe as you, the animator, define i. If your character pulls off hhead and dribbles it like a basketball—we dor question, we simply the audience's preconceptions are important. If your audience sees a eture-pertect digital human on the screen, it will expect him to walk, and act like a human. If he’s the slightest bit off from the way a real man acts, the illusion is lost. I'you give the audience a caricatured 1o0n hedgehog, however, there are absolutely no preconceptions. The dience will be much more aecepting of your character, and you, as the animator, will have the freedom to make him move and act however ‘want—even if it’s not exactly the way a real hedgehog would act. Caricatured design, however, is not a license to animate your charac~ poorly. Good animation makes even the simplest character appear to breathe. A simple box can be animated quite easily by ewisting, turning, moving it about so that it has weight, volume, and personality. The a stylized character such as this car- important thing about a character is the personality and whether toon cat may be easier to design and hat personality engages the audience. animate than a realistic human chapter 1: basics of design nas‘A realistic characteris between 6 and 8 heads tall digital character animation 2 Designing a Character ‘A characters design depends largely on the character and its personality Your character's design should also be an indication of its personality. and its role in the film. A character who's big and mean will have broad shoulders and beady eyes. One with big eyes and a potbelly will seem meek in comparison. A character with lots of big muscles a very small head might appear a bit clueless, and the character with a huge head and no muscles might appear to be a g OF course, these examples are stereotypes, and it is often a good idea to play against type.A good example might be a large. threatening charac- ter who has a small voice and timid personality. Design can also drive the personality of a character. A duck born with no feathers might be driven by his inadequacy to corner the market in down pillows, for Head and Body Proportions When deciding on the height of your character, use the size of the head as the guide, An average human is about 6 to 8 heads tall. If the charac~ ter is taller, it may appear more lithe and graceful.toon characters, on the other hand, can have much bigger heads in ‘elation to the body. Mr, Potato Head, in fact, has no body, he’s simply a “giant head with arms and fet. If your character has a body, it may be as large the head itself. A larger head in proportion to the body ds to make a character look cute. M characters are only 2 or 3 Ths cartoon character isa little more than 2 heads tal. The bigger the head in relation to the body, the cuter the character will look. size of the eyes in relation to the face determines how we perceive er. the eyes are small and beady, the character may appear gan or angry. Big eyes convey innocence and look cute. Placement of eyes albo plays a huge role in the character. Ifthe eyes are low on e head, the character seems older and more brainy with a bigger fore- d Placing the eyes higher on the head tends to make the character wk younger. he eyes of your character can also change shape. If your character has, eyes, you can bend and flex the eyes wildly with the character's If your character has eyes chat are inside the head, and more you don't have as much freedom in changing the shape at ani- time. If you want the eyes bigger than normal, you need to them big at the beginning, chapter 1: basics of design aeLe] digital character animation 2 2 @ The shape of your character’ eyes can Eyes that are wildly asymmetrical can Narrow eyes, such as these, can give also determine its personality, Rounk make your character look like he's beg- a character 4 mare devious look symmetncal eyes can look pleasant, ging for neurons. though a slight bit of asymmetry will add realism and character. Feet la foot should be almose exactly as long as your forearm. Many times your On real humans, feet are relatively small in relation to the body. Your cer will be wearing shoes. In this case, it is not necessary to model feet to place inside the shoes, the shoes can simply attach to the legs at the ankles ted A cartoon character, on the ether hand, may have the feet exaggera so that they are much larger than normal. If you decide to exaggerate the feet, remember that your character still needs to walk. Feet that are slightly longer than the legs can be animated using a number of tricks, such as walking with a waddle or the bods not impossible, to an 4. & This stylized character has feet that are bigger Too big, however, and he'll have problems walking, than normal ing the legs as the feet pass under Feet that are a lot longer than the legs will prove difficult, ifchapter 1: basics of design Hands Tike feet, hands on a stylized character can be slightly oversized to give ita cartoonish Jook. If your design is more realistic, you might want to jortion the hands realistically as well. If realism is the goal, you may to attach your hands seamlessly to the forearm. This may mean yg your character as one solid mesh or using an advanced seam- technique, such as a blend. to keep the skin smooth, These tech- ques are described later Here we have a character whose hand and body are one seamless mesh. avant, you can give your character cartoon gloves for hands, In the cartoon gloves were devised by animators so that hands and would be easier to draw: design stuck, mostly because it looks stylish and it works. For a 3D \ jr, cartoon gloves serve much the same purpose. They can be An easier to attach than a naked hand, because the glove can have a where it attaches, A cartoon glove is stylish, and it helps you design around the tricky problern Of seamlessly attaching your characters hand to the waistDo clothes make the man? The charac- ter on the left may be perceived diff ently than the one on the right, simply because of his clothes Clothing that behaves like real fabric, however, will equire custom software, which may affect your design decisions digital character animation z Clothing and Other Accessories Clothes make the man, or so they say. If your character is an astronaut, he will probably be wearing a space suit. A caveman may need only a bearskin. Clothing tells the audience, to a large degree, who the charac~ ter is and how he should be perceived. If you want a character to be mean and nasty, dress him accordingly. This doesn’t mean that every vil- lain you create needs to wear black, however. When designing characters and their clothing, try to be original and avoid stereotypes. Clothing can also help you design around the limitations of your model- er. If you model your character in a manner that forces you to place a good way to hide this. A dress oF a lor shire can help hide the hip area, which can be difficult to model and 2 that behaves like real fabric, how- ever, requires custom software, which may affect your design decisions, seam at its waist, a bele may be animate with some software. ClothitwT m1 may also be used to hide seams. This character The shirt hides the seams at the arms and neck, and the wel dressed ppants hide the seams for the feet ter has an unsightly seam at his waist. ‘Add a belt and the audience will never see the seam. igning Your Own Characters you have a bit of information under your belt, you can start your own characters. How you realize your designs depends skills and abilities. Some people like to draw, others like to and some like to work directly with the computer. chapter 1: basics of design =digital character animation 2 De: ‘The simplest way to design your characters isto sketch them on paper. A pencil is a wonderful thing, because it enables you to very quickly block out the size and shape of your character. If you know how to write your name, you can most likely draw a character. ining on Paper Ifyou've spent any amount of time modeling characters on the com- puter, you'll develop a visual sense chat will ranslate to paper as well. Ie seems that modeling characters improves your drawing, and drawing, characters improves your modeling. Another way to improve both skills is through a life drawing class. This is che single best way to aesthetically understand the human form, Regardless of your skills, after you've got a few sketches of your char- acter, you can use them as reference for modeling the character. Many people, however, need a more precise guide. This is particularly true when you need to model an existing character, such as the latest comic book hero who's appearing in your company’s new video game. In these cases, you need to draw several accurate views of your charac ter. Usually, this means a minimum of a front view and a side views When drawing these views, make sure to draw the character with its arms outstretched. This is the standard way of modeling characters, because it makes mesh deformation a lot easier. The final drawings are then scanned into the computer, where they are used as reference. However your character is designed, its a good idea to make an accurate drawing before starting the modeling processchapter 1: basics of design Using Images as Reference are several techniques for using your drawings as reference. Most only, the scans are simply used as a background image. This means configuring your software so that the image takes up the viewport of your odeling program, enabling you to trace the contours of the image. er way is to create a simple object, such as a plane or a thin box has the images applied as texture maps. This allows you to model imteractively, because you can move the camera to get custom views the character as its bei “After the drawing is scanned into the ‘computer, it can be mapped to a sees of planes, which act as references ven modeling The final, and very oldetashioned, way is not to scan the image, but to use a igitizing tablet to trace the drawing This method does work, but itis cer tainly far from interactive way to design characters is to sculpt them in clay. A three onal clay sculpture is much closer to the final character than dimensional drawing can be. Clay, however, is much messiet, tainly a great deal more time-consuming, sculpting is an art. There are plenty of books and classes to teach the finer points of sculpture, Those who understand ling should take to sculpture quite easily:You will probably of clay is far superior to any computedigital character animation 2 ‘As with drawings, a proper sculpture used for modeling characters should be in a neutral pose. The arms shoul! be outstretched with the palms forward and the legs tightly apart. Ie’ always tempting to sculpt your character in a dynamic pose. Although it may make a nice sculp- ture, don't succumb to this temptation—it will cause you nothing but headaches. If your character is sculpted properly, in a neutral pose, your digital character will be able to hit any pose you desire, Using Sculpey Most people who realize their designs as sculpture tend to use Sculpey Sculpey is a soft polymer clay that can be baked to a hard finish in a conventional oven rather than a kiln. This makes it much more acces ble for the average person, There are sewtal types of Sculpey. The original Sculpey is 2 plain white clay, and is the least expensive. For most purposes it is perfectly fine, but iv prone to cracking over time. Super Sculpey is more durable, and Sculpey III comes in designer colors. Which one you choose depends on your requirements. For character models that are disposable, Sculpey is more than adequate. Those wanting to keep their sculptures forever may want to invest in the higher-quality lines. Sculpey isa soft clay, so it’s best used with a bit of support. This means creating an armature, or a skeleton, for your character out of wire or aluminum foil. This skeleton will help support the clay until it is baked. In addition to supporting your character, creating a skeleton out of wire and aluminum fail enables you ta use less Sculpey. Not only does it save ‘you money, but it makes the curing process much easier: Sculpey needs to be baked at 275 degrees Fahrenheit (130 degrees Centigrade) for 15 minutes per quarter-inch of thickness. If your character is made out of pure Sculpey it will take a 4-inch-thick character 4 hours to bake, which means your outer layers will probably singe and burn long before the center is fully cooked. Always try to keep the Sculpey Jess chan an inch thick, After Sculpey is cured, it forms 2 very tough finish. Even a half-ineh-thick character can be sanded quite vigorously without breaking, though you should always be careful. For decoration, Sculpey ‘can ako be painted with standard acrylics.chapter 1: basics of design Getting Sculpture into the Computer Alter you create an acceptable sculpture, there are wo methods for get fing ic into the computer for modeling. The first method is to simply fake a photograph of the sculpture and scan it into the computer. The second method is to digitize the actual surface of the model, using a 3D Scanner or digitizer. The one you choose depends on your needs and ako your budget. Photographing Sculpture The simplest and cheapest way to get your sculpture into the computer isto take photographs of it at several different angles. Typically, a front View and a side view will sutice, but extra detail may need more ang Jase as with drawings, th computer and used as reference for modeling, photographs can then be scanned into the When photographing a model, try not to use a short focal Jength, or “fsh-eye” lens on your camera. This will distort the image, making, ‘modsling difficult. You should also try £0 light the model so that the contour of the sculpture is apparent. A washed-out photograph provides ‘very little useful information. Baulpture by Tuong wguyen) Photographing a sculpture is an easy way 10 get it into a computer. Take several diferent views of your character.=D) digivsl charaever anima Photography can also be used as reference for real-world items. A photo graph of a toy, an animal, or even a human can work just as effectively as a photograph of a sculpture. The photograph can then be scanned into the computer and used as refer- ence, just as with a drawing, Digitizing Sculpture ‘The second, and more complex, method is to digitize or scan the model. There are a number of digitizing devices on the market, which range fiom small mechanical digitizers to human-sized laser scanners. ‘There are also many service bureaus that can do the work for those without the money to purchase such expensive equipment, ‘There are two main ways of digitizing a sculpture. The first method, employed by most laser-based systems, scans the model in a grid-like fashion. This creates a dense cloud of points, which duplicates the sur- face quite accurately. Unfortunately, such a dense model will not ani mate with any degree of control. The dense model must be brought into the modeling package and used as a three-dimensional reference. From there, the animatable model is built by snapping vertices to the surface of the reference.basics of design A laser scan creates a model that is both dense and grid-lke. The scan must be used as reference for creating ‘8 model that can be animated. ial digitizers can ako scan clouds of points, but most of the ad-operated ones enable you to specify exactly which points are di | This makes it possible to draw out the copology of your wire eetly on the sculpture, The detail can then be digitized very precisely, feally modeling the character as you digitize. this frst exercise, you design a simple character. First, think about what pe of character you want to create. Is it real or stylized? What are that ters major personality traits? When you have a good idea of who ur characteris, ake a pencil and paper and sketch some character gns that you might want to model in the next few chapters. (If you're familiar with sculpting, you might want to do this in clay) These Jhen you have a number of designs, go through each character and think that character might be constructed in your chosen software. You st getting started, so these issues may not be readily apparent. Your gn may change as you learn more about how to build characters. This [be discussed in the next few chapters,ERIE: WLU Cy Le)Modeling Basics hen a character has been designed, it needs to be built. This means translating the desiga into a 3D model that can be ani~ mated. Building a character in the computer is called modeling, This rakes the basic design of the character and turns ie into a 3D) model that can then be manipulated. The basic decisions of how a character will be modeled affect the entire animation process. When modeled, the character must not only look good, it must animate well. This means that the surface af the character can deform easily and quickly for fast manipulation. A character that can be manipulated ins real-time makes animation a breeze. Building and configuring your character properly before you ever start animating saves you headaches in the long run.digital character animation 2 Surface Types When modeling, you can use many different surfaces to create a charac ter. These include polygons, patches, and NURBS-based surfaces. In addition to these three basic types, the such as metaballs, membranes, and hierarchical patches, to name but a few. If you understand the basics of polygons, patches, and NURBS (Non-Uniform Rational B-Splines), however, you'll have the basics that will enable you to work in almost any package. ¢ are also other types of surfaces, The surface you choose for constructing your character depends on a number of factors. Some of this decision is sofiware-driven. If your soft~ ware supports only polygonal surfices, for example, your decision is already made—the character must be built out of polygons. Another factor may be the number of tools, meaning that some packages are miore adept at one type of modeling over another. A package may sim= ply have more tools to manipulate NURBS surfaces, which might make NURBS the appropriate choice for the character. For those packages that support a variety of surface types, the decision comes down to the demands of the project, the individual characters, and, of course, person- al preference. Some characters lend themselves quite nicely to polygonal surfa and others may be approached via a patch-based or NURBS-based modeling method. A character in a gaming environment, for example, may need to be polygonal simply because the game engine supports only polygons. Characcers made for high-resolution output, such as film, typically are modeled as NURBS surfaces. Patches and subdivided polygonal surfaces, however, are catching up in offering the capability of deforming and can be even better for quick interaction within the software. ees, ‘The best way to make the best decision is to understand all the surface types equally. This involves working with all the different geometry types and trying your hand at modeling and animating a character with each, Of course, after you've worked with the different surface types for a while, you'll develop your own preferences, which will affect your future decisions, Alo be alert for new methods and technologies—the world of 3D animation is hardly standing stil Polygonal Surfaces Polygonal modeling was the first form of modeling developed for computer graphics, and, because it is the most fundamental form of modeling, it still holds up quite well, All other forms of modelingchapter 2: modeling basics 16 polygons at one point or another, because no matter what g tool you use, most software turns your model into polygons endering. Polygons are simply triangles or rectangles, each repre~ plane and defining a small area of the character’ sutface. nefit to polygonal modeling is in the number of different sur- es that can be defined. Many NURBS-based and patch-based are limited to surfaces that are topologically simple, such as a ‘ora sphere. In order to create an object that's more complex, a body, you may need to make ic out of several surfaces that are ed together. Polygons don’t suffer from this topological conun- ad with a polygonal modeler, you're free to make your surface as you desire models have three basic elements: the vertex, the edge, and zon, Essentially, these elements represent the three dimensions. rtex is a single point, an edge is a line connecting two vertices, and ygon is a surface defined by three edges or vertices. Polygons can nore than three sides, however, and many packages allow as many as you want to define a polygon, though internally these many- | polygons are resolved to triangles when rendered. tripe that people have with polygonal surfaces is that, in onder 4 reasonably smooth surface, the surface needs to have lots and “polygons. When animating, however, a high-resolution polygonal jis the last thing you'd ever want to work with, It’s slow to and tears easily. around this problem is to model a low-resolution polygonal that animates easily, and then use techniques to add the extra il needed for a smooth surface at render-time. As you will see is technique is called subdivision surfaces, and it can produce results. Surfaces atch is a surface that has curves for edges. These curves, in turn, a curved surface, Curves come in a variety of flavors, as do the patches. Curves can be called by a number of different names, cir degree. of the curve refers to the mathematical formula used t0 it.The higher the degree of the curve, the more calculation d to create it. The easiest way to remember the order of the that the degree of the curve is one less than the number ofdigital character animation 2 points required to define the curve. Therefore, a first-degree curve needs two points to define it, a second-degree curve needs three points, and so on. Some packages allow different degrees of curves to be combined to make patches—for example with those that are linear along the U direction, and B-spline curves along the V direction, A linear curve looks lke a series of lines connecting the con- A cardinal isa curve that passes through the control points. trol points. The curves defining the surface are equivalent to Each point also has a tangent control the edges in a polygonal surface. B-spline curve rarely passes through its control points, and Bezier curve is similar to those used in popular drawing having the control points far away can make manipulating a programs, such as Adobe Iustator. The curve passes surface a bit confusing, through each control point, and each point has two tangen- tial controls for adjusting the weight of the curve on either side of the vertex NURBS Surfaces NURBS are really just an extension of the B-spline patch described previously. The extension—and the big difference—with a NURBS sutface is the term non-uniform. This means that each vertex can also be weighted to affect the curvature of the surface more precisely. Weighting also enables @ simpler surface, because fewer points ate needed to define the same surface In many cases, NURBS can be manipulated in much the same way as other types of patches. A number of tools have also been developed specifically for NURBS surfaces. As you will see later in this chapter, these include the capability of creating curves on a surface, plus the capability of trimming surfaces and blending between these curveschapter 2: modeling basics jodeling Strategies fith a number of different surface types, the choice of surface depends on ny number of factors. Whichever surface type you choose, there are a few. gies you can use. p It Light though the type of surfaces you use is decided by a number of factors. the wg model ideally should be able to be manipulated as fast as you can jimate it.The last thing you want to be confronted with when animating is a few seconds (even a fraction of a second) for the screen to up-date time you pose the character. The most important feedback an animator get is the ability to scrub the animation in real-time as iv’s keyfiamed means you need to keep your characters light. The term “keep it ight” to make the most with each and every vertex, spline, edge, and patch our model. If you model your character with exactly the right detail, ic |remain light. This will make animation easier simply because thete will es vertices to calculate when moving the surface of the character peeding up interaction and making the animator’ life eas g exactly the right amount of detail also enables smoother deform- n of the surface, because it significantly reduces the possibility of rearing breaking. you start modeling, you need to familiarize yourself with the tools. few sections go over the basic tools found in the majority of sof packages. Many packages may have more tools; some will have less. By zing yourself with the basics, however, you should be able to get up ed on any software fairly quickly. problem when discussing modeling tools is that, many times, there is fusion over terms. Three different software packages can call the same | by three different names. In these cases, I provide a glossary table to ference the term defined for this book with the terms used in thedigital character animation 2 Basics of Polygonal Modeling Polygonal modeling can create excelle are still strictly polygon-based. This is by no means a bad thing, because a good polygonal modeler enables you to sculpt a surface almost as if it swete made ont characters, and many packages F clay Polygonal modeling also affords a great degree of topological freedom not found in many patch-based modelers, ‘The fundamentals of polygonal modeling are fairly simpfe. Polygonal models consist of chree basic elements: the vertex, the edge, and the polygon. There are a number of ways you can manipulate these ‘elements. These manipulations are listed in Table 2.1 Vertices Ve objects, Tivo vertices can be connected to create edges. Some modelers enable additional operations, suck as the capability of extruding and beveling a vertex, both of which add extra edges and polygons es represent a single point in space and are one-dimensional Edges Edges are defined by nwo vertices. When viewing an object in wire fiame mode, you ate seeing the edges. In effect, edges are lines and are ‘two-dimensional objects. When three or more edges are connected, they create a polygon. There are a number of operitions that can be performed on az edge, These include extruding and beveling, which again add extra edges and polygons. A collapse operation eliminates the edge, reducing it to a single vertex Polygons Polygons are defined by three or more edges. In effect, polygons are planes and are three. der, polygons are required. Mase packages enable a polygon to have ‘more than three sides, but these are generally tesselated to triangles when rendering time rolls around, mensional objects, In order for an object co ren= There are a number of operations that can be performed on a polygon. Collapsing a polygon removes the polygon, collapsing i to a singh vertex. Extruding, insetting, and beveling are all variations of the same operation, creating one extra polygon per edge of the original polygon. The difference is in how the original polygon is handled. In an extrude, it is moved outward or inward, typically along the normal (that is, at a righe angle Co the fice of the polygon}. In an inset, the original polygon is scaled, typically inward, along the surface of the original polygon. ‘The bevel is a combination of the two, both moving and sca original polyg ig thechapter 2: modeling basics jations Performed on Vertices Edge—two connected vertices form an ‘edge, which is aso a line. Polygon—Three oF mace connected, edges make a polygon, which is a plane. fe—Pulls the vertex outward and polygons. ions Performed on Edges Bevel—Expands the vertex to a polygon along the edges. extruded edge creates a lyon along the edges normal Bevel—Beveling an edge creates a new face,Basics of Polygonal Modeling Operations Performed on Edges Operations Performed on Polygons digital ee Colapse—Deletes he edge, creating a single vertex Extrude—Pulls the face in or out along the normal, adding four new polygons = Bevel—An extrude with an inset. Four new polygons are added. character animat Cuveonnect—Cuts the existing edges in half, and then connects with an edge. Inset—Shirinks the polygon along its plane, adding four new polygons. Collapse—Deletes the polygon, creat- ing a single vertex.chapter 2: modeling basics basic operations are known, modeling with polygons is ly seaightforward and simple. Typically, a basic shape is created, box, cylinder, or sphere. This basic object is then modified ‘a number of techniques to create a more complex shape for odeling polygonal surfaces, or any surface for that matter, it’s est to start with a small amount of surface detail and work your p to higher amounts of detail. Most polygonal modelers start with Primitive, such as a box or a sphere, and add detail, working and for modeling is to make the most with the least amount nil People who model for interactive games have it particularly because their models have to be incredibly light. If you want a on in simple, but effective, modeling, sit with a game modeler for a I. Loops the more important things co master when learning to model jons is the concept of the edge loop. An edge loop is a ring that define an area of the character. In the face, edge Loops are Alefine the mouth, brows, and eye sockets. In a body, they deter he outline of the torso, arms, and legs. to mentally picture edge loops isto think of a map that has jours that represent the elevation of the surface. The contours are dkage and the contours traced with edge loops, the map takes shape. a edge loop is created by cutting two existing edges in half ecting them to create a new edge. The way these edges are cut termines a great deal about how the resulting madel behaves. When edges of a rectangular polygon are cut and connected, the result o four-sided, or quadratic, polygons. When connecting adjacent you get a triangular and a five-sided polygon, which can prove 1 ptoblem when the character deforms. It is always best to create rates, because they (end to deform more predictably than a surface; so when sketching out an edge loop, try to avoid ecting adjacent corners.Ff digital character animation 2 ‘An edge loop, highlighted in red, is simply a continuous Even the contours of a model as compiex as 2 face have loop of edges that define the contours of the model recognizable contours defined by edge loops. The contours of the mouth and eyes, for example, can be defined by concentric loops. When opposite edges of a rectangular polygon are When connecting adjacent edges, you get a triangular and cut and connected, the result is two four-sided, or 2 five-sided polygon, which can prove problematic when quadatic, polygons. the character deforms. Subdivision Surfaces Not everyone models for interactive games, and most people who model for film or video Want their models looking as smooth and natural as, possible. At first, this may seem to be a difficult problem, because the extra detail can prove to be difficult to manage. Fortunately, with the right tools, a properly constructed low-resolution model can become a high-resolution model almost automatically. This automatic tech- nique, subdivision surfaces, is called a variety of names by various software vendorschapter 2: modeling basics gons Become Patches ‘you subdivide a polygonal surface, che software adds detail. It first to figure out where to place all the added detail. These extra poly placed according to any number of formulas. The theory of ¢ took is simple:The three or four points that define the edge of are just that—points in space. By interpolating these points of patch modeling. implest way of calculating this extra detail is by cutting some literally. A woodcarver sculpting a block of wood starts by ‘off the corners of the block. If you can find a way to mathe- whittle off the corners of your virtual character, you would smoother surface. This technique is called corner clipping. er clipping takes the midpoints, or centers, of your object’s pot and connects them with intermediate polygons, turning one poly four, four into sixteen. and so on, Each time this is done, the ‘get whittled away, and the object gets smoother. The interesting about corner clipping is that, when the clipping is repeated to the surface becomes exactly the same as a B-spline patch, = —*--, Sec t \ t Repeating this step further refines the lines so that they approach the outline of a B-spline curve. iding two lines is simple. Take ple lines, find the midpoints, then cut the comers In Other Words. ‘Subdivision surfaces have been around for quite a while, and, as usual, many vendors have adopted their own terms for the technique, most commonly dubbed smoothing: Houdini—Smoothing LightWave—MetaNurbs and Metaform MAX—MeshSmooth, Tesselate, and third-party plug-ins, such as Infogratica’s SurfReyes Maya—Smoothing Nichimen—Smoothing Softimage—Rounding trueSpace—Nurbs,digital character animation 2 For surtaces, the theory is the same, but each face is subdi- Smoothing can be configured to vided. This smooths the surface. produce a mixture of tranguar and quadratic polygons Lets wait a second while that last statement sinks in. Mathematically, a smoothed polygonal surface is exactly the same as a B-spline patch. The next time some wacko tells you that patches are better than polygons, just remind him or her of this litte tidbit. Of course, infinity is a pretty big number, even for a computer, so for most purposes, we need to only approximate infinity. Usually. only ewo or three rounds of smoothing are required to get a low-res polygonal surface t0 look perfectly smooth on a film or video sereen. ‘The one problem with a character who's been subdivided is that the extra detail that subdivision creates also weighs down the character. This makes speedy interaction and deformation nearly impossible, This is resolved through a number of tricks. Most of these center around the simple task of creating the subdivision surface after the character is ani- mated and before it is rendered. Some object-oriented packages, such as 3D Studio MAX, LightWave, and Houdini enable you to subdivide the surface at any point in the n process. Typically, this is after she character is deformed and after it is animated. Most of these packages enable you to switch the subdivision modifier ‘on and off. When the modifier is turned off, the character is low- resolution and animates quickly. When subdivision is turned on, the character becomes high-resolution for rendering, Because the sub- division happens after deformation of the low-res model, the smoothing does not exhibit any tears or creases.modeling basics chapter 2: packages, such as Nichimen and LightWave, enable this task to en automatically at render-time. Essentially, this performs the same asin the previous example, her way to perform this task is to keep both a high-and a low tion model handy. The low-resolution model is animated. The gion curves are then copied to the high-resolution model and re {-The one problem wih this method is that, because they are sepa~ lalebs, the deformation on the low-res character may not exactly the high-res character, causing possible tears and other small errors. mating using subdivision surfaces, animate the low-res mode! for speedy on (right). For high-quality output, subdivide after deformation and before 9 left). leans Are Evil GS right, they're just plain evil. Many polygonal modeling pack- 4 Boolean feature, which enables you to add and subtract sur- easily. This might be nice for solid objects, but not for those ed to deform. {a Boolean operation is applied to an object, it creates tons of use- gons. If you observe the seam cteated from a Boolean operation, tice a huge amount of tiny faces used to bridge inaccuracies in ng of the models. These polygons add very litde to the model, but rly double or triple the weightdigital character animation 2 ‘A Boolean operation quickly enables you to add and subtract surfaces, such 2 this cylinder from the sphere Unfortunately, when the object is sub- Instead of using Booleans, select the divided, the Boolean surface breaks up faces and use them to model the detail because of unwanted artifacts needed. This results in a perfectly clean surface that subdvides well Even worse, if you're using Booleans to model a low-resolution charac- ter, the seam gives you all sorts of trouble. A subdivided Boolean opera- tion looks like a bad surgery scar. This might appear good if you're modeling Frankenstein, but even then you want otal control over how your model behaves when subdivided, so just avoid Booleans instead of Booleans, smart polygonal modelers rely on the standard toolkit for polygonal modeling. If you need to create a branch, bevel or extrude polygons outward. If you want to cut a hole in an object, bevel or extrude inward. This way, the edges are clean and smooth, swith no problems Basics of Patch Modeling Patch modeling is another excellent way to create characters. Patches are surfaces defined by curves, such as linear, cardinal, B-spline, or Bézier. Typically, these surfaces are four-sided, or a conglomeration of individual four-sided surfaces. As stated previously, curves and surfices are defined by their degree, which is the amount of control you have over the surface (ee Table Wi Lincar—A first-degree curve, this looks like a series of li necting the control poinits. The curves defining the surface are equivalent to the edges in a polygonal surface. 5 con cardinal—A second-degree curve, this curve passes through the control points, Each point has a tangent control as well. As you ean see, patch created with cardinal curves has four rows of vertices in each direction, for a total of sixteen points. The control points also extend beyond the edge of the surface. This is needed because cavtinals do not have true endpoint interpolation, and the extra points define the curve.chapter 2: modeling basics a ch can cover only a surface that is regular, or Pinch up the ends to get a sphere es. A patch can be wrapped into a cylinder. ‘manipulating a surfice a bit confusing. In this type of x, the extra points are called knots. As with a cardinal curve, plines do not interpolate beyond the endpoints, This me: points are needed beyond the edges of the patch to define pure, “Ako a third-degree curve, this curve is similar to those popular drawing programs, such as Adobe Illustrator. The ‘pases through each control point, and each point has two controls for adjusting the weight of the curve on either ‘ofthe vertex.’ These extra controls enable endpoint interpol 0 the points are at the edges of the patch, thing about Béziers is that the extra control points enable a ange of control over the curvature, from a smooth surface to a z Another nice thing is that this control is localized, so ‘one part of the model do not affect other parts Surfaces surface to create with a patch is to a regular surface. This is a that is basically a single patch. Single patch models can by modifying primitives, such as a sphere for a head or a cylin~ ody with regular surfaces is that they cexture quite easily, The lines ofthe surface will map directly to the X and Y coordi ofan image. If you can create an object using one regular surface snot have to apply mapping coordinates to the surfice, as you th an irregular polygonal surfacdigital character animation z ‘When modeling with patches, think of a regular surface as a giant piece of wrapping paper. Anything you can wrap with a single sheet of paper, you can create asa regular surface. To create a cylinder, simply roll the paper in a circle, To create a sphere, pinch up the ends of the paper. Creating Patches from Curves ‘When creating a patch-based surface, one way to easily define the surface is to sketch out the shape with curves. After the basics of the shape are defined, tools are used to turn those curves into surfaces Gee Table 2.2 ME Extrude—Extrude works by sweeping a profile curve along a straight line or a path. This creates a shape that is essentially a cylinder. curve and BE Lathe/Revolve—A lathe or a revolve takes an our sweeps it around an axis to create a surfice. The resulting surface may be either a cylinder o be closed at one or more poles, like a sphere. WE Lofting—Lofting is similar to extrude but enables you to have suultiple outlines along the path, This can give you much better control of the shape. IE Skinning—Ako known as a U-loft or a V-lofi, skinning is closely related to lofting and enables multiple ate a surface fs to be used to gener- HE Ruled—Ruled uses two curves to define opposite edges of the surface, It is similar to skinning, bur the surface is built along the curve rather than perpendicular to it BE Boundary surfices—Boundary surfaces take three or f and use them to define the edges of a patch. HH BiRail—BiRail constructs a surface by sweeping one or two pro- file curves along two rail curves. If one profile is selected, the tool works much like an extrude or loft. If two curves are selected, the tool works more like a boundary surface.chapter 2 modeling basics ‘of Patch and NURBS Modeling Types of Surfaces a A first-degree curve, this looks {setes of lines connecting the points. jhe curve passes through each point, and each point has two controls for adjusting the Fo the curve on either side of Patch Operations Cardinal—A second-degree curve, this ‘curve passes through the control points, The patch has vertices beyond the edges. NURBS—A nonrational form of the B= spline, this curve has weights to adjust the tangency. B-spline—A third-degree curve, this type of curve rarely passes through its contral points. The patch has vertices beyond the edges. works by sweeping ‘cure along a straight line or A Latherrevolve—A lathe or revolve takes an outline curve and sweeps it around, {an axis to create a surface(ce al character animation 2 Basics of Patch and NURBS Modeling Basic Patch Operations Advanced Patch Operations Ruled—Uses two curves to define ‘opposite edges of the surface, NURB Curve Operations —- ad Skin—Also known a5 4 U-loftor a V- —_Loft—Loftingis similar to extrude, but loft, skinning is closely related to loft- enables you to have multiple outlines ing and enables multiple curves tobe along the path. used to generate a surface Boundary surfaces—These take three _BiRail—Constructs a surface by sweep: or four curves and use them to define ing one or two profile curves along the edges of e patch. two rail cues. Curve on surface—A curve on ‘Trims—A trim can be used to cut out surface can be Created by projecting the area inside a curve on surface (mid- it (ight) or drawing it directly on the dle) or autsie a curve on surface (ltt surface (left),chapter 2: modeling basics Surface Operations To connect a curve on surface to Blend—A blend! creates a similar sur- surface, use the curve as the face that automatically changes shape ‘lloft or skinning operation. to maintain tangency. filet isa blend that maintains Charnfer—A chamfer rounds off the semicircular outline. corners of a surface. hing Patches Together are times, however, when a single patch or regular surface gener- [fiom curves will not cut it. The problem comes when you get 10 a jing area, which is defined as something that can't be formed from ic cylinder, sphere, or torus. The human torso is a good example of ct that has branches—the arms and legs create a topology that’s ible to completely cover with a single rectangular patch. If we use {wrapping analogy, we would need more than one piece of and some tape to completely wrap a human body. If you don't this, get a hunk of wrapping paper and a Barbie dotl or a GI Joe try it yourself. ‘Way around this isto stitch the character together a patch at a time. ‘method can be painstaking, but it gives you the most control over mnposition of a surface. The eheory behind this method is very forward. Modeling stares with a single, four-sided patch created ing a single patch or generated from outline curves. A second is created, and its edge aligned with the edge of the first patch. A Sng function is then used to weld the vertices and edges exactlyTo join two patches, start with a single patch—in this case a Cardinal patch In this case, the front and back of the t0rs0, 25 well 35 an arm, are snapped together to create 3 seamless surface. al character animation 2 Bring in a second patch and roughly _Use a snapping function to snap each align three rows of vertices row together. With three rows perfectly aligned, the patches will act as one Ic must be said that the edges themselves aren't technically welded together; they just occupy the same point in space, Because the vertices are in identical positions, however, they deform identically and appear be welded together. Some packages, such as 3D Studio MAX, streamline the process by enabling you to attach patch edges automatically: This saves you the tedium of snapping the patches together, a vertex at a time, hough took: like this are more efficient, the underlying theory is exactly the same. Other packages, such as Maya, have a number of tools to help maintain tangency between the surfaces, enabling seamless connections ‘The best way to visualize a character made in this manner is to picture the patches as swatches of cloth. By “stitching” together the patches a vertex ata time, the patches a scamed together into a smooth and convineing character. You might want to study garment patterns as ref erence when using this techniquechapter a: Sided Patches packages, patches are four-sided. This makes it easier to make perfectly tangent. it also enables regular surfaces, which makes i much easier. Those modelers that support only four-sided however, may have problems with objects that have branches, ss the arms and legs on a body. se, any modeling package that can create a sphere from patches aably make a three-sided patch, becamse the top row of a sphere of triangles—or more specifically, four-sided patches that have collapsed to a point. d surface is actually created as a bit of geometric sleight of ‘surface created is geometrically a four-sided surface with the de scaled to zer0, making it look like a chree-sided surface. sin mind, a four-sided patch can be made into a three sided imply by snapping the points on one edge together so that they A three-sided patch sa foursided tch with one side colagsed t0 the same point in space. This exentialy produces the same note pot ny as the pole of a sphere, but in single-patch form. This works Tower-order curves, such as cardinals, but also works with other types, including NURBS. sare, however, a few packages that have support for three-sided es out of the box, most notably in Discreet and Hash’s products h splines—Hash curves are a hybrid type of curve similar co linals but with the capability of interpolating endpoints. This ns that Hash splines can define a surface wich only ehree or points. Hash ako supports triangular patches. This enables you sseate branching areas and nonregular topology, much like in a modeler, This keeps the character light while retaining advantages of patches. patches—Another unique type of patch are chase found 3D Studio MAX. These are closest to a Bézier-based patch, 3 each. point has extra controls to manipulate the tangency he surface, and the surface passes through the vertices. Like 3D Studio MAX also has the capability of creating lar patches. modeling basics iadigital character animation 2 Modeling with NURBS Tools that manipulate NURBS are very much like other types of spline modeling tools, Curves can be lathed or revolved to create surfaces such as spheres. NURBS curves can also be extruded or skinned to create patch- or tubular-shaped surfaces, Most NURBS modelers also support tools such as lofts, boundaries, and birails. [As has been stated before, one of the most difficult problems with patch surfaces, including NURBS surfaces, occurs in the branching areas. As with any other type of patch, NURBS surfaces can be seamed together by snapping the points together or by using special tools to help seam together the patches. These are shown in Table 2.2. Curves on Surface Another popular way to create branches with NURBS is by creating. a curve on surface. This is done by locking a curve to the NURBS surface and then using these curves as the root attach points for the branches. There are two basic ways to get a curve on a NURBS surface: WH Draw the curve—The first is to draw the curve on the surface, In some of the more sophisticated packages, curves can be drawn in real-time directly on a NURBS surface. When dh face, a curve on surface will say stuck to the surface no matter how the surface changes shape. wn on the sur WW Project the curve—The second way is to project an existing curve on a surface. Projected curves are simply curves that are projected onto a NURBS surface, much like a two-dimensional picture in a slide projector. Because the shape of the curve as it hits the surface is dependent on the angle of projection, you need to take this angle into consideration. Typically, the curve is projected at a right angle to the surface normals. When projected, a curve is locked to the surface; this means that if the surface changes shape, the curve will follow, exactly like a curve drawn on the surface, ‘One application for projected curves, and curves on surface in general, is to use them as the attach point for creating a second surface that remains seamlessly attached to the first. A projected curve, for example, can be used as the start of an extrude to create an arm or a tentacle.chapter 2: modeling basics Types of Curves on Surface m curve is essentially a curve on surface. The trim curve gets its som che fact that it is used to trim the surface away using the “much like a cookie cutter. This enables you to punch holes in a surface, and is the closest you'll ever get to. a NURBS Boolean Trims are actually just a rendering trick. The entire surface and the portion trimmed away is not rendered. m isa good base for creating branching surfaces. The trim avoids oblems with surfaces protruding through each other, because the ja protrusions have been snipped away by the trim operation, ods for seamlesly creating additional branched surfaces. The theory tools follows the same logie: The curve on surface is ased as a ice. The trick lies in making this seam trans- 10d of making the seam transparent to the eye is to loft the attached to the surface along with other, unattached curves to the skin. The base of the skin is locked to the originating curve, with this the skin, Remeber, that with NURBS curves, it takes four determine a curve. ff three curves are locked to the surface, should be perfectly tangent by the end of the third curve, he seam. Still, the texture might nor line up exactly, so it is 14 good idea to feather the texture so it is transparent at the seam. nice smooth transition more elegant solution of making the seam transparent is to use blend. A blend differs from a skin, because it automatically eS enough surface to fill che area between two curves attached to ¢ while maintaining perfect angency, Because the blend is auto- nd self-adjusting, itis a great choice for highly flexible areas, such ders. Blends are also handy in facial areas. A nose or the areaIn Other Words. Metaball implementations cross a large range of packages, and the names may vary from package 10 package. Some packages allow meta: balls modeling only, while others per- mit animation as well 3D Studio MAX—Has no native metaballs, but several third party implementations are available. Clay Studio Pro—Enables anima- tion, muiltiple types of primitives, sticky mapping and spline driven muscles, MetaReyes—Enables animation, sticky mapping, and the capability to create spline driven muscles, Blob Modeler—A freeware modeler with no animation Softimage—Meta Clay is Softimage’s implementation and enables spherical and ellipsoidal objects as well as animation. LightWave—Metaballs in LightWave enable spheres and are modeling tools only Houdini—Metaballs are spherical and elipsoidal and enable animation. digital character animation 2 around the eyes can be blended into the surface of a face quite easly The adaptive nature of the blend defily sidesteps many of the detail and tangency problems inherent in a direct skin method One problem with blends concerns performance. A blend surface is very complex and can bog down a system during animation. In most pack- ages, a blend shows up as a separate object, which means you can select and hide the blend objects while animating and then display them just before rendering, Fillets Other ways to create automatic surfaces is with tools that saw their gen= esis with CAD-type operations. The fillet is really just a variant of the blend that is constrained to a semicircular outline. Ic is great for creating rounded joints between attached surfaces. Chamfer The chamfer is the opposite of fillet; it rounds the outside edge of a surface, rather than the intersection of two surfaces. The chamfer is perfect for rounding off the corners on a box, for instance. Deformation by Surface Deformation by surface is another tool that has been used by vendors such s. Instead of locking down a curve, deformation by surface locks an entire surface to another. This can be used to add local detail or to create a base to connect a branched surface, such as an arm or a leg. In Softimage, this ccomplished through the use of a Zip patch, which skins between a deformed patch and a free surface. One problem with deformation by surface is that the patches are not mathematically locked, so there is Softimage to create seamless attach poi always a slight error in the fitting. Still, judicious use of textures can hide the se Metaballs Although these types of surfaces certainly are more than enough to create convincing characters, there are a number of other ways to create 23D surface. The most common of these other methods used for char- acters is metaballs. In their simplest form, metaballs are simply spheres or other primitive shapes that tend to blob together. Metaballs are a technology that 3D animators use to easily create round, blobby objects. As these objects move closer together, they seem to melt together.chapter 2: ng basics In addition to spheres, many of the better metaballimple- rmentations enable a variety of primitives, including cubes and spline-deven muscle shapes. in metaballs originally derived their name from simple which are the building blocks of metaball objects, though ‘ofthe better implementations enable 2 variety of primitives, By each sphere a weight and a sphere of influence, the metaballs fuses many spheres into a single blob. How much the spheres depends on the weights, influences, and distances between, simple example. Consider two spheres of equal size. The sphere uence surrounds each of the spheres like a shell. Any other sphere ‘within this range tries to fuse with the spheres. The weight exactly how much fusion actually occurs. If we have two ‘equal size, they'll fuse at the intersection of their spheres of The amount that they fuse depends on their weights. As ts and the spheres of influence change, you get different effects. sphere of influence makes an object softer and more willing se with another. On the other hand, if one sphere is given more ght than another, it appears more stable and solid. Il metaballs implementations are restricted to spheres. Many pacl nonspherical objects. Some go even further and actually ¢ you to create virtual muscles to give you some stunninglydigital character animation 2 The Trouble with Metaballs ‘The big problem with metaballs is that they create very messy surfaces that are difficult to anima marching cubes algorithm to generate the surface. This means that a three-dimensional grid is projected over the model, and the software uses this to calculate where the surface exists, The artifact of this method is a model with a fine unidirectional mesh that completely covers the surface. This type of surface proves very difficult to animate in its raw form. because most metaball modelers use the When a metaball-created model is turned into a mesh, the To solve this problem, the metaball primitives themselves ae mesh may look smooth, but the underlying wireframe is animated, and the mesh is resohied at rencer-time. Usually a mess and impossible to animate. Animating Metaballs There are two ways to circumvent the problem of metaball-derived meshes so you can animate the character. The first is to use the metaball generated model as a template. The model is loaded into a more tradi- tional modeling program, and the character is rebuilt using polygons, patches, or NURBS. ‘The second way to use metaballs in an animation environment is to ani- mate the metaball primitives themselves and resolve the surface only at render-time. This gives you a great deal of flexibility and enables very sofi and stretchy characters to be ereated. One good example is the motion picture Flubber, which used metaball technology for the jellylike characters that starred in the moviechapter 2: modeling basics ‘with this form of animation comes into play when a texture The more sophisticated packages enable “sticky mapping,” the of applying the texture co the metaball primitives themselves. ing then flows through to the surface and animates with the A less-sophisticated package may allow a texture to be applied the resulting surface, rather than the primitives. When animated, ‘will not adhere to such a surface. jusion ‘ean see, there are many different types of surfaces that can be build a character. [tis always a good idea to at leas try all these types to determine what works best for you. Having experience different geometry types will also make you more knowledge- valuable in the job market. few chapters discuss modeling characters. The characters are a number of different geometry types. If your software allows the time to work through all the exercises so that you under- characters are built using a number of different techniques.uolwewiuy 410f saipog BuljapowCHAPTER 3 Modeling Bodies for Animation hen you understand the basics of character design and the basics of modeling, you can begin building a character thar not only appealing, but animates well. The construction of your character is determined by its design and by your software's capabilities. Be sure to take advantage of your software's strengths when designing a character, Remember that your characters not only need t9 look good, they also need to animate well. This means keeping your characters light so they can be easily manipulated. A character that moves as fast as you can think makes animation a breeze. There is nothing more frustrating to an animator than waiting a few seconds for a character to update ‘every time ics moved, Fast interaction is accomplished by planning, your detail so that there is no waste. Less is more when it comes to building charactersdigital character animation 2 The Structure of the Body Whether you're designing a photorealistic human of a flour sack, you ned to understand the basics of anatomy: Even a simple Dour sack can appear to have feet and shoulders, so whether you're animating a flour sack to move like a human, or you're animating a CG human, knowl- edge of human anatomy is essential. This chapter gives a brief overview of the basics of anatomy. The sub- Jeet, however, is incredibly deep, and entire books have been written on the it, A good book on anatomy is always a great reference for modelers and animacors alike. Another good place to get a strong foundation in anatomy is through 2 life drawing class, Visualizing the human form helps train the eye as well as aid the understanding of the human form. The Skeleton If you want to design humanlike characters that animate well, you need to understand the human skeleton—how it is put together and also how. ic moves. A human skeleton has over 200 separate bones. Many of these, however, are small bones, such as those found in the wrists, ankles, and inner car. As an animator, you need to concern yourself with the major systems of the body and how: chose systems affect che shape and posture of the body as they move:chapter 3: modeling bodies for animation Three major bones comprise the arm; the humerus, the radius, and the ulna. Proportionally, the arms hang down so the wrists are even with the hips. ‘The humerus is the upper arm. Itis connected to the shoulder by a ball-and-socket joint. This gives the upper arm a wide range of motion The forearm consists of two bones: the radius and the ulna. They are connected to the humerus at the ‘elbow, which is a hinged joint, limating the motion to one axis. The radius and ulna twist around ‘one another to rotate the hand at the wrist. ‘Two sets of bones comprise the shoulders, Along the front of the body is the clavicle, commonly known as the collarbone. Along the back is the scapula, a broad bone commonly known as the shoulder blade. Shoulders connect the arms to the spine, but also have a range of motion that enables them to “shrug” and to move torward and back slightly.digital character animation 2 continued The Vertebrae The vertebrae, better known as the spine, are the support system for the upper body. All the weight of the upper body is transmitted through the spine to the hips. ‘Alone, each vertebra bends only slightly, but as 2 system, the verte- brae enable the spine to twist and to bend at the waist. Several bones comprise the hips, most important of which is the pelvis. The hips anchor the spine and transfer the weight of the upper body to the legs. As such, they are the center of weight distri bution for the entire body. The hips 4 themselves don’t move; the body moves in relation to the hipschapter 3: modeling bodies for animation Like the arms, three bones com- prise the legs. These are the femur, also known as the thigh bone, and the shin, which is two bones: the tibia and the fibula The femur attaches to the pelvis with a ball-and-socket joint. This, gives the thigh a wide range of motion, though not quite as much as the shoulder gives the upper arm: The shin is made up of the tibia and the fibula. They are con- nected to the femur at the knee, which is a hinged joint. The tibia and fibula also twist around one another to rotate the foot at the ankle.digital character animation 2 Muscles Bones define the structure of the body, and muscles, to a large degree, define the shape of the body. There are hundreds of muscles in the body, and each performs a specific function. Muse contracting, or pulling, the bones and skin. In the body, there are a few major muscles that the modeler and animator need to understand: les do their work by Deltoids—These are the muscles that run over the top of the shoulders and give the shoulders their definition, The deltoids raise the arm above the head. Peetoralis/Iatissimus—The pectoralis major moves the arm down and forward. The latissimus moves the arm down and back. Biceps/triceps—These are located around the humerus, or upper arm. The biceps pull the forearm up, and the triceps move it back. Supinator—The supinator rotates the radius and ulna around each, other, twisting the forearm, Flexors/extensors—These are a few of the many muscles in the forearm and are the muscles that open and close the hand, Glateus maximus—The rear end, which enables the leg 10 move backward. Sartorius/quadraceps—Alo known as the biceps of the leg, the sartorius enables the shin to bend back at the knee. It is comple- mented by the quadraceps, which pulls the shin forward, Gastroet and it is connected to the foot via the Achilles tendon. ius—This is the muscle along the back of the shin, Construction Methods Most 3D packages offer a wide degree of choices when building a char acter. The geometry can be either polygonal, patch, or some other sure face type, stich as metaballs oF membranes. The character can be baileas one solid entity or broken up into its component parts Breaking up a character in parts is also known as segmenting a charac ter. Depending on your character's design, it may be easier to segment the character rather than build it as one single skin. A good example of a segmented character might be an artist’: mannequin or a suit of armor. Single-skinned characters, such as a seamless human, are typical more organic and lifelike.chapter 3: modeling bodies for animation gmenting Bodies jut a doubt, the easiest way to create a body is to create it as a of segments. A true segmented character is constructed a joint at each joint as an individual segment. To picture a true segmented macter, think of a classic artist's mannequin, a character constructed individual blocks of wood and fitted together with metal pins. toys, such as Hasbro’ venerable GI Joe, are also good examples gmented characters ling a segmented character is quite easy, because each joint is con- ted separately. Modeling the joints individually makes them very to construct. Usually, the entire character is made of one type of ay; but any combination of geometry types can be allows it, Because the individual joints of a fully se uacter don’t change shape as the character is manipulated, this type of uacter is usually very easy to animate, Manipulation of the character t because the computer does not have to calculate complex shape on-the-fly “| "arists mannequin is a good example of a segmented A robot is another good candidate for e segmented characterdigital character animation 2 aarp Go € This space rangers body looks seamless. Pull him apart and you'll see that he is made of many Clothing is a good device to hide seams. The seams for the hands con this character are cleverly hidden under the sleeves. different segments that have been cleverly hidden. The downfall to segmentation is that, unlike a single-skin character, segmented character will almost always have telltale seams. This may not be a problem if your character is designed to hay mannequin of a mechanical robot, for example c seams—the artist’ Texture maps are very helpful in making a segmented character look real. A good, close-fitting seam covered with a complex texture will help hide, or at least minimize, the appearance of seams. Many popular inter- active games use these methods to keep their characters looking good. Another solution is to segment just the major chunks of a character. Many times, the head or the hands are created as separate objects that are connected to a seamless torso, The hands can be connected under- neath a shirt sleeve, the feet connected underneath a pants leg, and the neck beneath a collar. Hiding the seam in such a manner still allows a realistic-looking character, but makes it easier to model and set up.chapter 3: mode “Grates can be segmented in mary different ways “Hee isa simple female character with obvious seams This character also has segments, but the seams are 9t.s apparent g bodies for animation ‘She is segmented in a number of places, but the hands and, aims are seamless, as well as the feet and legs + Only the hands, head, and tail are separate. As you can see, there are many combinations when segmenting a character.« Model the legs first. Start with a sphere. 4. Add detail near the hip area, and other places as eded, so that the legs can be sculpted digital character animation z Modeling a Partially Segmented Body aracters. In this exercise, you create a simple character that has a seam at the waist, The character is Patches are another great way to creat constructed of two simple spheres. This particular character was created using NURBS, but because there are no blended surfaces, it can be made: i just about any package. . Rotate the sphere so that the 3, Select the vertices represents as lie along the horizontal ing the ends of the hot dog axis. Scale the sphere along and rotate them down to the horizontal axis to create c a horseshoe shape. a hot dog shape. 5. Seulpt the legs and hips Put the legs aside for now to get the desired shape. while you build the upper body. Again, start with a sphere.chapter 3: J Rotate the sphere so that it 8. Add detail near the armpits 9. Resculpe the upper body to lies along a horizontal axis. and along the chest so the give it the desired shape Select the vertices surround- arms and the upper body ing the poles and move can be sculpted. them out to create arms. 10. Finally, join the ewo objects to create the body. This cre- ates a seam where the two surfaces overlap. To hide this, a simple belt can be added around the waist to hide the seam. The hands, feet, and head can also be attached as segments. BS crcmmawacs A -copy of this model is on the CD. Itis named SegGuy. modeling bodies for animation aaatal character animation 2 Seamless Characters Although segmenting the parts of a body is one good way to build char: acters, many times you need the body to be one continuous skin. A m realistic or organic character, stich as a realistic iguana, could not hide a betw ead and body. In this case, you really need it to be al one piece, unless it's wearing a necktie. Many superhero characters wear tight-fitting garments, and can also be considered single-skin characters. the h A single-skin character is certainly much trickier to construct than a s mented body. Still, the construction of such a character is certainly wit the capabilities of most software. Both polygons and patches, as well as many other geometry types, can be used to create a seamless body. A seamless body must be deformed in order for it to animate. Typically, the body is deformed using a skeletal deformation system. This, how. can be problematic if the body is not modeled properly. Usually, the problems arise in the areas with the widest range of motion. A quick at our skeleton tells us that these spots are where the ball-and-socket Joints are located-specifically, the shoulders and the hips. If the mesh is improperly constructed, you get an unnatural movement of the skin may cause it to break or crease, Proper modeling first means understanding anatomy and the way a is put together. If your character’ shoulder is anatomically incorrect, for example, it simply will not move properly. The shoulder is supposed to above the armpit, for example. If the shoulder is modeled too far to the outside or inside of the armpit, not only will the character look odd, the shoulder will not move properly when animated. It is always a good idea to place your detail only where it is needed and try to run that det along the major lines of force exerted by the muscles, Modeling a Seamless Body Using Polygons Polygons can make excellent single-skin bodies. Because polygons give you wide topological freedom, the entire body can be modeled as a surface, including the hands and the head. As with other polygonal mo cling methods, itis best to model the hands at a low resolution and th add the required detail through subdividing the surface. The body presented in the exercise creates a simple male figure without the head and hands, which are discussed in the next chapters. The moi ing is done in two phases. First, the basic body is blocked out. Second, derail is added to the body where it is sculpted into the desired shape. When built, che model can be modified to create a body of just about any size and shape.1. Stare with a simple box that has three vertical subdivi- sions. The box should be of ‘equal width and depth, with the height approximately three times the width Select these vertices on the citer side of the body. Move these up vertically so that the shoulders are approximately ‘even with the neck. chapter 3: modeling bodies for animation Subdivide the box once. This should bevel the top, bottor and sides, as well as add detail within the box. Select the vertices surround~ ing the waist and seale them in. (If you are making a heeavyset character, you may want to scale these our instead.) . Select the vertices defining the top two rows of edges. Scale these down equally to make a neck, Select the faces at the base of the shoulders and extrude them out to create an arm stub.7. Conti to create enough detail to define the rest of the arm. The faces can be scaled up or down to help define the outline of the arm. If you extruding the faces ‘want, you can even extrude the hands from the ends of the arms, Construction of hands is covered in the next chapter, digital pe 8 10, Now move on to the legs. Select the outer faces on the underside of the body and extrude these down to create a leg stub. Now that they are horizon- tal, reselect the faces at the bottom of the leg stubs. Extrude these straight down to create the legs. As with the arms, add just enough detail to create the shape character animation 9. Select the vertices definin the bottom face of the left leg stub, Scale these ve to zero so that the face is horizontal. Do the same the right leg. 11, Ifall you want is a basic body, this model ean be smoothed and used imm ately (of course, hands at head need to be added). fine the body, tinue with the next pl further ryextra derail needs to be Create an edge loop bisects the front of the and another that bisects body along the side, the shape of the chest ‘and shoulder area. Select the “vertices that define the top of ‘he shoulders. Move these up vertically to define the del- toids. Sculpt the chest area as well to define the pectoral pter 3: modeling bodies for animation Add three more edge loops. One bisects the arms, run= ning across the chest and along the back. The second two bisect each leg, running up and over the shoulders, Select the faces in the top center of the body. Extrude these to create a neck stub. If you want, you can attach the head separately or continue extruding to create extra geometry, which can be fash- ioned into a head later (this be discussed in Chapter ‘Modeling Heads for Animation”). Select the vertices in the leg created by these new edge loops a ward so that they give the leg scale them out- a more cylindrical outline. Continue Co refine the out- line of the legs. Reepeat this procedure for the arms,| digital character animation 2 6. ‘To make feet, extrude the Afier the low-res model is faces at the bottom: of the complete, it can be subdivide leg, and then scale them ed and textured to make the up to create the heel area high resolution final. If the Select the forward faces and body is being deformed by 4 extrude them forward to skeletal deformation system, create the rest of the foot the subdivision should proba ee Hands can be made ina bly take place after the defor oe similar fashion, and detailed mation takes place. A copy of this model is on the CD. It is named PolyMan. construction is discussed in the next chapter. Modeling a Seamless NURBS Body Most of the better NURBS implementations enable you to create a blended surface. This exercise takes advantage of that feature to create a seamless female body. Of course, the same techniques ean be applied male, alien, or any other type of body you want to construct.hapter 3: The torso is created using ‘outline curves. First, create the profile of the front of the body: if you have sketch- es ora digitized model, use them as reference in creating, this outline. Select the front curve. Duplicate it and rotate the duplicate approximately 30 degrees to the right. ‘a second duplicate for the left side. Create modeling bo 2. Duplicate this curve and rotate it 90 degrees. Select the vertices and reshape the curve (o create the outline of the left side of the body. Again, reference may be helpful for this step. When completed, mirror this curve to create the right side of the body. 5. Repeat this procedure for the other three curves. You will now have a total of 12 curves. s for animation 3. Duplicate the original curve and rotate it 180 degrees. Reshape the curve to create the back of the body. You should now have four curves representing the front, back, right, and left sides of the character. 6. Loft or skin the curves to create a closed surface. Ifthe outline curves were accurate, the torso will be close to the desired shape.Exercise: continued 7. Insert detail as needed (red) and sculpt the torso to the final shape. 10. Move the arm close to the torso, Extract the curve at the shoulder end of the arm. Project this curve onto the torso, digital 8 character animation z Put the torso aside while the Sculpt the arm to get the desired shape. Select the inner side of the arm near the shoulder and expand it to create a fimnel shape. Mirror the object to get the opposite arms are constructed. Create a hook-shaped curve and revolve it into an arm shape. (lf you want to create an open-ended arm, such as a shirt sleeve, leave off the hook to get a cylinder open at both ends.) Trim out the surface within 12. Create a blend between the the projected curve. trim curve and the end of the arm, Adjust the blend parameters to get a seamless fit, Repeat this provedure for the opposite arm.modeling bodies for anima te a leg the same way 14, Connect the legs to the body 15. ‘The completed body is ready ‘ereated the arm. using a blend. This procedure for animation. Hands and ve an outline curve, is virtually identical to the feet can be created as sepa then sculpe the resulting arms. Curves ate extracted rate objects or blended to we to create the desired fiom the legs and projected the ends of the arms and shape. Mirror this leg to ‘onto the torso. The surface legs. Because the body is the opposite leg is then trimmed, and a blend open at che neck, a head can surface is ereated to fill be attached seamlessly, sepa~ the gap. rately, or through a blend. clusion is outlined in the exercises have been very simple representations of human bodies. OF course, one of the joys of animation is the capability ‘fr beyond reality. Any of the bodies can be reshaped and reworked into a that suits your needs. Think of these bodies as basic building blocks ‘are only the starting point for more witty, fantastic, or stylistic designs.uolewiuy 10f spuey BuljapowModeling Hands for Animation ext to faces, hands can be we of the more expressive parts of the human body. If you've ever watched anyone who gestures a lot, you know that hands can be a very important part of communication. In addition to helping with communication, hands also give umans the facility to manipulate and build things Animating hands can be particularly tricky. Like the person who's thrust ‘upon a stage with no preparation, many animators don't know what to do with their hands—or, at least, the hands of their characters The hand is a very complex structure, with dozens of bones, muscles, and tendons—all covered by pliable skin, Modeling the hand is a chal- Ienging task, and preparing it to be anirsated can prove frustrating as well. Like many complex tasks, however, modeling a hand can be broken down into simple steps that make the job easier.The palm is basically rectangular, and the fingers are basicaly cylindrical digital character animation 2 Examining the Hand‘s Structure Although hands are complex structures, they can be thought of as a cole lection of simple shapes. If you look at your own hand, you see that the lar box, with the fingers as cylinders attached along the top attached at about a 45-degree angle in the bottom corner of the box. Depending our character, you may need to ereate long, skinny hands, short, stubby hands, or something in between. Whatever the proportions, the basic shapes are essentially the same. palm is pretty much a rectan; ce. The thumb is a stubbier cylin ‘These basic shapes can simply be hierarchically linked, giving you a style ized, but serviceable, segmented hand. If you want « different look, you can expand and enhance these basic shapes with other, more sophisticat= ed modeling tools. You ean use almost any technology at your disposi Polygonal hands are probably the easiest (o model, Metaballs can alo make hands, and a good NURBS modeler with blending capabilities can create excellent hands as well, Another issue is how many fingers to use. realistic character will have four fingers (and one thumb) on each hand. For some unexplained reas gers, and alien eer fingers a character has, the easier son, many cartoon characters have only chree fi characters only one or two. it is to build, but the techniques are the same, less.chapter 4: modeling hands for animation can animate hands in a number of ways. If you build your hand wut ts, they can be joined in a hierarchy and manipulated directly. built from metaballs also can be joined in the same manner as nented hands and manipulated directly. Solid mesh of spline-based by a skeleton made of bones, which will be covered in a later ter You can also deform hands with shape animation tools, such iple target morphing uality. Each method has its advantages and dhantages, and which method you choose will depend on your char- requirements, as well as your software and its capabilities. ibility of the Hand and Fingers you animate your character’ hands, you should make sure that ‘hand is flexible enough to move realistically. How much the hand esis, to some degree, an aesthetic decision, Some projects demand ids that look absolutely real, right down to the bulge of the kmuckles id tendons. Other projects are less stringent and simply require that the gers move well decisions will affect how flexible your hand needs to be, as well as is built. A hand manipulated ising a skeleton or shape animation probably be more flexible than one made of hard segments, because in on the hand changes shape constantly as the hand flexes. It is a idea to know ahead of time how much detail the are direction of [project requires for your character's hanck, ve, the best way to see how a hand moves is to use your own as @example. Flex your hand and observe how it changes shape as it es: Here are a few pointers to help you understand the hand and aw it moves. Fingers 'yu look at your own hand, the motion of the fingers is pretty obvi- Compare your own hand ¢© the illustrations on the next page. last effect (the folding and creasing of the skin) is a tough one to te. Many projects don’t require this level of detail, but you can it with a skeleton and a mesh deformation utility chat enables to bulge and crimp the mesh according to the angle of the joint. ther way is to use a shape animation yystem hat actually morphs fingers to premodeled extreme shapes. There are abo other deforma- systems, such as spline of wire deformers, that can create the effectdigital character ani ation 2 They angle outa bit when the hand spreads. ‘Trey can also curl toward the palrn. When this happens, the pinky is usually the finger that curls fist, with the other fingers following in order ‘The skin creases and folds a5 the fingers bend (arrow) The Thumb The chumb’s motion is a bit more complex than the other fingers. Its three joints give it a much higher degree of motion than the three joints of a finger. Again, look at your own thumb and notice how it compares to the following illustrationschapter 4: modeling hands for animation joint pulls the entire thumb toward and away from the palm. S actually a third joint in the thumb, forming the base of the palm, hidden skin, This joint enables the thumb to move below the plane of the hand, st type of motion also causes the skin in the palm to crease, ost hands don’t need this level of detail. A skeletal deformation nbe used to manipulate the skin, and a multiple target mor- tem or other type of deformer is required for you to animate This flexibility enables the thumb to reach over and touch all the finger. This one bone is what oives humans their “opposable thumbs,” which separ rates them from the rest ofthe animal ingdom,1. Start with a simple finger joint This pa has two hemispheres at the ends sla joint digital character animation 2 Modeling Methods for Hands As with bodies, you can model hands out of segments, metaballs, poly= gons or patches, or any other type of surface, depending on your soft- ware, Most modeling methods can produce hands that look good and move realistically. In these exercises, construction of hands from a number of different methods will be discussed, Hopefully. one of these will work for you. Also, in each exercise, only one hand is modeled. It is simply a matter of mirroring your newly created hand to make its counterpart Modeling Segmented Hands Segmented hands are very easy to construct and, although they may not be seamless, they still can be quite expressive. Characters such as insects robots, and low-resolution characters for interactive use all are places where a segmented hand can be used. The characters in Pixar's A Bug’ Life are a terrific example of segmented hands. 2. Copy the finger joint to Align the fingers together. create nue more joints Because the original joint was made out of a hemi- sphere, you can align the Joints in wireframe so that the hemispheres line up. This will make the joint almost seamless when it animates,chapter 4: modeling hands for animation py the finger to create 5. Copy two joints of a finger 6. Model a palm, and then Tink fingers. to make a thumb. the hierarchy. to a seamless hand that needs to be deformed, a segmented sak easy to set up. Segmented hands also can be animated quite are great for practicing your animation chops. ‘course, is a very basic, bue useftl hand. The big poine here is that isa collection of objects that are linked irs a hierarchy, For ex creatures, the individual shapes can be molded to what- your character demands. ling Hands from Polygons s polygons offer a wide degree of topological latitude, they make t choice for modeling hands, As with other polygonal mod 10ds, itis best fo model the hands at a low resolution and then required detail through subdividing the surface presented in the exercise will block out the basic topology of sd hand. The techniques can easily be modified to produce any type of hand.Exercise #2: Building Hands from Polygons 1, Stare with a simple box that 2. Select the highlighted edges and move them toward the center of the hand. The spac ing between these edges and their neighbors will eventu- ally determine the spacing between the finger. has 8 rows of vertices in the X direction, 4 rows in the Y direction and 2 in the Z direction. 4. Select the face along the 5. Extrude the same face again front part of the box that represents the pinky and extrude it slightly to create a finger stub, The extrude should be the length of the first finger joint to make a narrow segment that defines the joint between the ewo joins. This will help the finger bend when it is deformed. Nest, extrude the face again to make the middle joint. Repeat this process once more for the last joint. digital character animation Do the same with the inner edges. Move them together so they reside in the center of the hand and are about the same distance apart as the other nwo sets of edges . Repeat this process for the other fingers.hapter 4: for the thumb, Select fice that resides at the comer of the palm, ly from the pinky ide this face slightly. The basic topology is now in place, Start sculpting the ‘hand to get the proportions correct. modeling hands for animation Extrude it again to make the 9, first thumb joint, then a sinall gap berween the joints, and finally dhe last joint. When sculpting the hand, be 12 sure to model the arch of the back of the hand, as well as the palm, To get the arch, select the vertices over the center of the hand and pull them up.You should also turn the hand over to sculpt the base of the palm in a similar manner. Select the faces representing. the thumb and rorate them along the vertical axis 45 degrees toward the palm. When the hand is the cor- rect shape, you can subdivide the surface to get a more detailed and smoother hand, If the hand is being deformed by a skeletal defor- mation system, however, this subdivision should take place after the deformation,digital character animation 2 When the hand is done, it ean be attached to the ends of your charac- ter’s arms. If you want a totally seamless character, the basic box used as the starting point for the hand can be extruded from the character’ arm. The hand can alo be attached by building faces from the end of the arm to the base of the hand, This capability is one of the big advan= tages of polygonal modeling. Of course, this hand is very basic, As you will see later, more detail can be added for an even more realistic look. And for a more stylized look, the hand can also be modified however you sce fit Modeling a Hand with NURBS NURBS can be used to model fairly realistic hands, though the meth- ods required are a little more complex than with polygonal modeling. Because NURBS surfaces are regular surfaces, the branching areas where the fingers attach represent the big problem. This makes NURBS hands a bit more difficult to manage at animation time. This exercise solves this problem using a NURBS blend to attach the fingers to the hand. The thumb and wrist can also be attached using a blend, but proper planning can eliminate a blend at these joints. The ‘thumb has the largest range of motion, so try t0 avoid a blend in this area because of the tendency for blends to break and crack. There are many ways to create a hand using NURBS; the choice will be dependent on the amount of detail required, the range of motion required, and the physical look of the character. The following sections describe wo basic methods that can be expanded upon as you need. Modeling a Hand with a Seamless Wrist Here is a simple four-fingered hand that has a seamless thumb. The hand is also arranged so that itis open at the wrist, enabling a seamles nv at the wrist. This red attachment to the for blends to four (one for each finger), which makes the hand easier to texture and deform, es the number ofchapter 4: MY Suart by making a finger. The finger is constructed by modeling a simple hook- shaped curve. |. After the finger is modeled, ‘copy it to make the other three fingers. Fingers are usually not the same size, so resculpt the new fingers accordingly. 5. modeling hands for animation . The curve is then revolved 360 degrees to make an open-ended sausage shape. Copy the original finger ‘once more. Rotate the copy 45 degrees and place it roughly where the thumb resides 3. This open sausage shape can then be sculpted a bit more to give it a more realistic outline. . We now need to make the surface of the hand. Select the vertices at the base of the thumb, Rotate them 45 degrees, and then scale them up to approximate the wrist opening,Exercise: continued 7. Now sculpt the rest of the hand fiom the base of the thumb, To get the shape right, you need to insert detail by adding rows of ver- tices into the hand. Add the detail gradually, so that you don't introduce unwanted creases into the hand. 10. Use a trim operation to remove the surface of the hand within the boundaries of the projected curves, digital character animation 2 8, When the hand is properly 9. Do the same for the rest of sculpted, the fingers need to the fingers. Make sure the be attached. Extract a curve projected curves do not from the base of the firse fin~ overlap each other. If this ger,and then scale it up by happens, reposition the fin= about 10 percent, Project this gers and try again, im curve onto the hand. Blend between the base of _—12, Because the base of the hand the fingers and the edges of 4s open, the hand can be the crims.The hand is now extended seamlessly from the complete, wrist (0 create ant arm. Conversely, an existing arm could have been resculpted to create the palm and thumb,chapter 4: modeling hands for animation ca jeling a NURBS Hand with a Blended W1 j next hand is very similar to the previous hand, but with a slight fendors.The alia dts is that the new sipaloey forces you tWadd a fith blend where the hand meets the wrist. 1, The fingers and thumb are 2. Close up the end of the 3. Move the vertices represent modeled the same as in the thumb by scaling the end ing the base of the thumb so previous exercise row of vertices (0 zer0. This that it rests immediately will make the thumb a below the pinky. Rotate closed sausage shape these vertices 45 degrees 4, Add detail into the thumb. 5. Select the vertices along the You should add at lease two front side of the thumb and rows of vertices for every pull them forward to make finger, plus a few more the front part of the hand.Exercise: continued 6. Select the vertices along the back side of the palm and move them back toward the wrist. The hand is now tak~ ing shape. Sculpe it to get the proper shape, adding detail if needed. Because the detail in this hand rans parallel to the fingers, it should be easy t0 create knuckles and tendons. digital character animation 2 7. When the hand is sculpted 8. Use a trim operation to to your satisfaction, attach. the fingers as in the previous exercise. First, project curves extracted from the fingers onto the hand 9, Blend between the base of 10. the fingers and the edges of the trims. The hand is pretty much complete, remove the surface of the hand within the boundaries of the projected curves, Because the detail runs par= allel to the fingers, the hand and the arm can't be pact of the same surface.To attach the hand to the arm, trim away the base of the hand and create a fifth blend su face in the same manner as you did for the fingers.lodeling Metaballs Hands etaballs are another way to make a smoothly skinned hand. The mod techniques for a metaballs-based hand are very straighe“orward, and basic hand is very similar to a segmented hand, In this exercise, we be using metaballs primitives, such as elliproids, which may not be sported by all the software on the market. For those who are limited spheres, it is quite easy to fashion an ellipsoid out of a few spheres. tis hand is very basic, and a more complex hand will require more imitives. One downfill co a metaballs hand is that it can be harder to 3 realistic details, such as tendons and knuckles, using basic primi- such as ellipsoids #5: Building Smooth Hands with Metaballs 1. Start with a simple ellipsoid Duplicate this shape three 3. Duplicate the one finger to shape. This will become a times, Move the shapes to create the rest of the fingers, finger joint create a primitive finger. Copy two joints to create the thumb. Position these in the appropriate place 4. Create a shape for the palm, In this case, a squashed ellip- soid is used, but you could certainly use more primitives to create a more complex shape for the palm. chapter 4: modeling hands for animation iSs0eca5. When these shapes are in place, the hand should start to blob together. Most likely, the fingers will tend to over lap, causing a webbing effect. digital character animation 2 6. To get rid of the unwanted 7. Animating dhe mecaball blobbing, create individual primitives animates the groups for each finger and keeping it seamless the thumb. Each group will contain only those joints associated with the finger, along with the objects for the palm, ‘This basic metaballs hand shows you yet another way to create a hi with the others, che hands are basic building blocks that ean be expand upon to create more complex or stylized hands. Regardless of the shape or the look of the hand, however, the building blocks are the same. Adding Realism ‘As you can see, the hand is a very complex object to construct. All these exercises created simple hands that will work quite well for simple char acters. The more realistic a hand gets, however, the more attention to detail is required. A totally real hand will have fingernails, skin that creas ‘sand folds, and knuckles and tendons that show themselves as the fine gers bend. For toually realistic hands, it s probably best to construct the hand out of polygons or patches. Because segmented hands, by definition, don't have a flexible skin, organic realism is nearly impossible. Metaballs-based hands can theoretically get quite realistic, but the tricks and techniques required are usually very software-specific, and outside the scope of this book.chapter 4: ingernails nails are probably the first detail that needs attention. A fingernail be created in one of two ways. First, che nail can be modeled as part the finger itself, For short nails this is fine. In fact, a simple indenta~ jn where the nail resides, combined with a good texture map of a fin- might provide enough detail, particularly if the hands are never up close. and talons can also be made as an integrated part of the finger. use claws tend to be cylindrical or conical in shape, they can be ¢ of as an extension of the cylindrical finger. Because claws tend to be cylindrical or conical in shape, they can be thought of as an extension of the cylindrical finger. The second way of doing nails is to make separate fingernail objects and aitach them to the ends of the fingers, much like fake fingernails in real Iife This is probably better for characters with longer fingernails, and ca ‘work quite well for creatures who sport claws and talons, Knuckles Along with fingernails, knuckles tend to be a good place to add detail to the hand. Knuc of realism that will go a long way in selling the hand. Modeling knuck- esis not much of a problem. Getting them to bulge correctly, however, requires a sophisticated deformation system. les that bulge out as the fingers curl in will add a sense modeling hands for animation js(On a polygonal hand, a knuckle can be made quite simply by faceting the faces along the top of the hand that lie at the base of each finger. you to model tendons, digital character animation 2 ‘Modeling a knuckle is fairly straightforward. You merely add enough detail in the area of the knuckle and sculpt it. On the polygonal hand we built, a knuckle can be made quite simply by faceting the fa along the top of the hand that lie at the base of each finger. On a NURBS-based hand, use the blended wrist model, which has the splines oriented in the proper direction. This should enable you to sculpt the detail properly Tendons ‘Tendons extend out from the knuckles and disappear under the skin somewhere before the wrist. Tendons can be made in much the same: way as knuckles: all that is needed is enough detail to get the effect ofa rising tendon. On a polygonal hand, you will already have a line of edges running along the top of the hand toward the wrist. These edges can be faceted te the additional detail needed. On a NURBS-based hand, itis jn best to use the blended wrist model, which has the splines oriemt= ed in the proper direction to sculpt the tendons ‘Adding extra detail along the back of the hand enables Move the vertices in the extra detail up; this helps create tendons, More detail can be added, it needed.chapter 4: modeling hands for animation in Creasing ve had your palun read, you know chat the hand is full of cteases . Not only do creases ran along the palm, but also along the ‘of the knuckles, at the wrist, and a number of other places. modeling elsewhere on the body, modeling with wrinkles and in mind can help when you need to add such detail. For exam- the derail of the hand is running along the same lines as the the hand may just naturally crease as it moves. If not, you need 1 mote detail and force a crease, ler efficient method is to create a bump or displacement map that be animated as the hand moves. If you want to create actual geome- for the creases, your best bet is to choose a polygonal surface, it enables this sort of detail to be sculpted in very specifically, bing last consideration is the slight amount of webbing between the fin This occurs between all the fingers, but is most pronounced sen the thumb and forefinger. This can be sculpted in quite easily polygonal model via a series of edges. A NURBS model, the blends at the fingers create an area that tbe sculpted, which makes webbing between the fingers a bie dif Webbing between the fingers is not nearly as noticeable as the od tegion between the thumb and forefinger, however. In the created in this chapter, the thumbs are not blended, which enables webbing effect to be sculpred as needed, ‘many characters, this effect is small enough to be negligible; buc on types of characters, such as older people, for example, the effect is pronounced. If the webbing is incredibly pronounced, such as a for the Creature from the Black Lagoon, the webbing may have to as a separate surface inclusion can see, there are many ways to build a hand thar can be animar- cis ahvays a good idea to try a number of different techniques and ¢ fluent with chem all. Also remember that you can mix and geomecry types. You may find that a metaballs-based hand works for your NURBS-based character. In any event, be sure 0 keep ‘options open when approaching your characters.uolewiuy 10f speay BuljapowModeling Heads for Animation rhe head and face are very complex structures that are not easy (0 model, let alone animate. Sculptors spend years learning how to cre~ ate convincing heads and faces, and modeling a convincing head on the computer takes just as much practice. Adding to this task is the need for your digital faces to also be animated easily. Modeling a face is one thing: modeling one that animates well is another. The human face ‘can express a wide variety of emotions through the subtlest changes, in shape, Don't despair, however. The goal of creating a totally convincing face in the computer is something that eludes even the best animators. The basic methods used to construct faces are pretty much the same, regard~ face may be. After you've blocked out a less of how real or stylized th d and shaped into almost any type of char~ construct heads basic head, it can be modi acter, With a few of these tricks up your sleeve, you and faces that work quite well.The head consists of two large masses: the skull and the jaw. digital character animation 2 The Structure of the Head and Face In order to build digital heads and faces, you need to understand the underlying anatomy of the human head and face. The face is by far ‘most important part of the equation, because this is where the major of the movement takes place. The head typically acts as a solid frame the face to live in. Outside the facial area, the rest of the head doesnt. move much. ‘The head is basically made from two large bones: the skull and the j ‘The skull is really a collection of many smaller bones that are stitched together by cartilage. It can be thought of as a single mass balanced on the top of the spine. The jaw is the second major bone. Its movement affects the shape of the lower part of the face. When the mout opens wide, it is the jaw’ rotation that does all the work. Although we don't see the skull and jaw, their influence affects the structure and moveme of the face and the muscles that cover it, The Muscles of the Face The face is a complex collection of nauscles that pull and stretch the skin in a variety of ways. Understanding how these muscles work will guide you in the construction of your characters head and face. Lets take a look at these muscle groups. 4. Levator labii superioris—This muscle starts at the base of the i and connects to the middle of the top lip. Used alone, it pulls the top lip up into a sneer. b. Zygomatic major—This muscle lays actoss the cheek and conne to the corner of the mouth. Used alone, it pulls the mouth into smile, c. Risorius—This muscle stretches over the jaw’ and attaches at the comer of the mouth, Used alone, it pulls the mouth to the side and down, as when crying, 4. Triangularis—This muscle stretches over the lower side of the j and pulls the corner of the mouth down. It is used when frow or scowling €, Depressor labii inferioris—This muscle connects the lower lip the chin. It pulls the lower lip straight down, as when speaking,chapter 5: modeling heads for animation Mentalis—This muscle connects to the skin of the chin. When contracted, it pulls the chin up, farcing the lower lip into a pout COrbicularis oris—This muscle is attached to the corners of the mouth. Ie purses or tightens the lips. Orbicularis oculi—This muscle connects the check to the inner- eye area. Contraction of this muscle results in squinting. Levator palpebrie—This muscle attaches to the upper eyelid and raises it when surprised. |. Corrugator—This muscle runs from the bridge of the nose to the iiddle of the eyebrow. It pulls the eyebrows down and in, as when frowning or concerned. Fronralis—This muscle runs across the forehead and connects to the eyebrows. It pull the eyebrows up.The muscles surrounding the mouth pull the skin radially ‘outward and also allow the lips to pucker. digital character animation 2 The muscles around the eyes move the brows up and down, as well as open and close the eyelids. All this anatomy is great, but how does it affect the way a digital model is created and animated? The trick is to understand how these muscles pull and shape the face to create expressions. Really, the groups of mus- cles fall into two categories: lower fice muscles that control the mouth and java upper face muscles that control the eyes and brows. Acthe mouth, the muscles do three major types of movements. First, the muscles that lie across the cheeks and jaw pull the lips outward radially from eight major anchor points. Second, the muscles surrounding the lips contract to purse the lips, forcing them together and forward, Finally, the jaw can drop. Acthe eyes, the muscles make three major movements. First, they open and close the eyelids, Second, the frontalis muscles on the brow raise and lower the eyebrows. Third, the corrugator pulls the eyebrows in toward the bridge of the nose, furrowing the brow. If we can create a digital model that moves easily along the same lines that these muscles are pulling, we have a much better chance at an ing the face convincingly.chapter 5: modeling heads for animation Basic Head Modeling Methods designing a digital head, it's best to take a good inventory of your tion tools before starting. It is very easy to back yourself into a orner by designing 2 head that cannot be easily animated. There are a thora of ways to create and animate a face. so it is up to your imagi- tion and creativity to think up new and innovative ways to get this done. and puppet animation. Think of imby or Mr. Potato Head. These characters had a simple shape for the gad with the features simply tacked on. The Rankin-Bass Christmas als of the "60s (Rudolph the Red-Nosed Reindeer, Santa Claus Is ing to Town) used his method effectively with puppets, and the principles apply to digital animation as well. paste-on features, most or all of the character's features are e objects. This fives the animator from complex modeling and animation, paste-on features can be made with the simplest of objects. Eyes be simple spheres, the mouth is easily constructed fiom a torus, d the nose isa simple sphere. Replacemeat animation methods can be used for che mouth. Animating shapes as simple as these is easy Iost packages. OTs) © -_ character has a simple face. ‘As you can see, mast of the facial features, Including the nase, are pasted on.N\ digital character animation 2 This ¥s another character with paste-on features, The mouth on this characteris built in, but the rest of the This face isa simple texture map, but it stil maintains a lot of character features are separate objects Another method you can use is to paint the features on with animated. texture maps. This makes the modeling task downright easy, because the head can be as simple asa sphere—the detail is in the maps, They are drawn and animated by hand or in a 2D paint/animation package. This technique can prove cumbersome, however, because matching up the 2D animation of the face to the 3D a ation of the body usually requires flipping between several packages. This sort of animation is best y of the texture maps or the visibility of the geometry containing the texture maps is animated to make the mouths appear and disappear. A Face with Flexible Skin lefé to techniques such as replacement animation; the transparer with cs. This requires more complex modeling meth For realistic and quasirealistc faces, you construct a head and fac skin that moves and fle ods. Ir also requires a package that can support sophisticated shape ani= mation of the many different expressions and poses the head and face will take. Polygonal Heads Polygons can make faces that are very realistic. The advantage to con- structing a face with polygons is that you have much more control in ning the surface of the face. It is much easier to cut holes for ¢} sockets and nasal passages with polygonal methods. To get a smooth head with polygonal methods, however, you need to have a lot of detail, Performing the complex shape animation required to deform such a head can tax even the fastest system. To circumventchapter 5: modeling heads for animation 3,2 low-resohution version of the head can be animated and added through subdivision after the deformation takes place. If posible, a matching high-resolution version of the head can dat render-time. With care and attention to detail, however, it ble w animate faces quite effectively with a good polygonal ani- package. Based Heads ed heads are another terrific way to create digital flesh. ig a head with patches enables you to create faces that closely the real thing. Because patches define a smooth surface with aller number of vertices than polygonal methods, controlling the of the fice and maintaining a smooth surface is quite easy. ling with patches is an exercise in Bauhausian simplicity. Less is and form follows function. Keeping your models light is the key. simpler the structure of your digital head, the easier it will be to = Every time you add a control point, vertex, or extra spline, exactly what its function will be. If you don’t need it, don't my patch modelers cat restrict you to surfaces without holes or ks To get around this limitation, you need a NURBS-based model- rth trims, which enable you to punch holes in the surface. Without eye sockets and mouth holes ean be somewhat difficult to model. 1 your particular modeler’s capabilities, patch-based heads can ructed using a number of different methods and topologies pology and Facial Modeling modeling a face, you need to consider the way the face moves rfl study of the muscle chart in the first part of this chapter shows the muscles around the mouth pull che lips radially outward, but @ brows are miote horizontally oriented. This knowledge can guide fn the consteuction of your face ako need to consider the design of your character. A realistic has diferent requirements than a stylized character—a cartoon diay not be ideally constructed, but it makes the rest of the body ay to managedigital character animation 2 ‘Topologically, a head can be thought of as a deformed sphere. The sphere can be oriented any one of three ways: with the “north pole” the head at the top of the head, at the front of the head (mouth), oF at the side of the head (ears). The orientation of the sphere when you st modeling affects how the final model will look when animated. The following methods apply primarily to patch- and NURBS-based. heads, but the techniques can also be applied to polygonal models. OF course, polygonal modelers have a bit more freedom with the topolo~ gies thar they can manage. The Pole at the Top of the Head The pole at the top of the head is the topology that usually comes fist to mind for most modelers, and many good-looking heads have been built chis way. This method has nwo main advantages. First, the pole at the bottom can be opened up, allowing the character co seamlessly co nect to the torso, This is important for NURBS modelers who want limit the number of blends in a body: Second, the detail in the head runs horizontally, which enables better control of the eyes and brows. The big disadvantage concerns the area around the mouth, Horizontal lines work great for the eyes, but not for the mouth. To get a decent mouth cavity, a number of splines must be bunched up at the corners of the mouth. This can cause creases that show up when the mouth is animated. If your character doesn’t talk, however, chis may not be With the pole at the top of the head, the detall This topology also enables an open neck, which mal is arranged horizontally. for a seamless attachment to the torso.Chapter 5: modeling heads for animation Pole at the Mouth nn as a radial head, the pole at the mouth method is very and certainly makes the best mouths. By placing the pole at th, the detail naturally flows along the same radial lines as the of the face. This topology makes animation of the mouth quite and easy: The radial method may limit you when it comes to internal eye sockets, however, because the orientation of the can be off. ‘ean be overcome by sweeping the detail of the sphere up toward top of the head. This enables you to sculpt the detail so that it is horizontal at the brows, giving you the capability to sculpt the more effectively : ping the derail up in this manner, however, causes another limita ‘where the head connects to the neck. The head wi the neck. Many times, this is not a problem. Ifthe neck needs to be ess, however, che detail must be swept down rather than up, and detail around the eyes sacrificed. Deciding which is more important inds on the design of your character. A character with separate eyes eyebrows, for example, will not need derail at the brows, allowing a seamless neck need a seam, disadvantage is, to get a decent mouth cavity, a number of splines must ‘be bunched up at the corners of the mouth, which can hamper animation,a digital character anima Putting the pole at the mouth creates a head that has A closer view of the mouth shows the radial mouth, a very flexible mouth. ‘where the splines are arrayed radially, just like the muscles that surround the mouth, Sweeping the detail of the head toward the top ‘Sweeping the detail of the head toward the bottom makes makes the brows and eyes easier to model, but the brows and eyes harder to model, but gives you an open forces you to place a seam at the neck. ‘neck, which enables a seamless attachment to the torso. The Pole at the Ears The pole at the ears is che least popular of the methods because it buys you very litle. In this method, the direction of the detail has a good orientation for eye sockets, but not much ese. Like with the first exam- ple, the mouth will be problematic because the detail will bunch up in the corners of the mouth. The neck will also need to be seamed. Still, skilled modeler can make a very convincing face using this method.chapter 5: modeling heads for animation the poles at the ears, the direction of the detail has The extra detail needed to make a mouth cavity causes ‘orientation for the nose and brow, ‘the splines to bunch up at the corner af the mouth. ich Topology to Use? re are many opinions as to which topology is best when construct a single surface head out of a sphere. The best way to choose is to fe your character and its needs. i character needs to talk, or move its mouth at all, place the pole mouth. This makes the mouth fir more controllable, though isa bit of a trade-off when it comes to attaching the head to the This, however, is a much casier problem to deal with than a poorly ‘mouth yur character doesn't talk or move its mouth, other topologies can ‘used. Having the pole ar the top of the head is probably the second fice because it enables a seamless neck, which can make attaching head to the body less complicated. the pole at the ears offers the fewest advantages of the three. only is the mouth hard to control, but the neck must be seamed lldigital character animation 2 Exercise #1: Modeling a Simple Head This exercise shows you how to create a basic head out of a sphere. Bec talk, it is modeled with the pole at the mouth. This gives you a radial mouth that animates quite wel. we want this character to This particular head was modeled using NURBS patches, but the same techniques could be used to model a polygonal head. The head being modeled is fairly realistic. Don't let this deter your imagina- tion. The same basic structure can be used for a wide variety of heads—both real and stylized. 1. Start with a simple sphere 2. Select the vertices on the 3. Pull these points straight with approximately 8 sub front part of the sphere back to create a hollow divisions in either direction mouth cavity inside Position the pole so that it the head. is facing forward,chapter 5: 4, Select the vertices around the lips and scale them down vertically: You now have the beginnings of the character’ lips. Sculpt the added detail to make the lips more realistic At this point, just try to get the basic edges of the lips defined. More detail will be “added later. modeling heads for animation 5. For this particular head, we 6 will sweep the detail up so that there is a pole at the top rear of the head and another at the back of the throat. Select each row of vertices, and then move and rotate them so that the detail sweeps up and back. Reshape the head as needed. 8. Add a few rows of vertices 9, near the brow. After you have a basic shape blocked out, start addin detail. Start by adding a few rows of vertices near the lips, Again, sculpt out the brow How prominent you make it will depend on the design of your character, but try to get the basic detail running horizontally,Exercise: continued 10. Now for the nose, Add side of the vertical center, columns of vertices on either digital character animation z 11, Pall out the vertices residing 12. After the nose is blocked just above the lips to create out, youre halfway there the nose. (This new detail The rest of the work is sime can also be used to flesh out ply tweaking the shape and the top lip) The exact shape adding extra detail where of the nose will be dictated necessary. Getting the shape by your character's design. exactly right requires time, More complex noses may patience and a good eye. require that more detail If you decide to add new be added. detail, do it gradually 1 keep the surface smooth, More Complex Heads Heads made by deforming a sphere are quite flexible, and can be turned into a wide variety of shapes. This type of head, however, always has trade-off. The detail around the mouth may need to be changed to accommodate the eyes or the neck. This can prove frustrating for more demanding characters In this case, a more complex head may need to be built. This can be accomplished in a number of ways. The simplest is to model the head as a polygonal surface and use subdivision to get it smooth Another way is to build the surface of the head out of multiple surfaces. ‘The eye sockets, for example, may be built out of a separate surface and blended on a sphere-based head. This can avert some of the topological limitations of the spherical head.chapter 5: modeling heads for animation The face may also be buile out of multiple patches seamed together. Depending on the software chosen, this technique may be more com- plex, but the basic topology is similar to that of a polygonal head. Modeling a Polygonal Head Polygonal heads may be either simple or complex, The big advantage t0 2 polygonal head is topological freedom. This enables you to arrange the detail so that ic works best in each area, Unlike a sphere-based head, a polygonal head does not need to compromise the mosth detail to accommodate the eyes or the neck, This particular model demonstrates this point, The mouth on this char- ‘acter is radial, much {tke in the previous exercise. The eyes and brows, however, are not compromised, Like the mevsth, each eye has a radial topology: The brow is naturally horizontal, which makes it easier to sculpt and animate. 1. Start with a simple cube 2. Start by moving the detail 3. Move the detail on the left This particular cube has 7 around on the face of the side of the center together. rows and 5 columns of ver~ cube. It is easier to manipu- Do the same for the right. tices, and is 3 rows deep. late the detail on a flat sur- face, so the head is kept as a cube while all the detail is added. First, take the two columns of vertices on either side of center and move them together.ry digital character animation 2 4, Take the two rows on either Now make some eye detail Select the polygons near where the eyes will be and inset them. If you don't have an inset command, you bevel or extrude the faces side of the center and move them closer together. n and then move the faves back so that they are flat with the rest of the cube. 7. Create detail for the mouth. Select the two rows of poly- gons beneath the eyes and inset them. If your modeler 8. Repeat this operation a few more times to get enough detail to sculpt the mouth. does not have an inset fane~ tion, extrude or bevel the faces out and then move back the faces so that they are flat with the rest of the cube, 6. Repeat this operation a few times to get enough detail shat you can sculpe the eyes. 9, Select the comers of the mout and scale the vertices inward t round off the outline of the mouth, (If you look closely, th decal will look a great deal Ii the radial mouth created in the previous exercise.) Round off the corners of each eye in the:chapter 5: modeling heads for animation 10. Up until now, the head has Work on the profile of the been a cube, which made it face. Pull out the cheeks, and ‘easy to add the required de~ sculpe the lips (arrow). tail. After the detail isin place, sculpt the cube into a more headlike shape. This can be done using a number of wols, such asa lattice or another type of deformer. It can also be done the old-fashioned way, by tugging vertices. . Create eye sockets by select- 14, Create the mouth cavity by ing the polygons inside the selecting the polygons along eyelids and extruding them the inside of the lips and inward. You may also choose extruding them inward, to delete these polygons (ight) to create holes for the eyes rr 2. Start sculpting the eyes. Push the detail along the top of the eye up to create a brow ridge. Sculpt the brow from this detail. Reshape the eyes to match your character's design. 15, Create the nose by selecting the polygons above the lips and extruding them outward,Eyes internal to the head give a realistic look when rendered. An eyeball is really just a simple sphere or hemisphere digital character animation 2 16, Round out the nose to 17. When the low-resolution the desired shape. Finish version is complete, it can sculpting the rest of the face be subdivided to create the to match your character's high-resolution version. design Eyes Eyes are one of the most expressive areas of the fice, so its essential to have eyes that are controllable in every respect, Eyes tend 10 fill into two broad categories: internal and external, Internal Internal eyes are akin to realistic eyes. The eyelids are part of the facial surface, with the eyeball inside the skull. If your character design dictates internal eyes, you plan for this by modeling eye sockets when construct- ing the head. This was done on the polygonal model chat was construct- ced in this chapter. For a NURBS model, one way to do this is to create a ietle posich for the eyeball te sit in This, however, can add a lar of detail that may get you at animation time. Another way is to trim out the eye socket and then blend a second surface for the eyelids to the original head.chapter 5: modeling heads for animation ry lds can be modeled as part of the surtace of the head. The lids can also be separate objects, such as a herisphere Cor these shutters that show only shen the eyes biink eyeball itself can be either a sphere or a hemisphere, because only front pare of the eye will ever show through the skin, The pupil can tade with a simple texture map or a second hemisphere sitting on first. big problem with internal eyes is that you need to keep the eyeballs in the socket while the head moves. This is usually accomplished ith 2 hierarchy—the head being the parent of the eyes. Another issue the movement of the eyelids. They must move along an arc that’s same radius as the eye. This ean be done using a multiple-target— hing utility. It also can be accomplished by animating the lids using ge or some other form of direct manipulation. wer Way 10 create lids for internal eyes is to model a separate lid animates and remains hidden until the eyes close. For fast eye clo- , such as blinks, this can be fine, but if the lid is closed too long, the lence may pick up the seam where the separate lid pokes through skin. rnal smal eyes are separate objects made from spheres or cylinders that ‘on the surface of the face. They can have a more cartoonlike appear- and are easier to control than internal eyes. Because they don’t we to line up exactly with the eye sockets on the face, you have h greater freedom over how they're placed. They're great for Avery-style eye-popping and afford the animator a variety of tic choices,108 | digital character animation 2 G a Eyes external to the head are made of spheres that sit External eyes can be exaggerated wildly to get a cartoonish Con the surface of the face. They work well for styized ook. characters. Eyes are easy to make and can be made from any type of geometry. ‘These can be used as internal or external eyes with the heads you made earlier. 1, Start with a simple sphere. 2. Create the pupil. This can This will be the eyeball. be done by creating a very shallow hemisphere widh a slightly larger radius than the eyeball, but which shares the same center. The pupil will glide above the eyeball much’ like a contact lens.chapter 5: modeling heads for animation . Alternatively, the pupil can 4. The lid is created in much 5. Fit the lid so that it shares be created as a texture that is the same way as the pupil. It the same center as the mapped directly to the eye- is a half-hemisphere slightly eyeball. Next, set up the ball. For a complex eye, you larger than the eyeball. hierarchy so that the eyeball can have vo maps on two parents the pupil and the lid. meshes: one for the itis, another with transparent edges for the pupil. The mapping coordinates of the pupil can then be sealed to make it dilate, A second lid can be created 7, The eye can be blinked | ‘by duplicating the first. simply by rotating the lids.na digital character animation 2 Asymmetrical Eyes Most realistic characters will have spherical eyes like those created in th previous exercise. A stylized character, however, may need eyes that are oblong or asymmetrical. This can be accomplished in a number of ways, For oblong eyes, you can use a trick that depends on how hierarchies work. Ifthe eyeball parents the lid and pupil, the eyeball can be scaled) either vertically or horizontally, Because the lids and the pupil are chil. dren of the eyeball, hey inherit the scaling, allowing them to move pet fectly along the surface, even though the eyeball is nonspherical. The scaling can also be animated to give a squash-and-stretch effect. You can take this concept one step further with the use of a lattice. Applying the lattice to the entire eyeball hierarchy enables you to distort the shape of the eyes even more. Because the lattice distorts the entire hierarchy, che lids still move along the eyeball’ surface. The latice can also be animated to give a very cartoonish effect. Oblong eyes can be created by scaling the eye vertically Applying a simple lattice to the eyeball hierarchy enables (or horizontally you to distort the shape of the eyes 10 get cartoonish shapes, while stil allowing the its ta blink. Eyebrows Eyebrows go a long way when you are trying to animate emotion, so be sure to plan for them when designing your characters. Depending con the design of your face, the eyebrows can be modeled as part of the surface of the face or as separate objects0 Es “Brows can be modeled by creating a. Eyebrow hair can be painted on as.a_—_—_This dog's eyebrows are separate “brow ridge that runs across the tops texture. You can also go further by objects that float above his head the eyes and the bridge of the nose. creating geometry to represent the They are very stylized, but can be individual hairs. moved much rove easly Eyebrows that are part of the surface of the face can be modeled by cre~ ating a brow ridge that runs horizontally across the tops of the eyes and the bridge of the nose. How prominent the ridge becomes depends on a ‘umber of factors. A large, lumbering character may have a very heavy ‘row, for example, but an infant would have a very subtle brow. After the Drow is modeled, you can create eyebrow hair along the brow ridge. These can be painted on as a texture. You can also go further by creating “geometry to represent the individual hairs Eyebrows separate from the face are fess realistic, but can take a wider “ariety of shapes and positions. These can be constructed from a variety ‘ofshapes and can float five or ride along the surface of the forehead. You animate the eyebrows using any number of methods, including shape ‘imation, cluster, lattices, or a skeletal deformation system. Teeth : use teeth are rigid objects, they ate fairly easy to construct out of ‘polygons, but they certainly can be constructed of patches as well Basically, teeth are rectangular in shape and slightly rounded at the cor~ ‘nets. Teeth can be constructed individually and grouped together in a Hierarchy or constructed as one solid object. Teeth are slightly curved Teeth can be modeled as a solid object and should be modeled to follow the inside of the skull and jaw. ‘ora tooth ata time, chapter 5: modeling heads for animation =zi digital character animation z ‘Teeth ate not usually modeled as part of the head, but are placed insid the surface of the head behind the lips. Because we usually don’t moc 1 skull or a jaw, the teeth are the audience's only clue for this underi structure. To maintain believability, its best to move the teeth the same vway theyre moved in real life. In real life, top teeth are attached to the skull and do not move in rel tion to the head as a whole. To get this effect digitally, simply use the head as a parent and don't move the top teeth, Bottom teeth are aitached to the jaw and rotate along with it when the mouth opens. One way to assure this is to make the bottom teeth pivot around the same axis that the jaw rotates. This is located slightly in front of and below the ear. That way, the lower teeth can give a good impression of a true jaw. Teeth are not ususlly modeled as part of the head, but are If you want, you can aiso model gums for your character placed inside the surface of the head behind the lips. The top teeth are stable to the skull The bottom teeth rotate with the jaw.chapter 5: modeling heads for animation gues tongue is important when animating speech, though not as critical ssome may think. Most of the characters in Toy Story, for example, did Ihave tongues, Still, its nice for your character to have a tongue, to an extra bit of realism ie tongue is a very flexible object and can be animated with a skeletal nation system, a lattice, a cluster, or some other form of shape ani The tongue is typically a separate object that floats inside the th cavity. Because we rarely fly our digital camera down our char~ ’S throat, the back of the tongue is rarely seen. easiest way to build a tongue is by squashing a sphere and putting a down its length. Texture and bump maps can also be used to make detailed surface. JtS a good idea to fade this texture to black at back of the tongue so that the back of the tongue remains hidden, Que in three easy steps: A sphere is created, then its Typically, the tongue is @ separate object that floats inside hed, and then a dent is modeled down the middle the mouth cavity‘You can model a neck as a small protuberance from the ‘A NURBS blend can aso be used to help hide the seam body and simply place the head over it. digital character animation 2 Connecting a Head and a Body After you've made the head, you create a neck and attach it to che body. There are three basic ways to do this, Neck on the Body You can model a neck as a small protuberance from the body and sie ply place the head over it. This gives you a bit of a seam where the head connects. Far those who model with NURBS, a blend can be used to help hide the seam where the head hits the neck. vuhere the head hits the neck. Neck on the Head You can also extruce a neck from the head and attach it to the body at the shoulders. This is good, because it leaves no seam on the open skin ‘of the head and neck. The seam of the neck and body can be better hidden on the shoulders or underneath a piece of clothing. For a patch: based surface, the surface of the neck can also be attached to the body, using tools that keep the surfaces locked together.chapter 5: modeling heads for animation us eeping the neck as part ofthe head For polygonal models, connecting the seamless connection for a NURBS- ples you to hide the seam on the head to the body is relatively simple. based character requires that the shouder or beneath a colar or other detail of the head and neck ine up, Bit of cotring. Head, Neck, and Body as One Object For a completely seamless connection, the head can be modeled as a part of the torso. For polygonal models, this is relatively easy. Many times, the head is built from polygons extruded ftom the body Conversely, the body can be constructed by extruding polygons from the lower part of a head. The vertices of the body can also be attached ‘by welding vertices together and building faces that span the two “objects. A number of packages have automated tools to help with this task. For NURBS-based characters, a seamless connection means planning the topology of both the head and torso so that they are open at the neck. Additionally, the detail lines of both need to line up exactly. A Iiead with the pole at the top of the head is a good candidate for this ‘ype of connection. A head with a radial mouth can also be used, but only if the detail is swept down to leave an open neck. Conclusion With the character’ head firmly attached to the body, you should have conseructed at least one complete character. When you've finished your character’ head and attached it to the body, you should be ready 10 ‘pegin animation. The next few chapters discuss setup and aniznation of the body, with the final chapter on facial animation.uolyewsofag ysaw pue suojaja4sSkeletons and Mesh Deformation fer your character is modeled, you get it ready for animation, If your character is fully segmented, this may mean joining the seg- ments into one hierarchy. If your character is partially segmented, or if its a single-skinned character, you deform the skin. Typically, this is done using a skeletal deformation system Ina skeletal deformation system, the skeleton is what is actually animat- ed. This makes setting up the skeleton a critical phase in the character- building proces. A properly built skeleton can be quickly and easily manipulated to attain any pose. In this chapter, we concern ourselves swith basic setups. Afer the skeleton is built, it will, in turn, deform the character. When rendered, the character will hopefully look alive to the audience.digital character animation z Hierarchies and Character Animatio! Depending on how you designed your character, you have either a collection of segments or a single skin to be deformed by a skeletal deformation system. Either way, you need to find a w: segments and/or skeleton into a cohesive character. You want the to move with the arm and the legs to move with the hips. You can your character together in this manner using a hierare! to connect A hierarchy is simply a way to tell the computer how the parts of character are linked. Taken literally, the hierarchy tells the computer foot bone is connected to the shin bone, the shin bone connected to the thigh bone, and so on. Technically, the hierarchy looks a bit like a tree, with each connection forming a branch, It can ako be thought as similar to the nested directories and folders found on your comp! The foot is a subfolder of the shin, the shin a subfolder of the thigh, 30 on, Manipulating Hierarchies When manipulating a hierarchy, moving the parents moves the chil This is also known as forward kinematics. In an inverse kinematics ronment, the opposite happens—the children drive the parents. This i discussed a litle bit later. In a forward kinematics environment, if you move a child object—a twig, for example—you won't move the parent—a branch or the t Bur, if you move the trunk, everything else, including branches and twigs, follow along. Just like a tree, a hierarchy must have a trunk, a sin gle parent that controls all the other branches. The same follows fora character's body. In a human skeleton, the trunk is almost always the hips or pelvis. The pelvis is close to the center of gravity of the human body, which makes it a good candidate for a parent. More importantly is the center weight distribution for the entire body. ‘The hips support the spine and the entire upper body, passing this weight down to the legs. This makes the spine and legs children of the hips. In turn, the spine parents the shoulders, which have the arms as their children, and the legs parent the feet. Having the trunk, or the of the hierarchy, at the hip area helps you create motion such as a back flip, for example, because that rotation is dependent upon the rotation the hips.chapter 6: skeletons and mesh deformation hy HEAD Tho ic hierarchy oF ain body NECK prmenenae tema fa body folows VY SHOULDERS BICEP 4 ‘ SPINE 2a JREARM 4 FOREARM HIPS 1 AND J \ HAND THIGH THIGH SHIN SHIN FOOT FOOT setups have everything linked in one single hierarchy. There are en you need to create a broken hierarchy. This actually keeps the joints separate to aid in animation. A character in a dress, Ke, might have the legs as separate objects, which could exist rate hierarchy. Another example might be an extremely simple msuich as a flour sack, where the leg skeleton is only one joint. ¢, it might be casier to keep the legs outside the hierarchy. and Joint Rotation ling a skeleton or segmented character, you need to tell are how the joints of your character's body rotate. The shin, = rorates abour the knee. When you create an object, such as ‘ofa shin, the computer has no way of knowing its supposed to and no way of knowing that it rotates about the knee. It sim- the pivot to an arbitrary point—typically, the physical center ject.pues! digital character a mation 2 The forearm is rotating around its default pivot point—the To make the forearm rotate propertly, you need to ma center of the object. Real forearms rotate, reassign the pivot so the forearm rotates at the elbow. Unfortunately, in character animation, arbitrary points are not accept able. As you assemble your character, you may also need to move the pivots on your joints so that everything moves properly. The forearm rotates around the elbow, the thigh rotates at the hip, the head pivots the neck, and so on Exercise #1: Building a Simple Segmented Character This exercise builds a segmented character out of spheres and cylinders. It can be constructed in almost any package. The character is simple, but itis designed to show how a basic hierarchy is put together. 1. Start with a sphere 2. Select the polygons on top of the sphere,chapter 6: ide these faces upward. will create a simple ball- tick shape chat will be for the joints of the eter. two cylinders. One presents the hips and is eed at the base of the ine. Make this segment the sent of the spine hase. The p evlinder represents the uilders, and is the child of top of the spine skeletons and mesh deformation | 4. Adjust the pivot of this seg~ merit so that it matches the center of the sphere. iit 7. Hang two more ball-and- stick joints off each side of the shoulders to make arms. ‘The shoulders parent the biceps, which in turn parent the forearms. Simple spheres at the end of the forearms can act as very simple hands. Copy the first joint ewice Place the copies end-to-end, as shown, to create a basic spine. Make the base of the spine the parent of the mid dle sewment, the middle seg ‘ment the parent of the top. Dopy the arm segments and translate them down to the hips. Make the hips parents of the thighs, and each thigh the parent of its corresponding shin,ital character animation 2 9, Adda simple sphere for 10. Add two flattened sph the head. for the feet. If the hierarchy is constructed properly, moving the hips should move: the rest of the body. Rotating the thigh should move the shin and f If the pivots are in the right place, all the joints should rotate around sphere at the base. Move your character around the screen and test rotations on all the joints, Put this character aside for later exercises. Skeletons and Hierarchies Single-skin characters do not have segments, so their meshes must be deformed. In this situation, the skeleton is most often used as the deformer. The skeleton is built from bones and joints, objects that are usually tetrahedral in shape and do not render. They fit inside a single- skin character much like real bones, and then the skeleton is conn in a hierarchy much as in the previous exercise. Bones typically don't render, which is fine because they are really only helper objects. They. simply act as a guide for the mesh deformation utility, telling the wii where to move the mesh. This is discussed a bit later. Many software packages also enable skeletons to be made of regular objects. These objects can be any type of geometry, but most people simply use boxes because they are easy to model. As with bones, the boxes act as guides for the mesh deformation plug-in, Because regular objects are visible to the renderer, these objects must either be hidden or have their rendering attributes switched off before the character is rendered,chapter 6: skeletons and mesh deformation = Bones are created in several different software packages. Even though the bbones lock slightly different, the pack- ages all work pretty much the same. In addition t0 bones, many packages enable other types of geometry to be Used in a skeleton. Here, we have 2 skeleton made of bones, one made ‘of bones and boxes, and ane made of just boxes, All can be used to deform a mesh, ipulating Hierarchies ‘character or skeleton is connected using a hierarchy, the joints s manipulated. There are two strategies for doing this: forward and inverse kmematics. As you will see, each has its own and advantages. Many characters are built so that some manipulated using forward kinematics and other parts using,rr 1. Placing this character's hand on the cup using forward kinematics is an exercise in rotations. digital character animation 2 Forward Kinematics Forward kinematics (FK) manipulates the character from the top of hierarchy down. This is what you did to test the character in the pr ous exercise. If you move the pelvis, the whole body moves; if you rotate the elbow, only the hand moves. When you use forward kine ics, you soon realize that, with the exception of the pelvis, rotating character’ joints is the only way to move the body parts around. If you want to place a character’ hand on a coffee cup, for example, ‘you first rotate the shoulder, then the elbow, then the wrist and fin working your way from the top of the hierarchy down. Each rotation brings the hand closer to the cup. You can't simply pick up the hand and place it on the cup, because the arm will not follow @ wr 2. First, rotate the arm at the 3. Next, rotate the arm at shoulder, elbow to meet the cup. working with a character using forward kinematics, you work from the top of the hierarchy down, Inverse Kinematics Inverse kinematics (IK) is simply another way of manipulating a ch ter. If set up correctly, it can be an animator’ dream. It is the exact opposite of forward kinematics, because moving the children rotates, the parents, Inverse kinematics is simple to use: You place the characte hand on the coffee cup, a the cup when you move it. This simple action is more complex that id the rest of the arm automatically follows think, however, because the software must solve the rotations for all joints so that everything looks natural, Luckily, most advanced software thas features to assist IK, such as a damping fonction that softens the motion at the joint approaches its limitdisadvantage of IK is that you get only one key for the whole hy, becauise each joint’ motion is being driven by an end effee~ For two joined limbs, such as human arms and legs, this may not problem, bue in a many-jointed limb, such as a tai, you lose con the limbs shape. In this ease, FK may be the best bet. computer doesn’t know how the joints of the body are supposed Ifits just as easy to bend the elbow backward or sideways the problem, thar’ fine with the computer. Unfortunately rthe animator, backward-bending elbows are neither a natural nor ed solution. Elbows are a hinged joint that can bend only When positioning the hand with d. Without this information, the computer can’t position the inverse kinematics, the hand is simply opeciy pulled to the coffee cup and the est of the arm follows. three diferent ways to rotate the joints so that the hand rests on the cup. Only the first of these Jscorrect, but the computer doesn’t know that. You need to tell the computer exactly how you {your joints to behave. ating IK Chains paler for the computer to position the joints properly. it needs to abit of information. Firs, it needs to know how many bones, or ate in the chain, as well as their respective lengths. This is typic: when creating the j hy is created automatically within the software. ws on the computer. As the joints are drawn, first joint in the chain is known as the root. Ie also the root of the g It may also contain data helpful for the positioning of the depending on the software. Moving the root moves the entire When drawing a chai Phierarchy. If you are drawing a leg, for example, start with the hip ork down to the ankle and foot. Iways start with the uppermost joint in chapter 6: skeletons and mesh deformation a-——— Joint +——— Bone Effector The parts of an IK chain Translating the effector bends the Ik chain digital character animation z ‘The tip of the last joint of the chain is known as the effctor.The controls the position of the end of the IK chain. The software a tries to position the chain so that it runs between the root and the effector. Between the 1001 and the effector are bones connected by joints’ are simply articulated points in the chain. An arm has one joint elbow—and a spine may have many joints. Manipulating a Chain ‘The chain is manipulated through the use of the effector. As the el is moved, the joints of the chain rotate to meet the effector. This posing and animating a character easy, because you need to consider only the position of one effector rather than the rotation of many joints. You give up some of the control you had with FK, bur. in re ‘you can animate with ease, much like working with a puppet. In fact, some studios call the low-res version of the skeleton “the puppet.” What happens when you pull the effector beyond the limits of the Joints? Most software keeps the joints at a fixed length, so the fally extended chain aims itself at the effector. Some software, however, enable the joints to stretch, expanding their length of the entire chain meet the effector. This sort of effect can be good for squash-and-st Squash-and-stretch on bones enables a large foot to pass under a sh body by stretching the leg as the foot passes, for example, At the other end of the chain, translating the root typically moves the entire chain, including the effector. This is because, hierarchically, the effector isa child of the root. This means that when the root moves, the effector moves, Here's an example: If your IK chain is a character, = Sy When the effector is moved beyond —_f the root is moved, however, the the limits of the chain, the chain simply entire chain, including the effector, aims itself atthe effector also moves.hapter 6: skeletons and mesh deformation he root is the hip and the effector is the foot. If you move the top ie leg, the foot moves as well. This is not the way to set Up a skele- "you want the feet to remain fixed co the floor. aints around this problem, many packages use a device called a con- A constraint tells one object to stick to another object. Ifyou «da ball to a box, moving the box also moves the ball. On the ce, constraint seems to work the same as a hierarchical link. You d just as easily parent the box to the ball to get the same result— the parent (the box) would move the child (the ball) The dif- skeletons constrain the effector to an object outside the leg hier Because this outside object does not move with the root. it fixed in space. The effector, in turn, remains fixed to the outside “This enables the root of the chain and the effector to be moved tly, which makes posing the character much easier. ent elbows (and knees) from bending backward, you need t© the software exactly whar the limits are for apy given joint. Most es enable this to be configured o1 have different types of j n specify a joint as either {evo-dimensional) or ball-and-socket (three-dimensional). -hy-joint basis, Some itions and Goals ages now have incorporated the idea of a default position, own as a rest position or a goal, for a series of skeleton joints. «es the chain to return to the default shape when the effector is ito its default position, Typically, this is the position that the joints jt created, but some packages enable this position to be rede- any time. helps considerably because it makes the behavior of the aguite predictable. It is also exceptionally good for chains with [pist, most machines simply weren't fast enough to solve IK equa- jaime. This meant that the equations had to be solved in a € pass, which “baked” the joint rotations into the joints. Most machines today are fast enough so this is no longer a limitation. ‘A constraint is used to stick the effec- tor to the box. Because the box is out- side the hierarchy, the effector remains in the same place, even when the root is moved. \ ‘A rest position ahvays returns the chain to the same shape when the effector is in the same placedigital character animation 2 As such, most 3D packages offer real-time IK, which simply means mnhances interact the equations are solved in real-time. This greatly and posing of the character. There are times, however, when you may still need to calculate the IK solutions and bake in the joint rotations. Most of these situations involve transporting the animation to other places. One good examy is in video gaming. Most game engines accept only rotational data, forcing you to bake in the IK solutions. Those wsing a third-party renderer may also need to solve IK in order for the animation to be properly managed. Which Kinematics to Use? As you can see, inverse kinematics systems have a few more options, which make them a bit more complex to set up. After they're set up, however, animation is much faster. Forward kinematics, on the other hand. is very straightforward. The question of which technique to use depends on each individual part of the skeleton, Forward kinematics is most commonly used in the upper body. The spine, shoulders, and arms can be animated using forward kinematics, a character is standing and gesturing with his hands, this works just Ihe takes his hat off and throws it, touches another character on the shoulder, holds hands with another character, or basically interacts wit anything else, IK may be required. ‘The spine and hands are usually mated with FK, though IK can be used. ‘One place where IK is almost always used is the legs and feet. IK is ideal for keeping a character’ feet locked to the ground. If the charac ter is swimming, however, FK may need to be used on the legs and fe as well as the arms. As you can see, many times the choice of one type of kinematics over the other depends on the specific demands of the shot.To get around this, many productions have more than one setup. for a character. Which particular setup gets used depends on the task at hand. Building Skeletons Building a skeleton is similar to putting together a segmented character, The big difference is in the use of IK chains, as well as constraints, The next two exercises build a skeleton for a body and for a hand.ing a Body Skeleton ‘on for a body should be built with your mesh as a reference. It good idea to place the bones inside the mesh as you build the to ensure a perfect fit, Some people have been known to build cons and fit them to mesh after it built, This techn cause a problem when scaling the bones to fit the mesh. your skeleton inside 2, Start with the legs. Draw a 3. Copy the first leg to make character’: mesh, using simple IK chain with two second. Align the second to as a reference, Joints. The mesh has been the mesh. removed here, but make sure the knee of the chain aligns with the knee of the mesh. “two bones to create a5, When viewed with the 6. Draw three bones to create a Make the pelvis the mesh, the pelvis and legs spine. The pelvis should of the legs. should look something be the parent of the lowest like this. vertebrae. pter 6: skeletons and mesh deformation ca7 10. digital character animation 2 Ac the top of the spine, cre- 8, Draw an arm and link it to 9. When viewed inside the ate two bones for the shoul- the shoulder. Copy this to mesh, the upper body shou ders, Make the top of the make the other arm, look something like this, spine the parent of the Make sure the shoulders shoulders. attach to the arms above th armpit, not inside the ton Add feet to the body. These «Ifyou are using IK, set up When the constraint is can be single joints or, to the constraints. Create a null vated, the leg should snap t better articulate the foot, two or dummy object near the the object. Moving the joints per foot. right foot. You can also we object will now manipulate other objects if you desire, the leg. Move the leg back such as text of a nonrender- a neutral position. Repeat ing box. Constrain the ankle the procedure for the othe to the object (arrow). leg. If you want to use IK 0 the arms, repeat the proce dare for the arms as wellchapter 6: When all che constraints are 14. in place, manipulating the body is accomplished by ng five objects: the hips, the leg constraints, and the constraints. One prob- mm with this setup is that, to he scene, you need to move I five objects. skeletons and mesh deformation final skeleton should fit nicely within the body, ‘This can be overcome by adding one more null oF dummy object and making it the parent of the hips, leg constraints, and arm con straints. This keeps the legs, arms, and hips at the same level, while keeping all the character's parts within a single hierarchy.1. Start with a bone stretching. from the base of the palm to the middle of the first knuckle, Now make the thumb, Draw a bone that stretches from the base of the palm to the middle of the thumb’s knuckle, digital character animation 2 Hand Skeleton Skeletons for hands are reasonably straightforward. It is simply a 1 of creating one bone per finger bone, plus an extra set of bones for palm. As with a skeleton for the body, its skeleton inside the hand mesh, ways best to build the hi 2. Create three bones for the 3. Repeat the procedure for first finger. Align the bones rest of the fingers. so they run through the cen ter of the finger and the joints line up with finger joints 5, Then draw the two joints for 6. The final skeleton can be the thumb itself. The skele- attached to the arm of your ton is now done. The hand character through a simple skeleton is attached to the hierarchical link. \wrist simply though a hierar- chical link.chapter 6: skeletons and mesh deformation lesh Deformation a skeleton is built, it can be used to deform the skin of your char- t-This is done through a mesh deformation utility, which uses the esition of the bones to determine the shape of the mesh. As the bones the skeleton are animated, the skin of the character also animates. this is done depends on your sofware and its capabilities. Most ges use a. combination of methods, but they fill into the general tegories of direct assignment and envelopes. his method, each vertex in the model is assigned to its corresponding ‘or joint. The vertices in the thigh are assigned to the thigh bone, se in the shin assigned to the shin bone, and so forth. For simple dels with few vertices, this method is quite easy and predictable the calculations are easy, deformation is usually reasonably fast, Jing the character easy to manipulate in near-real-time, Most soft- me does this automatically, assigning each vertex to its closest joint, igh some packages may force you to do the assignment manually ‘cleanup anyway to manage vertices assigned to the wrong joint. other way to accomplish this task is to create clusters for each part of body (chat is, a cluster for the lefi thigh, another for the right thigh, [so forth) and assign each cluster to a joint. This is done either ough a simple hierarchical link or via a constraint or expression. sure if you do this that all the joints and/or bones are named ily, because the clusters will be named that way too, and it ‘be confusing if you do not name them right in the first place. ‘one problem with direct assignment is that it will not produce ‘organic results. This is because most parts of the body are affected than one joint, For example, the knee is affected by both the ind the thigh. Directly assigning those joints may cause tears or ifthey are moved to extremes. Still, or basic characters, direct gument is simple and works quite well.me digital character animation 2 Direct assignment links the vertices in the mesh directly to. _-When the forearm moves, the mesh deforms to match the skeleton, In this example, the vertices in the forearm its motion, (red) are assigned to the forearm bone. The vertices in the Upper arm (green) are assigned to the upper arm bone. When the upper arm moves, it moves its vertices. Because the forearm is a child of the upper arm, it moves to0, as da its vertices, The big problem with direct assignment occurs when joints move to their extremes. In this case, the motion the forearm causes the vertices surrounding the elbow severely crease, Weighted Assignment/Envelopes For a more organic deformation, you may need a vertex to be aff by more than one joint. The vertices around the elbow, for example, be affected by both the upper and lower arm. This brings in the co of a weighted deformation.chapter 6: skeletons and mesh deformation jope is basically @ range of influence, Where the ‘When the arm bends, the vertices around the elbow are es overlap, the vertices are affected accordingly. weighted according to their envelopes, which eliminates the crease. ghted deformation enables more than one joint to affect a given ex. How this is done alo may vary, depending on the sofware, Many have a visual reference, typically known as an envelope. An enve- is basically a range of influence. Where the envelopes overlap, the are affected accordingly. 1e software also enables numeric weighting of clusters. A given clustes, ple, can be affected 70% by joint A and 30% by joint B. In , this is similar to envelopes, but can be more flexible, because the ex asignments are not constricted to a specific area. You could theoret~ ‘sign any vertex to any bone with any weight. Of course, without visual reference, it is certainly a bit more difficult to set up. r method is much less visual than envelopes, which are displayed nubblelike icons that can show what is being included in the deforma- iny packages overcome this by coloring the vertices acco This can act as a visual reference, easing setup. dvanced envelope systems enable you to move the control points envelope’ bubble to include or exclude the stray vertices. The one that might crop up with simpler implementations of an envelope system is that you may not have the capability to reshape the enve- is gives you no way to deal with stray vertices. A good example ‘be the fingers in a hand—the envelopes for one finger might over- wd affect che vertices in another. If your sofiware doesn’t allow you omize envelopes, you can avoid these sorts of problems with clever You might model the fingers farther apart than in the normal hand to keep the envelopes from overlapping, and then move the imo a relaxed position afier the deformation is applied.The envelopes in these legs overlap slightly at the knees, which is good, but this also causes the right leg’ envelopes to overlap the lets. digital When the thigh rotates out, the ver- tices from one leg are stl affected by the other. “Another common problem is the area ‘underneath the arms. Typically, vertices in this region get assigned to the clos- est joint, inthis case, the upper atm, This causes a teltale bulge as the arm character animation 2 To fix this, either the envelopes bbe reshaped, or vertices on one need to be excluded from the ot é ‘The vertices under the arm are part the torso and are affected more by spine than by the arm, Reassigning these vertices to the spine helps the problem and eliminates the is raised above the head, because these vertices rotate out from the body with the arm, Advanced Deformation Tools One step beyond weighted envelopes are tools that enable you to « trol the shape of the envelope, based upon the position and angle of Joint. These tools are aimed at simulating advanced effects, such as bulging muscles and stretching skin. These tools are great for gaining extra edge when it comes to realism, and can also be used for other types of effects. They are not, however, absolutely essential when it comes to learning the art of animation.chapter 6: skeletons and mesh deformation vanced deformation systems enable you to define how As the skeleton bends the arm, the biceps bulge like shape changes as the angle of the joint changes. Here, _real muscles, have an arm with lots of muscles. eformation and Surface Type ly, most mesh deformation tools work the same regardless of ie geometry type. The tool simply takes control of the vertices on ‘model, moving them along with the joie to deform the character. ier those vertices belong to a polygonal, patch, or other type of fice really doesn't matter much to the software, Polygonal and patch dels, however, do have a few unique issues that need to be addressed. forming Polygons nv-resolution polygonal surfaces deform quite easily, but do not good when rendered. Conversely, a high-resolution polygonal renders well, but can be a real pain to deform. As pointed in Chapter 2," Modeling Basics,” the solution is to deform a low- tion version of your character, and then subdivide the surface er it has been deformed, rendering the high-resolution model. If sr software supports this type of functionality, itis simply a macter ling up che resolution before the animation is rendered. your sofiware does not support subdivision after defarmation, the tion will be different. In this case, the animator will manipulate and fa low-resolution stand-in model. When the shot is complete, high-resolution model is swapped in, the animation from the low ter applied, and the shot rendered.‘A low-resolution model is deformed, digital character animation 2 The model is then subdivided after the deformation, keeps the model easy to animate, and it makes it when rendered This is where enveloping comes in quite handy. If the high-resolut model is simply a smoothed version of the low-resolution model, envelopes applied to the low-res model should affect the high-res pretty much the same, This enables you to use the same skeleton and envelopes on both models, When render-time comes around, you si ply hide the low-res model and unhide the high-res model. Of course, theory and real life never quite match up, so you will cer tainly have many headaches when using this method. Most often, the high-res model will not deform exactly like the low-res model, caus problems. Ifthe hands dont deform exactly the same, it may mean difference between holding a drink and having the fingers pass dl the walls of the glass. The only way around these problems is to thoroughly test the setup before it goes into production. This is absolutely the most important part of your success. Test the character against a skeleton animated through a wide range of motions. This way, any problems will show up and be able to be managed before animation begins. Ifthe low-res model is as close to the high-res in deformation as you can possibly it, you will spend less time working on damage control. Some high; productions have been known to have animators move vertices on a frame-by-frame basis to clean up these sorts of mishaps. This is not, by any means, a pleasant way to spend a week's worth of workhapter 6: skeletons and mesh deformation ing Patches patches define surfaces of variable resolution, you do not need divide their surfaces. This means that a patch surface can be ren~ directly without fear of jaggies or artifacts. Most patch surfaces art rly light, so deformation is usually reasonably fast. has a lot of detail modeled into its surface, however, a wed character can get as complex as a polygonal surface. In this ou again may need to have two different models: a low-res one ion and a high-res one for rendering. If you must do this, the olved are exactly the sime as a polygonal model, so refer to the section. BS-based patch surface, you will find that the most prob- with che blends, Blends are a calculated surface that automati- ges shape to fill in the gap benween two other surfaces. If 0 other surfaces exhibit a great deal of motion, the blend may ‘or cause kinks, Ir always best to keep the blend outside the ange of motion for a joint. As mentioned in Chapter 3, ling Basics for Animation,” don’t put the blend right on the - Instead, place it on the torso to lessen the strain on the blend. ther thing that must be mentioned is that blends take quite a bit of power to calculate, It is always a good idea to hide your es while animating. Wt wil eliminate these costly calculations ced up your interaction with your character. ting a character with NURBS blends, hide the increased speed, ‘When the animation is finished, unhide the blendsa digital character anima Skeletons and Segmented Characters If your character is made of segments, or partially segmented, you sill can use a skeleton to manipulate it. The individual segments can be hierarchically linked to the corresponding joint in the skeleton, ‘Manipulating the skeleton will then move the joints This technique is redundant for joints manipulated using forward kit imaties, because they can just as easily be manipulated directly. For thi Joints that need inverse kinematics, however, this technique can wotk quite well Other Deformation Methods Skeletons are not the only way to deform a character. There are plenty of other options available to the animator. Most of these methods are used to deform parts of a character, but they can also be used for enti characters, if necessary. Lattices ‘One of the more popular tools is the lattice. A lattice is simply an array of poinss, usually box-shaped, that surround a character or a part of a character. Moving the points on the lattice will, in turn, move the region of the character surrounded by the lattice. Lattices have many uses. As you saw in Chapter 5,"Modeling Heads for Animation,” lattices can be used as an aid to facial animation, partie= ularly for cartoon eyes. They also can be used for things like muscle bulging, For simple characters, such as a flour sack, the lattice may be all you need to deform a character. A lattice is placed around this oylinder. When the lattice is moved, the vertices on the cylinder fllchapter 6: skeletons and mesh deformation tt an also be used for animation. This flour sack is Animating the shape of the lattice animates the shape of fed using a simple lattice. the flour sack are another way to animate the shape of your character. In its plest sense, a cluster is a collection of vertices. This enables you to ontrol many vertices with just one cluster. More complex systems you to weight vertices between several clusters, which can give ‘very sophisticated deforma isters are used in a number of places. A cluster linked to a skeleton Abe used as a skeletal deformation system, Clusters can be used (0 effects such as muscle bulging. Some people use clusters to ntrol parts of the face—for example, the eyebrows oF the tongue— ficial animation. line Deformation deformation enables a simple spline to control the shape of an ect. Animating the vertices of the controlling spline can animate the {One good example where this technique might be wsed isin an tail. Another place it might be used is in a character who has142 digital character animation z The spline just above this cylinder is used to control Move one vertex of the spline, and the cider changes its shape, shape to match, Morphing Morphing enables one object to assume the shape of another. Morpl is most commonly used in the area of facial animation, Because the is such a complex shape, it is best fo sculpt the many individual poses character's fice may take. This enables precise control of the specific shapes.A morphing utility enables you to smoothly animate between these shapes. Muliple target morphing enables multiple shapes to be mixed and blended to create new shapes. For facial his enables even more control over the shape af the fa © extremes can be mixed and blended in a yy ratio or combination. Facial setup and ani tion is covered in depth later in the book.chapter 6: skeletons and mesh deformation clusion can see, skeletons and mesh deformation can help you bring a Imesh character to life. Although many packages have tools that are in the way they work, there are many package-specific differ- that you need to consider when deforming your own character. pmeans cracking open the manual for the litte tricks that are 10 your chosen software just learning animation, these cools can get very complex, 50 to get too bogged down in the fine-tuning of your character’ tions. If your character is overly complex, you can spend more seting up a character than you do animating it. It is better for the wer to create a simple character that deforms easly rather than a character that is difficult to control rememiber that the goal of a good setup is to make the character animate. If your character is hard to animate, you will take much to Jearn the craft, The best idea is to invest the time needed for up to make your job easier when you animate. You don't want t0 j-taping” your animation together because it wasn’t set up right first place.FES boLSULU po UU ELeYe|Posing Characters ow that you've built a few characters, you can start on the fundamentals of animation, Before you actually move your characters, however, you need to understand where and how they will be moving, Specifically, you need to understand how your characters are going to be posed—exactly how they will stand, sit, or walk in relation to the camera. A pose is a fundamental building block of animation, and many animators block out a character's major poses while animating, Posing determines, to a great extent, how your audiences perceive the characters and their emotions. Intimately involved with posing is the concept of staging. This determines where your characters are in relation to the camera, their environment, and each other. Again, these relationships determine exactly how your animation plays to the audience. Until you under stand how to place your characters properly in font of the camera it makes no sense to move them. So, lets look at the basies of posidigital character animation 2 Posing A pose is simply the way a character presents itself to the camera. If character is sad, happy, frightened or brave, you should be able to read that emotion in us pose—the way che character stands, where its han are placed, the position of its head. Every part of the body has a role creating the pose. Theater, dance, mime, and countless other performing arts involve posing to a great extent. Similarly, animation is another ar form that relies heavily on clear poses and silhouettes t0 convey a message. ‘When animated, your characters need to show emotion. Even the simplest shots require a character to hie 4 strong pose. Whether the character is sad, happy, proud, or surprised, the emotion shows in the body and the pose, ‘A simple character can show a great deal of emotion simply through the pose of the body. Animating with Poses In the golden age of animation, anim basic methods for animating a scene: straight-ahead and pose-to-pose, Each method uses posing differently and has its own place and its own advantages ‘ors discovered there are two Pose-to-Pose Animation Pose-to-pose animation is the more controlled of the two. In this ‘method, you plan your shot and get the main poses of the character within the shot blocked out. If your character is standing up from a chair, for example, the poses may be leaning back, grabbing the armrest for support, leaning forward, and, finally, standing up.A character winning the lottery may read the lortery ticket and then show disbelief, 'chapter 7: posing characters re ck, and joy. The theory is that every action can be broken down basseries of distinct poses. From there, its a matter of creating in- or letting the computer in-between the poses for you —and, tweaking these as necessary It is a good way to animate and tightly choreographed shots, as well as slower subtle lime moves, because those do not have many surprises or quick to them, Pose-to-pose animation is also good for dialogue, each pose can be fit to the mayor points in the dialogue track mall of the method is that it may lack spontaneity light-Ahead Animation ght-ahead animation is pretty much what the name implies. In method, you start on fame one and animate “straight ahead” fro “This method is more improvisational in nature and can sometimes 9 very spontancous and complex motion. It's also a great method icker action motions because of the spontaneity. Straight-ahead is thod closest to “acting” a frame at a time, and ie is very sitnilar to echniques used in stop-motion animation. If you are trying to eve a stop-motion look and feel to your animation, this is certainly to go. Still, this method can make it hard to achieve well- xd and solid poses, sometimes making animation that is hard to "It also makes animating complex shots difficult mmbining the Two Methods 1, however, combine the two methods and get the best of both. puter animation gives you the bonus of being able to do this rather j Most fast machines can play back an animation test almost antly. This makes it easy for the animator to block out a series of rather quickly, almost in a straight-ahead fashion, or animate a at a time in those sections that might need more spontaneity. je question that still remains, however, is one of thought process. How approach animating your scene? Do you plan your shot carefully o-pose) or do you improvise (straight-ahead)? This is not an easy stion to answer, and the best advice is to use your intuition and ence. Overplanning a shot may very well sap the life out of it. ‘more improvisational can add unexpected touches and details you uid have never dreamed of. On the other hand, complex shots need plinned very carefully or all the elements simply won't sync up. en you combine the two methods, you plan the extreme poses and tighten that up with straight-ahead interpretation between those es The first part lays the groundwork, and the second part spices itA simple flour sack is « good character to pose. digital character animation 2 up and gives it life, leaving behind all the computer interpolation that makes an animation look mechanical. With a computer, you can exsily use both. | highly recommend you use both techniques to get a good animation; if you rely on just one or the other, you easily see where their weaknesses are, Posing a Flour Sack One easy character to use to begin studying posing is a simple flour sack. This litle character has become a classic tool for the study of animation. The flour sack is simple to construct, but it has volume and. weight and can be easily posed, How a flour sack relates to a human body may be a bit confusing to the newcomer. The easiest way to conceptualize this connection is that each corner of the sack represents the arms and legs of the character. course, these are very stubby arms and legs. Another way to think of character is that the corners of the sack are the hips and shoulders of the character. This is a bit more accurate and leads you to a better understanding of how the body works. Almost all motion in the body starts at the hips and then translates to the limbs. This is also how you should pose characters—from the center of the body outward. If you pose the hips, spine, and shoulders properly, the arms and legs will be much easier to position naturally. This is because almost all motion begins at the hips and continues down the tree of your skeleton. The only time this changes is when a character is reacting, from being hit or pulled with or by something. When the characteris interacting with something else, this rule changes to lead with whatever part of the body is being hit or pulled Ifthe corners ofthe sack are stretched, If separate arms and legs are added, you can see how they can be thought you Can see how the corners of the ‘of as arms and legs. sack can also be thought of as hips and shoulders.chapter 7: posing charact ing Up a Flour Sack Jour sack can be deformed using any number of methods. Some tse a simple method, such as a rectangular lactice. The points of fice are deformed, which reshapes the sack into the desired pose. skeleton, without arms and legs, can also be used. A skeleton is particularly for learning, because it lends itself to an understanding he underlying anatomy. sack can be deformed using a simple lattice, ‘lo be deformed using a simple skeleton and a mesh deformation tool jing the Sack the sack is modeled and deformed, the task of posing the sack is 4 matter of manipulating the skeleton, lattice, or other type of mer to get the pose you want. When posing the sack, try to make ‘poses convey emotion and meaning.ry digital character animation 2 On the CD is a flour sack mode! ready to be deformed. Set it up in your favorite package and use it to create a dozen or so interesting poses. As inspiration, here's a few to get you started: 6 Posing Characters Now that you've had a bit of practice posing a simple character, you & move to a more complex character, such as a human body: When posin a human body, you need to understand anatomy, the skeleton, and how they all tie together, (Much of this was covered in Chapter 3,"Modeli Bodies for Animation.”) This knowledge will help you create strong. poses. Strong poses are one of the fundamental building blocks of animation. They're read well the instant they're viewed. If your poses are strong, the auidience knows exactly what's happet stand your character’ actions better. Also, a good pose is almost always balanced, giving your digital characters the distinct feeling of weight and life ng and will under Balance Keeping a pose balanced is very important. The body is simply a system of joints that is trying to stay in balance. Each bone acts as a tiny levchapter ing the weight of the body through the spine to the hips, and down through the legs to the ground. If a pose is out of balance, acter will appear as if its sbout to tip over. If the body is slightly if balance, the eye will still pick it up and see the pose as wrong Wan apparent reason. gisan unnatural pose. Rarely does Placing the weight on one leg, however, Twisting the spine so the shoulders are test equally or both feet forces the hips to lean. This puts the turned opposite to the hips places the body out of balance, making it lok like bady in balance. its about to tip over: human body, all balance starts with the hips. The spine and the body rest on the hips. and the hips rest on the legs. Any forces wed by the legs reach the upper body through the hips. The hips alto close to the center of gravity of the body and thus are close to center of most motions, ase everything stemts from the hips, they're the best places to start posing a character. In a relaxed stance, the body usually rests on Teg, not both. If you've ever watched people waiting in a long line, know what [ mean. People constantly shift their weight from one to another as they wait. Rarely do they place their weight equally both feet, except when standing at attention. As the poses get more , you need to pay attention to balance very closely. Ifa character up an object, for example, the object’s weight should be balanced ithe rest of the body. When a person rests on one leg, it throws the le system off center. When the weight is on one leg, the free leg the hip down and out of balance. This, in turn, curves the spine forces the shoulders in the oppesite direction to maintain balance. body is about to move forward, the shoulder may drop on the side as the hip. Either way, the body’s natural sate of balance is trical, posing characters has‘This pose is symmetrical in almost every This pose breaks symmetry in a number And so does this one. Both of them respect. Its also boring in almost every respect. This character is locked in a very symmetrical pose. This is Rotating the hips forces the spine to twist to maintain full of twins, such as the legs and arms, and itis not very balance and makes the pose asymmetrical. This makes the interesting, al character animation 2 symmetry Because the body is so syinmetrical, itis very tempting to place it in symmetrical poses. Not only is this dull and boring, itis very unnat To keep your characters looking natural, you need to keep them asy metrical in almost every way—from the positions of the eyes, hands, and feet to the motions and actions that they do, Symmetry has this odd habit of creeping in at the worst times. Avoiding symmetry means avoiding what animators call twins. A nui simply a part of the body that mirrors another. Even minor details ina pose—suich as both feet pointing in the same direction—can make a character look strange. of places. more interesting and look more nat character look more natural+ point to consider is the weight of the character ay well as the ght of everything else in the scene. Consider a character standing the edge of a diving board. A heavy character will bend the diving and considerably, and a tiny character will hardly bend the board at A character lifting a heavy suitcase will need to lean its body avsay om the suitcase to get it off the ground. heavy isthe ba? Until the This bag is very light You can tell this This bag is very heavy. Again, the pose elif it, we do not know. by the pose es us the information. nother important point to consider is the weight of the character itself, ny characters have no problem supporting their own weight, In a hurricane, however, they may rend to blow away. Heavier characters, a the other hand, will be more stable in hurricanes, but may have lems supporting their own weight when they are simply standing can cause the knees of a heavy character to bend outward to nt the extra weight. Be sure to take these factors into consideration hen posing your characters When animated. a heavy character moves much slower than a light mnacter and needs a lot of force to begin moving. Think of a dinosaur it needs momentum to get moving. But a mosquito is quick and light nd needs barely any momentum to get moving, rust still be presented to the fence. You must always remember that your 3D animated creation is st always going to be shown on a 2D screen. The screen is your , with the audience viewing everything through the camera. As chapter 7: posing characters SaPr digital character animation 2 such, your characters must present their actions clearly to the camera that the audience can read and understand them well Probably the best way to pose characters for the camera is by studying methods pioneered by magicians, mimes, and stage actors over the centuries. Good silhouettes and a strong line of action help present your characters to the camera properly Creating Strong Silhouettes ‘One of the most important fundamentals is presenting a strong, silhouette to the audience—or, in our case, the camera. ‘The human eye first picks out the silhouette of an object and then fills in the rest of the detail, Ifthe action is presented so that its outline is clear, the action will be much clearer as well Think of a magician pulling a rabbit our of hat. Hee always pulls the rabbit out to his right o his let. This way. even the person in the back row can understand what going on. If the magician wants to hide something, he'll usually do it when his hands are passing in front of his body. The body, along with the motions, serves to conceal the action: The same principles apply to animation. Animacors usually don't have @ worry about performing sleight of hand, but they're very concerned about making actions read. The silhouette is the key A good sithouette wil tell fou what this character is doing he's about to pull something from his hat. If the silhouette reads, so does the pose.fp check out your character's silhouette, simply pose your character the computer. Next, apply a matte black texture to the character. nder che pose, You'll have nothing but the silhouette. Another way {g0 about itis to simply look at the alpha channel matte used to posite the character into the shot—it's always the silhouette nn you have this wwe, ask yourself, “What is this character doing?” Tfyour silhouette is clear, the action and the pose will read bf {A gp0d example of sihouetting is this kick. The image reads well both as an image and in silhouette , f yn the camera 90 degrees, however, and nothing reads. The angle from which you view your character is as important as pose ise ine of Action addition to a strong silhouette, a good pose should also have a inte line of action. This is a strong line that you can follow from ir character's fect to the tip of its fin also adds beauty to the pose. The human eye is surally drawn to a good line. 1s, This not on chapter 7: posing characters czdigital character animation 2 For example, if your character is throwing a ball, arch the back and make the arm follow along the same are. This gives a much more pleasing line than if the arm just moves back. If you put the whole body into the throw, it looks more convincing, 1 This pose is weak because it has a weak line of action, Getting the characters whole body into the action creates stronger line of action, as well as a stronger pose. If a character is in a tug-of-war, he digs his heels into the ground and arches his back, putting every muscle he can into the effort. Even simy actions should follow a definite line, Ifa character is proud, he arches his back and throws out his chest. Ihe’ tired, he slumps over and has a completely different arc. x This tug-of-war pose shows a great line of action. If a characteris tired, he slumps over in a different line of action.chapter 7: posing characters 157 Staging Multiple Characters Not only does the animator need to know where to place the charac in relation to the camera, but also where to place the characters in tion to one another. How do you make them interact and read? i solution depends on a study of acting and interpersonal relation- ips. Characters who are friendly can be staged closer than characters are enemies. Every situation has its own demands. For example, a rine drill sergeant intimidates a recruit by yelling at him a few slim ¢s from his face. Lovers who are angry at each other turn away and id practically back to back. v you place the characters depends on the dramatic requirements of scene as well. If you want to highlight 2 character’ facial expression, that character so the camera is closer to his face. Similarly, 110 make a character seem distant and alone, place him off in the snce. Placing characters at different distances from the camera also your shot more depth and makes your film more dynamic. These stagings might work for a simple conversation, To get more depth or to highlight ‘one character over the other, different stagings may be required158 | digital character anima ion 2 ‘When you have more than wo characters in a scene, use the comp tion to direct the eye to what’ important, Think of the characters as weights. If you put one character off to the side of the screen, you m need several toward the center of the shot to keep it balanced. Ifyou have too many characters on one side of the screen, the siot will lose ind look odd. If the shot requires the characters to be toward one side of the screen, other items, such as furniture or foliage, can be balan used to help balance the composition, Characters, particularly crowds, can also be used as a way to frame ber that the composition affect how the audience interprets the scene, For example, think of those what's important in the shot. Re times where a crowd parts to reveal the hero, or those shots where a gang of people surrounds the villain. A crowd is a shape that can be animated in its own right. Use this tool in your compositions as wel Think of your characters as weights. In this poorly staged To balance things out, bring the isolated character desert shot, the three characters on the left tit the weight of the _the camera to make him bigger and give him more weight composition to the left This makes the shot more pleasing, Characters can also be used to frame the action. Here, the Looking a bit deeper, notice how the heads of the charac characters in the crowd afe used to frame the character inform a “V" shape, which leads the eye directly to the cent the distance. character. The shape of the crowd is a handy composition toolStaging and Camera Placement Proper camera placement is important for anyone involved in film, Knowing where to place the characters in relation to other characters, camera, and the environment is essential to achieve the maximum Mdamatic effect Plicing the character in just the right place at the right Mie isan art that every director needs to master. A good example Upper camera placement is a simple example, What if you have a Stharacter dangling from a rope over a deep crevasse? Hou a the camera from the side, the audience will not Pointing the camere down int the pit reveals the danger Jw that the characteris in danger other important factor in staging is how the camera position affects perception of the shot. When the camera is placed low and is shoor- Up at a character, it makes him seem more menacing, Think of any WW5Vs sci-fi movie where the giant insect is menacing a Us.and you'll get the idea. Conversely, keeping the camera high makes 2ior metropo- Sducters seem helpless and insignificant, much like ants when viewed hom above. ‘Moving the Camera Computer animation makes it extremely easy to move the camera pmputer animators probably move their cameras fir more often than throughs, this may Bhe great; but for character animators, this can spell instant disaster— sary. For spaceship flybys and architectural wal pecially when you are trying to present an action, feeling, or emotion Dthe camera. When the camera moves, the motion draws attention yy fiom the actors and toward the camera. Because moving the Ginera effectively changes the pose, it also forces you to rethink your every time the camera moves chapter 7: posing characters nzdigital character animation 2 Still, moving the camera can be very effective when it’s done at the right time and for the right reasons. Panning with a running charact gives the feeling of motion. A rack focus blurs the background or f ground, giving strong emphasis to another part of the scene. You may want to move in the camera quickly in a moment of fear. These all simple wicks that live-action directors have used for decades. Rem: the most important thing: As with motion blur, you should never n it. The cameva should embellish the scene, but never move so much it becomes a character in itself, In a digital environment, you can easily place several cameras in a scene (© get the equivalent of the live-action, three-camera shoot. Us fone lens for the establishing shor, one for the action, and one for the close-up. Even in 3D, lenses tend to have different effects. Many live= action directors shoot close-ups with a longer lens, because it flattens the perspective and focuses attention on the actor. On a panoramic shot, the director may use a shorter, wide-angle lens to get as much, of the landscape as possible into the shot The same goes for 3D. Instead of just using one lens and moving the camera, change the field of view to get different effects. As with live action, try to keep the lens the same for the duration of a shot, and change lenses only after a cut. Changing the field of view within a is typically not done in live action character animation, either. Again, do not draw attention to the cam and it is not a good idea for most For those whose 3D shots are composited with live action, knowledge of the lenses used in the live-aetion shoot is critical. For animators wi use software without stock lenses, the lens length is equivalent to the angle of the camera’ field of view (FOV). The wider the FOV, the shorter the lens. These values are shown in Table 7.1. Table 7.1. Conversion between the Lens Length and the Camera Field of View Field of view Type of lens 115.0 degrees Ultra Wide Angle 94.286 degrees Very Wide Angle 28mm 76.364 degrees Wide Angle 35 mm 63.0 degrees Medium Wide Angle 46,0 degrees Normal/Standard 28.0 degrees Medium Longchapter 7: Lens Field of view Type of lens posing characters 135 mm 18.0 degrees Long/Telephoto 200 mm 12.0 degrees Extra Long/Super Here’ an e: same pose rendered through three lenses: 15mm, 50mm, 135mm. The longer the lens, the tighter the angle and further back the camera needs to be to frame the shot Jn lens eliminates the fsh-eye effect and causes the Pe to recede a bit. The SOmm lens is a good all lens ple of how lens choice affects the composition of a shot ‘Through a 15mm lens, the bear is distorted for a fish-eye effect. Also notice how the landscape is included in the shot. A lens this short might not be so good for close-ups, but its great for panoramic shots, A.135mm lens flattens the character even more, practically eliminating the background. The 135mm lens is great for close-ups.: digital character animation 2 Conclusion This chapter focused on the importance of posing and staging your characters. Remember to keep your poses natural and balanced. Avo symmetry in your poses and also be sure to keep a strong line of. and silhouette. You must akways keep in mind thar your audience sees your animation only through the caetera, so be sure that you place camera properly and ase the right lenses. As we get into animating characters in the next chapter, keep these concepts firmly in mind, ‘The best animation will never come across if the camera doesn't see it properly.TCT ese sal ee OR rs PelilThe Basics of Character Animation Ye are now finally ready to begin animation. As you have seen, animation is more than just moving things around. It involves proper construction of your characters, as well as propet staging and the ability to pose your characters naturally. Now it’ time to delve inta the nitty-gritty of motion and timing, the heart of animation. Understanding Newton's laws of motion is certainly one component of the equation. Watching and observing the way people and animals move ant comedies, oF clasic is another. Watching references—nature films, cartoons—is yet another way to understand motion. It$ also important 1o familiarize yourself with basic animation concepts, such 3s squash- and-stretch, anticipation, avershoor, follow-through, and many others. In addition to this, time is a precious quantity that must be used wisely ‘Timing is a raw material that actors, comedians, and musicians use constantly. A comedian with good comic timing knows exactly when to spring the punch line. An animator who has good timing knows exactly when a character should react, blink, or pull chat huge mallet out from behind its back. Timing is the only thing that separates animation from illustration. With proper timing, your characters will appear to live. Without it, they'll look like nothing more than inanimate puppets.rz digital character animation 2 The Importance of Timing Timing affects every aspect ofa film, and on many levels. First, the film is a specific length—from a 30-second commercial to a 2-hour feat Second, the cutting of the scenes within this time constraint affects the mood and pace of the film. Third, the acting and timing of the character's actions affect how each individual scene plays. ‘Think of your film as music. Both mediums rely intimately on time, The film is your entire work, much like the song or symphony isa composer's entire work. Your film’s scenes an be s choruses, or movements. The individual notes of the instruments are the same as the individual actions of your characters. Each action, a in each musical note, must be in the right place at the right time. As in music, bad ciming in animation sticks out like a sore thumb. The Process of Timing Timing a film starts at che length of the film itself. A commercial may be only 30 seconds, but a feature will be over an hour in length, Wh this length is known, a storyboard and a script are prepared. The dialogue is then recorded. Dialogue is a big factor in determining how Jong an animation will take, so use ie frugally. Long stretches of dialo can eat up time very quickly. A slow-talking character can also waste vast amounts of time. In any event, afer the dialogue is recorded, an animatic or leica ree! i created to rough out che timing of the whole film, This is accompli by scanning the storyboard panels into the computer and timing them to the dialogue tracks, using a video editing program such as Adobe Premiere. The goal is to make the film the desired length and to block out the timing for the individual scenes. After the timing of the shots is blocked out, each shot can then be broken down into the specific actions a character will perform. How this is done depends on the production. Most of the larger, traditional productions use exposure sheets (also known as X-Sheets): most of the newer all-CG shops leave the timing to the animator.hapter 8: the basics of character animation rames-per-Second and Timing hen you are breaking down a shot for animation, the first isue is the me base, or frames per second, that the production will use. For the ist half-century of animation, this issue was simple. All animation was thot to film, and all film ran at 24 frames-per-second. proliferation of video, CD-ROM, and other media has clouded this n the USS., video runs at 30fps; in Europe, it runs at 25fps, and video games are animated at only 15fps. With so many time bases being thrown at the animator, it can be sult to learn how many frames a certain action will take. It used p be that an experienced film animator could tell that a particular jon would take a specific number of frames at 24fps. To transfer this Jedge to video, for example, the filot animator multiplied this time 1.2 to get a correct time (24fps x 1.2 = 30fps). Doing this sort of nati in your head can be awkward, at best With this in mind, what time base do you use for learning your timing The best advice is to simply learn to time at both 244ps and 30fps, and then you can svitch-hit as the project demands. Most high-end productions are still film-based, so this skill will serve you well in this context, Although many animators may work in video or game time 1€5, 30fps is the proper choice, A good animator’s stopwatch might ko be a good tool to own, because it can convert to both 24 and Timing Using Exposure Sheets n exposure sheet is simply a sheet of paper with an array of rows and columns. Each horizontal row represents a single fame of animation. tical rows contain information, such as the contents of the dialogue trick (explained in Chapter 10, “Facial Animation and Dialogue”), eamera instructions, and a margin for the director to plan the timing the shot. This can be done via sketches to explain a character’ poses, is well as lines to indicate when the actions start and stop.‘An exposure sheet can be used to time ‘animation, Each horizontal row repre sents one frame of animation, digital character animation 2 TERRE eee eer Timing on the Computer Some productions do not use exposure sheets. If this is the case, tim is done directly on the computer by the animator (with the director giving feedback). Most high-end 3D packages allow the dialogue to be loaded in with the scene data, On a fast machine, this gives the animator the ability to scrub the dialogue in real time, reducing the need for the track to be read manually on an exposure sheet Some productions do what is called an animatic, where blocks are placed in the scene and moved about to get a basic idea of how the scene will play. The director then approves the animatic, and the animator uses the animatic as a template to set up the scene and gat the basic timing. It is fairly easy to accomplish, and when played agai dialogue, it serves the purpose of the exposure sheet quite well. How Much Time? This is the number-one question on every animator’s mind: How mix time? Time is a very tricky substance. Too much time makes your film slow and boring. Too little time makes your film fase and unclear. As develop your skills, experience will hone your sense of timing. Trial error is also a good tool. If you're not sure how long a particular acti will take, guess. You can always add frames if the motion is too fast or subtract fiames if it’s too slow. Computers afford the animator the abil to change timings very quickly.apter 8: the basics of character animation Thing at a Time thing at a time” is an important rule in timing for animation. The perceives things best sequentially, so you should present your actions that way—one at a time within a smooth sequence of ‘A character stubs his toe, recoils, and then reacts. Ifthe reaction quick, the audience won't have time to read it, so the recoil acts bridge between the two main actions. must remember that the audience is usually seeing your film for first time. As the animator, you need to guide the audience and tell viewers exactly where to look at each point in the film. Think of a winer cartoon. When Wile E, Coyote steps out over the gorge, he a while to notice hes suspended in mid-air. He typically reacts and looks directly at the camera with a pitiful expression. This one ing can be held for almost a second, He may blink or his whiskers ‘witch, but it is esentially a still pose. He then zips off the screen in few frames. This is a great example, because in the span of time that pose is held, the aud te—he’s doomed. It also draws the audience's full attentio coyote, so when he does fall offscreen, they see that clearly as well nce comes to the same realization as the to of the most important lessons you can learn about timing is to attention to what is about to move before it moves. An action acs only when the audience is fully focused on it. As the animator, must guide the audience’ eyes only through the character’ actions. veloping Good Timing A good sense of timing is something that animators develop through years of practice. Observing the world around you is the firse step in Ibis process. Breaking down these observations into their exact timing ‘isthe next. Many animators dissect classic animated and live-action films fiame by frame just to understand the timing. AVCR with a good ze-frame capability is an excellent investment for an animator “to make, You should dissect live action as much as cartoons, because all motion is based on live action. This is very important to understanding timing, ‘Cartoonish action is an exaggeration of live action. les sort of like studying life drawing first so you understand the human figure; then, and only then, should you go on to drawing cartoon stuf, or you ‘would never represent the figure correctly with so few lines. This applies to motion, too; if you understand how the body moves in real life, you ‘an sprinkle a little exaggeration onto that to get a cartoonish motion.digital character animation 2 Another essential tool isa stopwatch. Whenever you need to time a shot, the first ehing you should do is reach for your stopwatch and out the scene, timing the approximate length of each action. Plug.in these amounts as your rough-pose timing and then eweak it in your program. The stopwateh is one of the most important tools an anima ‘can own. It enables you to physically act out your scene and get a fi 200d idea of the timing, If you're shy, close the door and draw the blinds, but act out your shots, regardless. Timing and Software There are a number of ways to view and manipulate time within a program. These include curves, dope sheets, and paths. Each method ha its own strengths and weaknesses, and which tool you use depends on the needs of your shot. Some animators gravitate toward one set of over another and have heated debates with other animators over whi tool is best. Of course, the choice is personal, so Jet’ go over the main tools that most 3D programs offer. Curves Most good animation sofiware presents you with a graphical represe tion of how your objects are moving. Curves ate an invaluable tool fo the animator in diagnosing and fixing animation problems. Knowing how to read and manipulate animation curves is an essential skill. Ev software package is different, but most curves work in similar ways. ‘Typically, the horizontal axis of the graph represents time, and the vertical axis represents the parameter being changed—position, rotati scaling, and so on. These parameters are plotted graphically to tell you exactly how an object moves. ‘Take, for example, a character walking down the street. This motion can be represented as a graph, First, the character may be walking al the sidewalk at a steady rate. The character may slow to a stop at an intersection. When the intersection clears, the character accelerates agnin t0 a steady rate of speed ‘The motion graph would look something like the figure on the next page,ation ra chapter 8: the basics of character a ‘An animation curve of a car in trafic. DISTANCE ——>— TIME —————>- When the character is walking at a steady rate of speed, the graph is finear—a straight, diagonal line. When the character is stopped, the change is zer0, so the graph is flat. When the character is accelerating, decelerating, or changing its velocity in any way, the result is a curve. A ‘curve that slopes toward vertical represents acceleration, and a curve that slopes toward horizontal represents deceleration. In most software packages, you have three motion graphs for the object: ‘one apiece for the x, y,and 2 axes. This way, you can move an object finearly along one axis while giving it variable motion along another.digital character animation 2 In this software, red represents the x axis, green represents the y ans, and bive is the 2 axis, Notice that the blue line i fat indicating there is no motion along the 2 axis. Motion graphs are also good for finding and fixing annoying glitch your animation. Generally, animation problems show up as distinct spikes in a motion graph. These may be places where a key has been altered, or is in the wrong position. Another use is determining whe something is moving. If you want your character's feet to remain f 10 the floor, the foor’s motion graph will also remain flat at the level the floor. Here is an animation curve for this character's foot. When the curve is at 10, the foot is flat on the floor.chapter 8: the basics of character animation Spike in the curve indicates a sudden change that may When the curve dips below zero, so does the foot. problems in the final animation, iting Curves animation curves is very similar to editing curves used for g Each curve has a consol point where each keyfiame is cl. Most software will add Bézicr-style control handles on the curve ble you to change the curve's slope. You can also move the contol ts themselves to change the timing or the value of any particular tion to Bézier-style handles, many packages also give you other of default interpolations. A linear cane does not do any sat of -in or slow-out. A step ave looks like a square wave and jumps cone vale to the next, Most of these types of interpolations can derived using the Bézier handles as well. Bt cures are manipulated via Bézier linear curve has no slow in or oxt, A step curve simply pops from one and thus appears asa series of straight value tothe next, much ikea square lines. wave.rz digital character animation 2 Using Curves for Pose-to-Pose Animation Ifyou decide to animate a shot pose-to-pose, a good knowledge ‘curves can help you with the process. In cel animation, the poses drawn first, and then these drawings are timed to get what is called, pose test, It is much easier to get the timing of the major poses di before drawing the in-betweens. It also helps the animator focus on global timing issues without distraction, ‘The same paradigm can be used for computer animation. In this you block out your major poses on the time line using a step cu ‘This makes your character appear to “pop” from one pose to the Ifa linear or Bézier curve is used, the animation appears “floaty”, the timing is hard to read, Of course, with a character popping fi one pose to the next, the animation won't be realistic, but the idea concentrate on just the poses and the timing of the poses In this simple pose-to-pose animation, the character is standing, leans over to look at the ground, and stands up again.hapter 8: the basics of character animation ng with a step curve makes the character pop After the timing i set, simply copy the pose keys toa few the three poses. Tis isnot reatistic, but itis frames later in the time line ful in blocking out the pose timing 92 all the curves to Bezier. You now have the hold a pose, then smoothly hit the next pose, and the animation should look fairly real ‘the timing is set, you can in-between the poses, The idea here is old each pose for a brief period of time, and then make a transition next pose.A good rule of thumb is €0 allow 6-8 frames as a (which can change drastically depending on the needs of the The transitions are then set to Bézier curves. From there, its of tweaking the motion, offietting frames to get overlap, and Sheets pe sheet is a much simpler representation of an animation. A curve is ional representation of the motion (a value plotted against and a dope sheet boils this down to one dimnension, It is simply a containing marks or dots that denote where the motion changes. In us example, when the character starts walking, there's a mark dope sheet. When the character stops, there’s another mark, anddigital character animation 2 when the character changes direction, there's another mark. This is much like an animation curve, but without the information about value of the curve, ‘A dope sheet represents keys linearly. In this case (3D Studio Depending on the software, a dope sheet can also MAX), keys are represented as a series of dots represent keys as a series of bars, Ths is how it looks in Maya, ‘Many times, it's easier to visualize the animation as events in time than values. In addition, too many curves on a screen can get cont A dope sheet makes it easier to manipulate a large numbers of keyframes, Most dope sheets enable you to select and move dozens of keyframes simultaneously, which is helpful when retiming entire so Paths/Trajectory Another way of looking at your animation is chrough the use of paths, ‘Most packages display the path that your object takes through space. This isa good way of detersnining exactly how an object is moving. By adjusting the shape and position of the path, you can alter the way an object or character moves through the scene Paths are not used that much in character animation, but they can be helpful in places where you need to visualize an action. Ifyou had a character whose arms were manipulated via IK, for example, will see later, should move along an are. Another e character flying through a room. The path indicates the motion of the body in space. In addition to generating a path for keyframed motion, ‘many packages also enable you to draw a spline and use that line as the path, If you want your character to follow a straight line, for example, the path will be straight as well.chapter 8: path (blue) indicates the motion of this character's \Whem Ik isused, the arm follows the motion of the wrist. Using Timing to Suggest Weight ss your software calculates real-world physics, there is no way to dicate an object's weight. Think of a simple sphere sitting on the ind, as shown in the figure. Is it a bowling ball or a basketball? tl it moves, you have no idea whether the sphere is filled with air lead. Once the sphere is in motion, however, its characteristics ome apparent. bowling ball is heavy. It moves slowiy, and a great kal of force is required to change its direction. A basketball, on the ther hand, is relatively light. It moves fast, bounces easly, and it takes very little force to change its direction. Which sphere is filed with air and which one is filed with lead? Until they move, there is no way of knowing the basics of character animation Cai digital character animation = ‘One of the best ways to develop your sense of timing is to do some experiments with simple objects, such as spheres and boxes. These are great because they can be modeled and animated fairly quickly in any 3 package. By animating simple objects, we also ‘can focus on pure motion and timing. 1. Model a sphere, a box, and a ground plane. Position the sphere and the box on the ground plane with some distance berween them, 2. Animate the sphere so thac it travels in a straight line toward one face of the box. (3. When the sphere touches the box, change the direction of the sphere so that it moves off in the opposite direction. Keep the box absolutely still Render the shot.apter 8: the basics of character animation x does this example teach us? In both shots, the ball and the box identical; the only thing that changed is their motion. In the first iple, the ball appears light, and in the second, it appears heavy. As ‘can see, motion and timing affect our perception of an object’s ight. ie Language of Movement animator must also understand the language of movement. By T mean animation has a very specific vocabulary of motion that jmators ean draw from. This vocabulary includes such things 2s ares, icipation, overshoot, secondary motion, follow-through, overlap, and ing holds, among others. These motions are the raw material; good ing is the glue that holds ic all rogether. ‘This gives us a good example of the box being very heavy, with the ball being relatively light. Let's now reverse the process . Using the same setup 2s before, animate the sphere so that it travels in a straight line toward the box. . This time, when the ball strikes the box, keep the ball moving in a straight line Next, animate the box by rotating it, as if it were pushed aside by the ball. Render the shot. . This gives you a nice exam- ple of the ball being much heavier than the box and knocking it out of its path.The rotation of an arm's joints causes the hand to move along an arc ‘ball thrown in the air also follows an arc digital character animation 2 Arcs In nature, nearly everything moves in ares. This is partly due to pl and the way muscles move, but itis aso due to the fact thae neatly every joint in the body rotates. The arm doesn’t just move forwa shoulders and elbow rotate in combination to move the arm, which produces a natural rotation of the hand and fingers as well Gravity is another big factor in causing objects (0 move along ares. Falling objects and cartoon characters’ bodies follow ares. Gravity planets to follow elliptical orbits. It also causes a thrown ball to folk parabolic trajectory. As you ean see, ares are everywhere in nature,hapter 8: the basics of character animation me good example of arcs in action is the simple head turn. Suppose want to turn our character's head to the left, Our first instinct is to nply rotate it along a horizontal straight line. This makes the head mechanical, In real life, heads dip slightly below horizontal as they ysmaking a shallow arc. Heads do not turn on straight lines. Instead, they dip slightly in the middle Of the turn, forming an arc ow-Ins and Slow-Outs ether fact about natural motions is chat things don't just start and p suddenly. Rather, they accelerate from a start and decelerate to a Force plays a role in this. Newtonian physics states that any object jected to a force will accelerate. Think of a ball thrown straight up. force of gravity will slow it to a stop and then accelerate the ball as Is back to Earth. In animation, this effect is known as a sloii-in and out ough gravity is certainly one force that acts on your characters, there plenty of others. Forces can be external—a character being pushed, d, or moved by some outside object, for example. Forces can also be I, coming from inside the character's body—the muscle puts a pstant force on the joint, for example, which accelerates the limb in direction. Another muscle pulls it in the opposite direction, slowing (a stop. Expanding upon this concept, even a character's internal jougits and emotions can be thought of as forces—an indecisive acter moves much differently than a self-assured one. en performing a slow-in or slow-out, its best to use a Bézier curve, naturally slows in and out. There are many times, however, when might not want to do a slow-in or slow-out. If'a character hits aleasisieg digital character animation 2 brick wall, for instance, it stops dead in its tracks, without decelerating, In this case, a linear ot step curve might be more appropriate The animation curve for a slowsin and slow-out looks something ike this Notice howr the curve slopes toward horizontal as the speed nears zero. Force and Drag When animating, you also need to consider the effects of dragon an object. Force transmitted to an object does not affecc all parts of the object equally. Imagine two sticks connected through a flexible joint. If you pull one of the sticks straight down, the second stick takes a while to “get in line,” so to speak. This effect is called drag, and is also known as lag,apter 8: the basics of character animation ry Nf [causes the second joint to take the second stick must rotate to get Drag causes a delay in the two sticks time to follow the first. f you pull into alignment with the first. lining up, on the frst stick ler point to consider is how a multijointed object will move, Ifan has more than two joints, each joint drags behind the other. A stick added to the first ewo simply drags behind the second. sticks simply means more drag. The thied stick drags behind the md, which drags behind the first. YOR same principles apply to the joints of your character. The spine is just a collection of similar joints. Force transmitted to one end of spine takes time to reach the other end. Force applied to the arm time co reach the shoulder and even longer to reach the feet of a dog’ tail. The joints in the tail behave exactly like the joints ‘our stick example. The base of the tail rotates in a cycle, causing the er parts of che tail to drag,a) digital character animation 2 ‘The body's skeleton is really just a collection of joints. Force applied 10 ‘one part of the skeleton takes time 10 reach the others. A dog’s tail behaves exactly like our sticks. The end of the tail drags behind the base. da! a! be roochapter 8: the basics of character animation Squash-and-Stretch Unless you're animating the Statue of Liberty, your characters will ‘usually be made of flesh. Flesh is very pliable and flexes and bends considerably when moving, Think of how many different shapes the human face can make. The same goes for just about any part of the body: This characteristic is known by animators as sguash-and-stretch, The easiest way to illustrate this concept is with a bouncing ball. When the ball hits the ground, the force of impact makes it “squash” from a ‘sphere into an oblong shape. As it recoils, you can see it “stretch” in. the ‘other direction, The same principles apply in character animation. When a character Jumps in the air, i stretches as it takes off and squashes when it lands The same goes for things such as muscles and limbs. This does not mean the character has to literally stretch in volume. Realistic animals and ‘humans have squash-and-stretch as well. If a character lands on the “ground, the skeleton bends at the knees (squashes) to absorb the shock of impact. Ifthe character leaps off the screen, the knees staighten ‘Guretch) when it fies through che air ‘When a ball hits the ground, the force (of impact squashes it. Upon recoil, it Ce e@® mt'n much the same way, a flour sack also squashes when it hits the ground, A slightly more complex character squashes in much the same way. For an exaggerated, cartoonish effect, a character falling down a hole can stretch quite a bitin a few frames. digital character animation 2 When squashing and stretching, you must remember to maintain yo characters volume. Consider a balloon filled with water. If you stretch it or squash it, the water in one part of the balloon simply moves to another area. Never does the volume change. People are mostly water as well, so the same principle applies. No matter how squashed or distended the character, its volume always remains the same, If the volume increases or decreases, it appears as though the character is ‘growing or shrinkingchapter 8: the basics of character animation 187 When the rings move upward, this Just making the body longer will not Narrowing the chest area (arrows) character stretches, ‘work—the character gains volume and maintains volume and makes the simply looks talec stretch look natural. How to Squash and Stretch a Character Exactly how you squash and stretch a character in a computer depends a great deal on how the character is built and how it is deformed, Let's Took at son methods for squashing and stretching your characters Proper Posing Many times, particularly with quasi-realistic characters, y want to do a hyper-cartoonish stretch and you will not want to play vou will not Wwith the volume of the character. In these cases, the best squash-and- suretch is to simply pose the skeleton properly. When a character squash~ ts,bend the knees and curve the spine to make the character squash realistically: When the character is stretching, straighten the legs and spine to get a stretch that is anatomically correct Squash-and-stretch need not be a cartoonish effect. Ie is a part of every ‘motion of every creature on the planet. The cartoon character is simply exaggerating this effect for maximum contrast, Even when doing these Iypes of effects, you need to first get the pose of the skeleton correct before manipulating the volume of the character. Exaggetation is always based in realityScaling an object along one axis alone increases its volume. You must com- pensate by scaling back along the other axis digital character animation 2 Scaling ‘The easiest method for squashing and stretching is to simply scale all part of your character along one axis. If your character is bei squashed, reduce its size along the axis; increase it to stretch the chi ter. To maintain volume, however, you also need to increase or reduce! the volume along the other two axes accordingly. Skeletal Deformation Ifyour character has a skeleton, you can get squash-and-stretch faisly easily by manipulating the bones. Most packages that support bones enable you to scale and resize each bone. The affected vertices of your character's mesh will scale accordingly. If you want to stretch the arm, for example, simply scale the bones in the arm, Most of the same pri ples of scaling apply here as well; if you stretch the bone in one direc= tion, shrink it along the others to maintain volume. Another way to achieve stretch with a skeleton is to manipulate the space between the bones. This really works only with forward kinemat= ics, but if you want a character to stretch, simply move the bones ft apart. A similar technique works for IK systems that ate set up using. constraints. In the skeleton constructed in Chapter 6,"Skeletons and Mesh Deformation,” the foot was constrained to the leg at the ankle, the foot is moved beyond the length of the leg, the foot will separate from the ankle, causing the same effect as separating the bones (altho the effect is localized near the ankle). Some, but not all, IK systems have features that allow stretch to occur automatically. When the joint hits its limit, such as in the previous ¢ ple, the skeleton automatically stretches the bones to meet the effector: ‘This means the entire leg stretches to meet the foot19 a bone along its length squashes and stretches nt. dL his leg is manipulated via IK with a constraint to the foot the legis pulled past the IK limit, the bones separate at ankle, which stretches the leg at the ankle SSS—= systems, however, allow the bones to automatically to meet the effector chapter 8: the basics of character animation With forward kinematics, you can also stretch by manipulat- ing the distances beteen the bones, === beyond the limits of the joints, the bones remain the sare length,‘Another way to squash or stretch all or part of a character is vvath a deformation tool, such as this lattice. digital character animation 2 Lattices and Other Deformations Lattice deformations—as well as global deformations that allow you to bend, per, or otherwise modify the shape of your object—are great creating squash-and-stretch. You can apply these to the character asa whole or just to patt of a character, if your software allows it. Manipulating the lattice causes the character to stretch, Automatic Stretching Another method of squash-and-stretch is to let the computer do the stretching for you, Many packages contain soft body dynamies systems, which can simulate the motion of soft objects, such as skin. This mal it easy to create characters with Doppy ears or jiggly bellies. Soft can also be used to create other types of motion, such as overlap and follow-through, two more techniques that are explained a bit later. An Anticipation is also the body’s natural way of gaining momentum bel an action begins. Think of a man jumping. He swings his arms behind him, bends his knees, and actually moves down slightly to get momen- wor before jumping up.A baseball batter moves the bat back before ‘swinging it forward. A character leans back before getting out of a c Its sort of like getting a head start on the action pationfcipation can also be used to control the audience and get your vers to look exactly where you want them to. The human eye works ewhat automatically, [t is naturally drawn to things that move, By Wg an object in the opposite direction, you draw attention to it fore it moves. When it does move, the audience is already watching, the motion can be much faster. Anticipation is a perfectly natura! part ‘motion, and by exaggerating this, you can keep the audience's atten jon and achieve crisper timing. times, you may need to animate a very fast action, such as a 7ip- , where a character stiddenly dashes offscreen. The most important e a fast action, your audi- needs to be fully aware of the action before it occurs. This makes ticipation of the action very important. int to remember is that when you a + underestimate the importance of anticipating your character's ns when you are animating simply because it’s the way the body rally moves. An anticipation does not have to be a big motion or a pause in the action, It can be as short and subtle as a few frames, a small action. Anticipation means moving left before you move right and mowing down before moving up. Before this charac ter zips off, she cocks her body in anticipation of the move chapter 8: the basics of character animation eScharacter animation 2 Before this character puts his hand in the hat, he anticipates the move by flipping his wrist back before moving it down, 2 Exercise #2: Anticipation Simply by making a character leap out of frame, anticipation works, Before you do this exercise, stand up and do a you can see how small jump. Pay attention t0 how your knees bend and your arms anticipate the jump by moving 5 forward when you jump. If you want (0, back before swin use a stopwatch as you jump to 1. Place your character in the 2. Rotate the arms back and get a rough idea of the timing, shot in a simple. relaxed pose. bend the knees in anticipax tion of the leap. To maintain balance, the body should dt forward and down at the hips. This anticipation should take about 1/4-1/2 second past the first pose.chapter 8: the basics of character animation 3. Make your character leap. Straighten the body and swing the arms forward. Straighten the legs as they push against the ground and send the character forward. This action should take only a few frames co complete. 4. Render a test and adjust the poses and timing until you are satisfied. simple leap can be expanded with other little details. First, when ing, the arms lead the action by s frame or two. The feet do not to leave the ground at the same time; dragging one foor by a frame two behind the other may make the leap more natural as well. ly, remember that the path your character will follow is the begin of a parabola. As the character leaps out, make it follow an are. vershoot we've seen, anticipation is used t@ make the start of an action more like. At the tail end of the action, you have overshoot, Many times, body will not come wo a slow and perfect stop. Instead, it will over~ joot the stopping point for a few frames and settle into the pose. Like pation, itis a natural part of motion and can be exaggerated to the ator’ advantage. of a character throwing out its arm to point a finger. Before the ron starts, the arm anticipates the move. If the motion is quick, the racter’s arm naturally overshoots the pose so that the arm is absolute straight, It may even stretch past this point for an even more exagger~ effect. After a few frames, the arm settles inte a more natural, pose. Overshoot can be used to give your character’ actions snap, If you're animating pose-to-pose, you can overshoot a pose a few frames and then settle in,digital character animation 2 When this character throws out his arm, he overshoots the final pose for a few frames and settles in Secondary Action We've been concentrating on the primary motions of the body; many parts of the body move also. This is known £0 animators as secondary action. This broad category of actions includes the little oweaks, gestures, and touches that add extra life and personality to your character. It could be as simple as a charaeter tilting his hat before storming off, or a baby in a high chair wiggling her toes while she eats. The primary action is eating, but the secondary action—tilting the hat, wiggling the toes—adds personality and life to the character as a whole, Secondary action is best animated as a layer on the primary action First, animate the main motion. When this is correct, add the secondary actions. They should be little extra touches and not detract from the main point of the shot. Secondary actions are usually bracketed around the main action, but they do not interfere. Remember, the cardinal rule in animation is to do one thing at a time. The secondary actions serve 3s subtle bridge between the main actions. If they get too wild or selves and detract from noticeable, they become primary actions the the shot Another angle to this is the psychological gesture. Ifa character is talk~ a psy- chological place of interest. Perhaps the character really needs to sneeze ing and keeps brushing at his nose, the character is giving aw: Another example might be a character looking at his watch while he walks. This secondary action has now become a psychological gesture explaining that the character is in a hurrychapter 8: the basics of character animation Mar outside the main action, such as the motion of hair or clothing, Although this has become standard lingo for many animators, technical Jy, many of these types of action fall under the subcategory of overlap and follow-through, yy people use the term “secondary motion” to describe any action Overlap and Follow-Through When a character comes to a stop, not every part of its body stops on the same fame. Momentum carries some parts of the body past the sopping point. When parts of the body still move afer a character has come to a stop, that movement is called follow-through. As part of sec- ondary motion, follow-through motions are driven more by the laws of physics rather than the internal motivations of the character. Think of a dog with big, floppy ears. After the dog has stopped walki its ears continue to move, much like a pendulum. Drag also plays a big part in this. Overlap is almost identical to follow-through, because it involves secondary motion of the body. The body itself has many parts that don’t all move at once—a fat person’s belly bounces up and down during a walk, overlapping the action by several frames, for example. As with squash-and-stretch, these effects can be achieved through a soft- body dynamics module Follow-through also applies to the limbs of a body. The hips usually start an action, with the spine and arms dragging by a few frames. Think of an athlete completing a broad jump. His feet may hit the ground, but it may take many frames for the arms to swing forward and come to a ‘complete stop. Again, they act a bit like a pendulum, The dog doesn’t stop aif at once. His front feet stop first.digital character animation 2 e then his ears, collar, and tail follow through ‘and come to rest a few frames later, Eni UO SET ENNTE Let's make that other half of the leap you just created by making your character land. This shows you how different parts of the body follow through, overlap, and overshoot poses as a character moves Again, before animating this shot, act out the action. Stand up and take a short hop. Notice how, as you land, your knees bend co absorb the shock and your arms naturally swing forward, Ie also important to have overlap in the actions of the scene as a whole ‘Things do not move all ar once. Think of the situ characters are walking through the scene. If you start all the walks on frame 1, they appear to march overlap the motions of the walk, offSetting some of the characters. tion where many unison. Instead, the best choice is to 1, The character should come 2. Two frames later, the charace in from the leap in a few ter touches ground. The feet frames. As the character is hit heel-first, stop, and lock about to land, his knees to the ground. The rest of should be slightly bent, his the body, though, is nowhere hips forward, and his arms near stopping. and shoulders back.3. Two frames later, the momentum of the bods still downward and forward, forcing the knees to bend and the body to bend for- ward at the waist to absorb the shock. The arms are still back. . Four fames later, the body is ‘overshooting the final pose. ‘The spine and legs are almost straight, and the arms are coming to a stop above the head and beginning to move back down again, chapter 8: the basics of character animation ‘Three Games later, the body has now recoiled and is beginning to stand up. The spine is straightening out, one vertebra ata time due to drag, and the knees are straightening, The arms swing forward and start following through. Six fiames later is the final pose. The spine and legs bend a bit for a nice relaxed stance. The arms hang loose. 7. Render a test and adjust the poses and timing until you are happy with the shor.Riding a bicycle is a good example of the repetitive motions that can be han- dled by cycles. digital character animation 2 For additional practice, you can hook up this Janding with the leap the previous exercise, When doing this, remember that your character’ center of mass will follow a smooth parabola during the leap. Cycles One way to save time when doing ani is to animate in cycles. A cycle is simply the same sequence of keyfiames repeated. Think of somebody riding a bieycle. The feet on the pedals undergo a regular, cyclical motion. After you've animated one rotation of the pedals, you can simply copy the keys to create as many repetitions as you want In cel animation, the process is simply a matter of repeating your draw= ings. In 3D animation, you copy the keyfiames to other places on the time line. One problem with doing this is you need to remember that the last keys of the first cycle need! to flow into the first key of the next Many programs have a function that aligns your keys automatically for a smooth transition and repeats the cycle for as many times 3s you want, If you don't have such features, the best way to do this is manually. You simply copy the first frame of the first cycle to the frst frame of the next cycle before you begin animating. That way, you always have a tar= get so you can see that the last few frames of the cycle match up with the nextchapter 8: the basics of character animation Da Another problem with cycles is that they are repetitive motions, ard as such, can get stale and boring very quickly. You can avoid this by cycling ‘only part of a character’s motions. Let's say the character is walking. You ‘an cycle the feet and legs, but change the animation in the upper body to give the shot more variation Moving Holds There are many times where your character may need to hold a pose or be still —from a few frames to a few seconds. This is known as a hold. In ‘el animation, this is typically done by simply holding a single drawing fon the sercen for the duration of the hold. In computer animation, this can be very problematic, For some unexplained reason—maybe because of the realistic natute of the medium—a digital character held in a sin- gle position for more than a few frames will completely die. It looks as, though you hit the freeze-frame button on the VCR, A character left immobile on the screen looks as if you hit Move the character slightly over the duration of the hold the freeze-frame button, Here, the character has shitted his weight to his right sight- ly. Be sure to overlap these motions In order to get around this, you always keep your character movin even slightly: There are any number of tricks for doing this. One is to create moving holds Ifthe hold is for a handful of frames, you can sim- ply create two slightly different variations of the same pose and have the computer in-between these over the course of the hold.[200 | digital character animation 2 If the hold is in the range of seconds, you have to use another meth Secondary actions help a lot, but you should overlap many subtle actions in the hold to keep the character from looking too mechaniea You may want to shift the character’ weight from one foot to ano} Move the arm slightly. Til the head. Add some blinks. Ifyou have d ability to animate your character's chest, make it breathe, Be stire to overlap these subtle motions as well. Ie really a matter of keeping character still, but still alive. ‘A good exercise is to stand in the corner for a minute or two and try remain still. If you pay attention to your body, you'll soon realize that constantly moving. Your weight may shift from foot to foot, you're stantly breathing, you're blinking, and so forth. These are the same actions you can use to keep your character looking alive. You should also remember environmental things, such as wind, temp ature, gravity, and so on. If the character is holding its arm out, gravity will affect it, Let the arm droop slightly over time. If your character hair, slight breeze might help bring the hair to life. 1 its cold out the character may shiver. Putting It All Together ‘These last two exercises use most of the techniques you've learned 50 far. Use them as a way to develop and hone your skills By making a character stand up to anticipate the move, and then Load a character into your 3D. from a seated position, you use you lean forward and stand up. program. Next, model a simple anticipation, drag, overlap, follow- This gives you a guideline for the box for the character to sit on, through, and overshoot, among —_maotion, but what about the you want to create a chair ora other techniques. timing? If you have a stopwatch, sofa for the character to sit on, time yourself standing up and use that’ fine, but not essential. Before you go any further, you thiceotblotieta your ania: should get up fiom your chair several times. Notice exactly how your body moves. You lean back otherwise, use the timing | give you and adjust it to your needs.chapter 8: 1. Seat your character in a relaxed position. This is your first pose, the basics of character anima For your character to stand 3.- As the body gets out of the up, it needs to anticipate the chair, it leans forward quite a move to help it get some bit, with the momentum of momentum, Rotate the the upper body pulling the upper body back at the hips rear end out of the chair as As in the jump exercise, the the weight is transferred to arms should go back slightly the feet. Rotate the upper as well, This is your second body forward, and rorate the pose, Place this keyframe knees so that the rear end is approximately 1/4+1/2 sec- slightly out of the chair. Also, ond after the firs. swing the arms forward to help maintain balance, but drag them behind the body by a few frames. Again, 1/4-1/2 second is a good general timing, |. Create a standing pose, Remember to keep the feet firmly on the ground as the character stands, either by locking them in place with IK or by watching and manipulating the keys carefully. This timing should take anywhere from 1/2 second to 1 second after the previous key. The arms also do a bit of follow-through and come to a stop a few fiames after the body is standing, . Let the computer calculate the in-betweens. If your computer is fast enough, press Play; otherwise, render a test to see how you did. Go back and rework the timing and poses until you are happy with the shot.di ‘This last simple exercise makes a flour sack jump off a box and land. Animate this shot pose-to-pose by blocking out each of the major parts of the shot. Then, time the poses to get a good sense of weight. Finally, let the system in- berween the poses and tweak from there. 1 3. To make the sack jump, it needs to get its weight into it. This is done by bending the character at the waist and anticipating the jump. This should take about 6-8 frames. 4. gita character animation 2 Create a simple set contain— ing a box and a floor plane. 2, Place your flour sack in the set. Give the flour sack a natural starting pose, Asymmetrical poses are best. Create the take-off position Stretch the sack upward while keeping the feet plant ed. This should take 4-6 frames. Pose the top of the leap. The: sack will not be stretched at the top of the leap, Remember slow-in and slow-out, which tells us the flour sack moves along a parabola. The feet alo rotate forwand a bit in preparation for landing, This should take 8-12 frames.chapter 8: the basics of character animation Bs 6. Pose the landing. As the 7. As the flour sack’s weight hits character falls back to the the ground, it will squash, ground, gravity stretches the pethaps bending over a bit as lower part of che sack: well, Squash it anywhere ‘Typically, both feet won't hit fiom 4-8 frames the ground at once, so keep this pose asymmetrical. This should take 8-12 frames 8. The sack finally stands up in 9. After these major poses are a stable pose. It may need to set; cweak the titng to get take a step to get into this ‘overlap and follow-through. pose. Affer the poses are The timing given here is blocked out, play with the very rough. Learn to derive timing to get the best sense the timing yourself so that of weight. Standing up you understand what feels should take 6-12 frames. right and what doesn't. Conclusion In this chapter, you explored the basic animation concepts, such as squash-and-stretch, anticipation, and secondary motion. These techniques are the palette of motion and the colors with which you bring your characters to life. These techniques tell you why and where to place the keyframes in an animation. The chapter alo presented the basics of most 3D packages and how they relate to these concepts, which showed you how to set and manipulate the keyframes. Understanding these tools helps you control your characters and their motions better.Uol}oW0907 pue BulyjemWalking and Locomotion Jow that you've had a bit of practice with animating the human figure, it time to start moving your characters around. By moving, T mean walking and running, Walking requires tons of balance and coordination; its amazing how easy people make it look. ‘Walking conveys a great deal about a character's personality. The next time you're in a crowded place, notice all the different types of walks that people have. Some people waddle, others saunter, and some drag their feet. It’ amazing how almost everyone you see has a unique walk Mae West, Groucho Marx, John Wayne, and Charlie Chaplin are all characters who have very distinctive walks. If you want to know who a character is figure out how he walks. Computer animators have a number of tools available for animating walks. lt seems as though software vendors have focused on this prob- lem with considerable effort, and quite a few new and innovative tools are available for automating all or part of the walking process. These sophisticated tools can be both good and bad. As the animator, you should still understand exactly how characters walk and how you want your characters to walk. If a piece of software enables you to get this done in half the time, that fintastic. Just be sure that you are the one controlling the process, not a piece of software. If the computer does the animation, it will look computerized—stale, mechanical, and lifeless.On the CD are four animations that show you this walk from four different views. These files are called WALK 1 AVI through WALK4.AVI (Mac users should load WALK. MOV through WALK4.MOV). Watch these ani- mations for reference. digital character animation 2 Understanding the Mechanics of Wal ." Every time you take a step, you actually lean forward and fall slightly, and are caught by your outstretched foot. If you failed to put your foot forward, you would fll flac on your face. After your foot touches the ground, your body's ‘weight is transferred to it and your knee bends to absorb the shock. The front leg then lifts the body and propels it forward as the rear leg swings up to catch you again, and the cycle repeats. Before you read any further, get up and walk around the room for a bit. Pay attention to how each part of your body moves. You'll soon notice that every part of the body, from the feet to the arms to the head, has it own unique set of motions. As you walk around, notice how you lean forward into the walk, and how your legs neatly catch your body to prevent it from falling, If you purposely hold your foot back on a stp, you'll fall lat on your face. The process of walking is very complex. Not only do the feet have to move across the ground, but the hips, spine, arms, shoulders, and head move in sync to keep the system in balance. Although these movements are complex, if' you break them down joint by joint, the mechanics of walking becomes clear ‘The following sections break down a basic walk, step by step. For clarity, I've animated a simple skeleton so you can see how each joint moves. Feet and Legs The feet and legs propel the body forward. To keep your character Jooking natural, you should always keep the joints bent slightly, even at full leg extensionchapter 9: walking and locomotion Ao The welk usually starts with the feet As the weight of the body is wens- This figure illustrates the point halfway at the extended postion, where ferred to the forward foot, the forward through the first step. As the character the feet are farthest apart and where knee bends to absorb the shock. This moves forward, the forward knee the characters weight shifts to the position is called the recoil position, _straightens out and lifts the body to its forward foot and is the lowest point in the walk. highest point. This position is called the passing position because the free foot passes the supporting leg at this point. “= As the character moves forward, the The free leg makes contact with the weight-bearing foot lifts off the ground ground, completing half the cycle. The atthe heel, transmitting the force to. second half is an exact mirror of the the ball of the foot. The bady starts to first. if it differs, the character may fall forward. The free foot swings for- appear to limp. ward like a pendulum to meet the ‘ground and catch the body's weight. The Hips, Spine, and Shoulders The body's center of gravity is at che hips; all balance starts there, as does the rest of the body's motion. During a walk, it’s best to think of the hips’ motion as two separate, overlapping rotations. First, the hips rotate along the axis of the spine, forward and back with the legs. If the “right leg is forward, the right hip is rotated forward. Second, at the pass- ing position, the fiee leg pulls the hip out of center, forcing the hips to ‘ock ftom side to side. These two motions are then transmitted through the spine to the shoulders, which mirror the hips to maintain balance.digital character animation 2 ‘vanen the feet are fully extended, the hips must rotate along the avs of the spine. To keep balance, the shoulders swing in the opposite direction. From the front, the spine is relatively straight, but from the top, you can see how the hi and shoulders twist in opposite directions to maintain balance. ‘At the passing position, the front view shows the hip being pulled out of center by the weight of the free leg, causing a counter-rotation in the shoulders. From the top, the hips and shoulders are at nearly equal angles. The hip also moves slightly off center, moving away from the free leg. If all the weight of the charac: ter is evenly distributed in the center all the time, the walk looks static and dead, At the extension of the second leg, the hips and shoulders again are flat when viewed from the front. From the top, you can see the completed rotation of the hips and shoulderschapter 9: walking and locomotion is holding something or gesturing, its arms general~ ly hang loose at the sides. When walking, they act like a pendulum, dragging a few frames behind the motion of the hips and shoulders. Even at full extension, the arms should be slightly bent at the elbows to keep them looking natural The Head and Spine from the Side Ifyou look at a character’ spine from the side, you can see its general posture, which may be very stiff or slouched over, depending on the character’ attitude. The spine also absorbs some of the shock transmitted to the hips from the legs, making it flex from front to back a bit. In a standard walk, the head tries to stay level, with the eyes pointing in the direction of the walk, but it bobs around slightly to stay balanced. If a character is excited, this bobbing is more pronounced, The head may also hang low for a sad character, or may look around if the scene requires it Because walking is kind of like falling forward, the body For example, if characteris terbly sad, he tends to hunch should be angled forward slightly at the hips for most forward and hang his head low. This posture forces the hips walks. The spine arches up slightly to keep the chest and to rotate in the opposite direction, giving the body a differ head over the hips. However, ths line of action can change ent attitude. He'll most likely drag his feet as well With the character’ attitudei When you're animating multiple characters, it's always tempting to make them walk at the same rate to simplify the animation process—not a good idea. Giving your characters the same gait makes your characters look a if they're marching in unison, which ‘can detract from the shot. its ‘always best to stagger walk ‘cycles and give characters differ- ent gaits. I one character is walk- ing at 12 frames per step, give the other a gait that’s slightly slower or faster—maybe 10 frames per step, or 15. By mixing it up, you make your shots more digital character animation 2 Animating a Walk Now that you understand the underlying mechanies of walking, you can attempt to animate a walk, Using traditional animation techniques, a walk cycle is tough; it can be just as tricky on a computer, Timing the Walk ‘The first thing you need to concern yourself with is the timing of the walk, How many frames does it take? That's not an easy question to answer, Is your character large and lumbering or small and scrappy? Is your character running or walking? Happy or sad? All these factors determine the amount of time it takes your character to take a step. Ata normal walking gait, a step takes anywhere from 1/3-2/3 second (8-16 fiames at 24fjs, or LO-20 frames at 30fps) with 1/2 second per step being about average. A full cycle (both right and left steps) takes about 1 second per cycle. Larger characters tend to walk slower, and smaller characters walk faster. In general, men have slightly slower gaits than women, and sad people walk slower than happy people. One nice thing about working with a computer is that many programs enable you to scale the length of your animation. If your character is walking too slowly, you can speed it up a bit by reworking the keys. Keeping Your Feet on the Ground ‘The most important thing to remember when animating a walk is to keep your character’ feet firmly locked to the gwound. The friction between the feet and the ground propels the character forward. If the feet slide around, the illusion of friction is lost and the animation will sm realistic. (If your character is walking on banana peels or an oil sliding may seem hilarious.) How do you keep the feet locked to the ground? It really depends on your software and its feature set. Many packages have tools to asist you in this task; others do not, The following are a few of the more popular methods: Inverse kinematics (IK) with constraints—IK is one of the pre nation methods because i ensures that a character’ ferred walk~ grounded foot stays in the sane spot for the duration of the step. Knowing that the (eet are locked frees the animator’ mind from annoying details and enables him to concentrate on more impor tant things, such as posing the character.chapter 9: walking and locomotion Footstep generators—Some packages have what | call “footstep generators.” These plug-ins enable you to toss down footprints on the ground, up hills, down ladders, wherever you want. The pro- gram then automatically moves the character's legs and feet to match, Typically, the motion is generated by an algorithm that computes the character's physical weight and dynamics. Using such a plug-in is a great way to get the first pass of a walk: into the computer, but that’s about it. One bad thing about this type of automation is that che walk is typically generated from the feet and not the hips. Animators must keep this in mind when tweaking the animation, because most all movemert in 2-walk leads from the hips. As the animator, you still need to control the process and add personality to the walk. You must go back over the animation and tweak the poses and motion on a second and third pass to add the flavor and individuality that makes your char- acter unique. Inverse or broken hhierarchies—This method does not require IK and works with any package that supports forwand kinematics, tn a normal skeleton, che hips ate at the top of the tree, with the spine and legs as children. Ifyou don’t have a way to lock the feet, this setup can cause problems, because moving the hips moves the spine, legs, and feet. In the inverse hierarchy method, the hierarchy is turned upside down and broken at the hips. The feer become the parents of the legs, enabling you to place the feet where you want them, Because theyte the parents, you cat rest assured that they won't move unless you absolutely want them to move. What about the hips? In a hierarchy, because an object can't have ‘ovo parents, the hips must be disconnected from the legs and are forced to float free above them. The hips are then animated to rnatch up with the legs, Because exactly matching the hips to the legs can be problematic this method works best for characters vhose hips are hidden—a soda can with feet or a character in a Tt also works well when using bones, because the character's skin tends to obscure most hip-alignment problems. Besides, view~ ers usually notice the fect slipping long before they notice that the hips are nor quite connected; its the lesser of two evils.al character animation 2 ae) Es a SPINE rere] E t tae am] al RE t t_ Gel a] oe ay cb ‘na normal hierarchy, all joints point to the hips. Thus, by moving the hips, you move the entire body. This method can make it very eas) for the feet to slp inadvertent. THIGH THIGH ¥ ¥ saw] [ane ¥ z ' = In an inverse hierarchy, the link is broken at the hips, and the feet become parents of the legs. The feet never move ‘unless you tell them to, and they never slp. 1 Forward kinematics with guides—If, for some reason, none of these other methods works, you can tough it out and keyframe the walk by hand. To aid in this process, you can always use guides to aid in the placement of the feet. In any package, you can place null objects, dummies, or transparent objects along the character's feet and use them as virtual placeholders. If the foot moves inadver- tently, you have a reference point that enables you to put the foot back where it belongs.chapter 9: walking and locomotion yy This box is being used as a quide in the placement of the foot. By aligning a reference point on the foot to the guide, you can be reasonably sure it won't sli Some packages enable you to simulate an “onionskin” effect by ghosting the previous frame, providing a good simulation of the traditional animator’ light table. Because you can see where the feet are on the previous frames, i's rather easy to align them on the current frame. Finally, some animators have been known to use dry-erase pens to draw the guides directly on the computer sereen. The marks can then be erased simply by wiping them off. Still, 'd hate for you to grab the wrong pen and make permanent marks on the screen, so be carefull A good substitute might be a sticky note placed on the screen and marked with a pencil. To Cycle or Not to Cycle? Because walking is a cyclical motion, you may want to create the walk~ ing motion as a cycle rather than as straight-ahead animation. If done properly, a cycle can save a lot of animation time. One cycle can be applied to a number of different environments. Your character and its cycle can just as easily be placed in a cityscape as a country road, because the walk motion is essentially the same. Why dupiicate your efforts? Clasical animators use this «rick a lot, simply repeating the same Sequence of drawings and swapping only the background painting — plicing the character in a different location. In 3D, you have the flexibil- ity to change the cameras, lighting, and environment to make the shot look completely different.i ‘This multiple exposure of a walk cycle shows that the body does not move forward; instead, the feet move beneath it. The red marks show how the foot on the ground moves the same distance on each frame to prevent slipping, digital character animation 2 Many times, one cyele will look completely diferent fiom a diffe camera angle. This can work for and against you. If you are work a game animation and the character has an identifiable walk, it m imilar enough to recognize from any cantera angles. If you are fon a crowd scene, many times you can use the same walk from a cent angle and it looks like a different character walking. Those who work in the interactive and gaming industries deal with cycles every day. Most game engines require that you animate your charactets in cycles, which are then called up as the player uses the Joystick. In these cases, you may also need little 3-6 frame animati called links to bridge the gap between cycles—a walk and a run, example. There are downsides to using cycles, First, because the eycle is re tive, it can seem sterile and flat, particularly when viewed for an e3 ed period of time. Second, cycles work best on level cerrain. If your character has to walk around a corner or over a hill, the cycle might not match up properly. Animacing a cycle is similar to making your character walk on a t mill. The body does not move forward; the feet simply move beneath it.To maintain the illusion of walking, the entire character must be moved actoss the ground (or the ground moved past che character) at the same rate that the feet are moving. Ocherwise, the character's feet appear to slip. Also, the foot on the ground needs to move the same distance on each frame. Again, if the length of the steps varies, the feet appear to slip.chapter 9: walking and locomotion Although it doesn't really have legs or arms, the basic dour sack is a great place to begin when animating a walk. If, as in Chapter 7, Posing Characters,” you picture the sack as the hips and shoulders of a character, you will understand how these parts of the body move when walking How you animate the sack depends a bit on how it is deformed. The basic way to deform a sack is simply by placing four bones, one at each corner of the sack, with a fifth bone at the center. Because this structure is0 simple, you should not need to Tink the bones in a hierarchy. Not having them Tinked also avoids foot placement and slippage problems. This walk will be timed at 12 frames per step, 24 for the whole cycle. a & 1, Start with the feet slightly Move to fiame 6, halfway apart, Place the left foot for- * pea the first step. Move ward and the right foot back the right foot up and forward To keep the character's so that it is even with the weight balanced, make the lefé. Set a key. shoulders mirror the legs. Place the right shoulder for~ ward and the left shoulder back. Set keys for both the shoulders and the feet, ‘Make sure the bone in the center of the body also moves along, Move it forward so that the weight is centered and set a key.digital character animation 2 4. Balance out the upper body, 5, Get the end of the step (and 6. Move the center bone for Recall how the shoulders the start of the next). Move ward to keep the weight mitror the hips in a walk. the right foot forward and centered between the two Move the right shoulder place it firmly on the fect. Set a key down and back. Move the ground. Set keys for left shoulder up and forward. both feet. Set keys for both. 7. Mirror the shoulders to the a hips by moving the right shoulder up and back, Move the left shoulder down and forward. Set keys for both. This finishes off the first step. Repeat the procedure for the next foot to get a fall eyele, and then again for as many steps as the character needs to take. To add a bit more life to the walk, you may want to give the sack’s belly a bit of bounce. Remember the concepts of drag, overlap, and follow-through. As the sack walks, its belly (controlled by the center bone) tends to drag behind the action a bit. Gravity also tends to pull it toward the ground. Go back over the walk and add a bit of up-and~ down motion to the center bone. As the right foot is moving up, for instance, the bone drags behind by 2 or 3 frames, so it doesn't move up until slightly later. If you did the walk properly, you'll sce how even a very simple character can be made to walk. Remember this lesson as we move on to more complex characters.chapter 9: walking and locomotion | Exercise #2: Animating a Walk Cycle Using Forward Kinematics In this exercise, you animate a walk cycle without the aid of inverse Kinematics or locks. Many animators consider this method to be the kind of a fail-safe technique; it should work just abour anywhere with almost all packages. This cycle is timed at 16 frames per step, 32 for the whole cycle. You're doing this example with a simple skeleton, but the techniques should transfer well to any two-legged character. The ability to lock your char- acter’ feet to the ground is important, but not necessary, as [' explain along the way. To begin, load up your character and hierarchical skeleton in your favorite 3D animation program. Create an animation chats 33 frames Jong. I've added the extra frame to the end as a target for frames in the second half of the cycle. It will not be rendered. Start by animating the hips and shoulders. Its bes: to start here, because all other motions derive from the hips. The hips have two separate, over~ Ihpping rotations that are mirrored by the shoulders, The first rotation is along the vertical axis of the spine and follows the position of the legs and feet. Go to the halfway point (frame 17) in the cycle and rotate the hips and shoulders in the opposite direction. Remember, they should still be mirroring each other. At this point, the hips should also move sideways slightly to maintain balance over the planted foot 1. Stare the walk with the right 2 foot; the right hip must go forward, as well, On frame 1 of the animation, rotate the hips around the y axis so that the right side is forward. From the top view, rotate the shoulders to mitror the hips’ rotation. Ifyou don’t have a character, you an always use one from the CD. On the CD is a character named TINKRBOY.DXF or TINKRBOY 305; you can use either if you want. i 3. Go to 1 frame past the end of the cycle (frame 33) and copy the first frame's keys here. This frame is a target for the frames chat are in- betweened on the second half of the cycle.ai | digital character animation 2 4. Next, you need to cteate the sway of the hips. Go to the frame in the middle of the first step (frame 9). IF your rotations are correct, the hips and shoulders should be parallel when viewed from the top. At this point, the passing position or the point of highest leg, extension, the body rests on the right leg and the left leg pulls the hips out of center From the front view, rotate the hips around the z axis so that the right hip is higher. Adjust the spine and shoulders so you get a smooth line of action and the shoulders mirror the hips. 6. Go to the middle frame of the second step (frame 25) and reverse the rotations that you just made at the middle of the first step, The body rests on the left leg at this point, and the spine curves in the opposite direction. Again, che hips move slightly 10 the left to balance over the planted foot. 7. Adjust the spine on frames 1 and 25 to give ita forward lean and a nice curve, 8, Go back to each keyframe and adjust the Jegs and arms so that they hang vertically throughout the cycle. 9, Play back the cycle, If it looks smooth and balanced, move on to the next step. Otherwise, tweak the keyframes until you have a nice smooth motion.chapter 9: walking and locomotion Ea 10. You now need to move the legs and feet—the trickies part of the process, First, set up the extreme poses. From a side view, go to if you in-berween the foot linearly at the two extremes frame | and set the first pose, where the legs are at maximum extension. Copy these keys to the end of the cycle, at frame 33. 11. Go to the middle of the cycle (frame 17) and mirror frame | so that the left leg is forward. A ghosting feature would help consider- ably in this proces. “fo aid in the anienation process, use a guide to belp position = (vith no slow-ins of slow-outs, the foot moves across the floor os ct. If you are using locks, you probably won't have to do automatically. Still, guide acts as a nice double-check. 12, To create the guide, model a box and place it directly beneath the floor near the character’ forward foot. If i's below the floor, it won't show up at rendering time. You could use a null object. 13. Go to frame 1 and move the guide horizontally to the place where the toe hits the ground. 14, Go to the middle of the cycle (frame 17) and position the guide at the same place on the toe as you did in frame i 15, Sct up the guide so that it in-betweens these two positions at a linear rate. The guide will tell you exactly where the toe needs to be at any point in the step. (If you have an IK system, you exa pin the foot to the guide for a surefire solution.)al character animation 2 16, Now, you need co tweak the poses, Because you know where the feet need to be, you can concentrate on the legs. About a quarter of the way through the first step (frame 5) is the recoil position — where the leg absorbs the shock and bends to its lowest point. Move the hips down so that the shin is forced to rotate forward a bit, giving the knee a nice bend 17. The body recoils upward into the passing position at frame 9. Move the hips up so that the forward leg is fairly well extended. Its very important to keep the knee bent slightly to make the action look natural. 18, Ac this point, the weight of the body is on the ball of the foot. The heel lifts off the floor as the body falls forward. The hips are moving down at this point. There may also be problems with the free foot as it swings forward: if you have extra-big shoes, they'll hit the floor unless you bend the toes slightly 19. The first step is now complete, Create a second guide and repeat these procedures for the left foot on the second haif of the cycle. Be careful to make the second half as close to the first as possible. Render a test and go back to tweak any inconsistencies. Now you need to create the motion of the arms and head. In the simplest case, the arms swing back and forth to maintain balance in opposition to the legs. The arms also drag behind the action a bit, placing the arms' extreme poses a few frames behind the legs20. chapter 9: walking and locomotion Rotate the arms into position on the first frame. Because the 1 d, the right arm is buck, and the left arm is forward. leg is forw Go a few frames (from 2-5) in for the armis extreme pose. 1 chose frame 5. On the left arm, rotate the forearm back to a hice extension, On the right, rotate the forearm usp slightly: Go a few frames (from 2-5) past the start of the second step and mirror the extreme from the previous step. Finally, copy the keys on frame 1 to the last frame so the arm will swing through to the end of the cycle. Now you need to do the head. Go to the keys at the start and halfway through each step and rotate the head so that it remains vertical and the eyes are facing forward. The head can bob from side to side a bit, as long as it not too distracting, You can also make the spine bounce up and down a bit. Remember the Chapter 8, ““Basies of Character Animation.” This applies to the spine as well. effects of drag that you learned in If you want, you can make the spine flex between curved and straight, dragging behind the up-and-down motion of the hips by a few ffames. As the hips go up, the spine compresses into a curve. When the hips come down again, drag causes the spine to straighten out.Prscencsrnnewsrs On the CD is a finished movie of the walk. The title is TINK- WALK.AVI or TINKWALK.MOV for Mac; you can watch it as a refer- ence, if you like. Inverse kinematics is by far the most popular way to animate the lower body. IK is one of the preferred walk-animation methods sures that a charac because it 1 ter’s grounded foot stays in the same spot for the duration of the step. Knowing that the feet are locked frees the animator’s mind from annoying details and enables him to concentrate on more important things, such as posing the character. digital character animation 2 Now that you have a convincing cycle, you need to get your character off the treadmill and out into the world. To do this, you can do one of ‘hwo things: Bither move the character along the ground or move the ground under the character. Moving the ground is best when you want ‘ing camers that is locked on the character; because the character is still, he camera can remain still. Moving the character is best in cases when you wane the camera stable and the character to walk past. wo use a In this example, move the ground. If you used a guide to assist in your animation, the task is simple. Find the absolute position of your first guide in frame 1, and then again when it stops in the middle of the cycle. For example, in my shot, the guide moves along the x axis. The position of the guide along x is 300 units. At frame 17, the guide is at 100 units; he character’s foot moves a total of 200 units per step (300 = 100 = 200). This value is known as the stride length. Doubling it equals 400 units for the total cycle. On the first frame, move the floor to the starting position. On the lst frame (fame 33), move the floor 400 units along the x axis. In-berween these frames linearly: That’ it. The shot is done. IK chain to an effector that resides outside the hierarchy of the skele- ton. This exercise has the leg chains constrained at the ankle to outside effectors. These effe tors, in turn, parent the feet chains Animating an object outside the hierarchy of the skeleton has an advantage because moving the ips does not affect the position of the feet. Implementation of IK differs between packages, but most soft- ware enables you to constrain an | As with the previous walk, this 1, Start with a neutral skeleton. walk is timed at 16 frames per for one full cycle.chapter 9: walking and locomotion =a 5 i \ ’ ft ev. S&S a2 = AA 2, Start with the lower body. 3. Move the slider to frame 8 to 4. Position the right leg. Move Create the first pose on the get the passing pose. Move the right ankle so that itis first frame, Position the left the hips forward so that they directly above the left ankle ankle forward, the right ate directly over the left and slightly below the left ankle back. Position the hips ankle. Move the hips up so knee, Sec a panition key for tulfinay between the feet. that the left leg extends to the right ankle, Rotate the hips along the nearly straight. Rotate the vertical axis so that the lef hips along the vertical axis hip is slightly forward. See back to zero, Set position and position and rotation keys for rotation keys for the hip. all these objects, Create a pose that is the muir ror of the first. Place the j f right ankle forward so that the foot rests firmly on the \ ground. Move the hips for~ \ ward and down so that they are halfway berween the feet. Rotate the hips along the vertical axis so that the right hip is slightly forward. Set position and rotation keys for all these objects. Also set a position key for the left toot Scrubbing the animation ——- - — 8. From a front view, go to frame 8, the middle of che cycle. The hips should now be rorated slightly to the right due to the weight of the free right leg. Rotate the hips and set a rota- tion key. Move the hips slight- ly to the left to place the weight of the body toward the left leg, Scrubbing the anima- tion should now show a basic lower body step, digital character animation z 6. Now tweak the step. Go back to fame 4, the recoil position, where the hips sink as the weight is transferred to the forward foot. Move the hips down at this point and set a position key. 9. Move on to the upper body, starting with the spine. At frame 1, rotate the joints of the spine so that the shoul- dets mirror the hips, Set rotation keys for the spine. 10, at At the recoil position, the k foot needs to be rotated slightly downward. Rotate the foot and set a rotation key. frame 8, the middle of the firse step. Rotate the spine so the shoulders are even with the hips when viewed from above, From the front, rotate the spine so that the shoulders nuirror the hips, Set rotation keys for he spinechapter 9: walking and locomotion [5 ah 11, Move to frame 16, Adjust the 12, Move on to the arms. With Repeat steps 1=12 to create the spine so that it mirrors the the right leg forward, the left step. When finished, render pose on frame 1. Set rotation right arm needs to be back. a test, keys for the spine. The step is Conversely, the left arm. almost complete. needs co be forward. Set rotation keys for these objects. Move t0 fiame 16 and mirror these poses, set ting rotation keys once again. Rather than working a step at a time, in actual production, you proba- bly want to block out all the steps first. Next, you move on to the upper aan body and set all the keys for a full cycle, and then copy these keys for the duration of the walk. On the CD is a test of this walk. Itis titled IKWALK. AVI or IKWALK.MOV for Mac.digital character animation 2 In this example, you animate a You will show this walk on a slight gaps between the legs and walk using inverse hierarchies. standard skeleton for reterence. hips will usually not show up on This method works well in almost The one thing you will notice isthe deformed mesh any package. One nice thin, that it is difficult to align the hips about it we thae this tet locks exactly with the tops of the ee Begin by getting the hierarchies down the feet quite easily and Because of this, this type of walk Set UP Properly. The hierarchy for forces you to move the hips prop- is best for nonsegmented charac~ ‘he leg is broken at the hips. For erly. Also, you can animate this ters. An ideal character is one each leg, the toe parents the foot, walk straight-ahead rather than in where the hips are concealed, such Which parents the shin and then aeycle, Animate this walk at 12. as 9 Woman in a skirt. This the thigh. It is the exact opposite frames per step for film (15 frames method ean also work with a of the standard skeleton, but can per step for video) deformation program, because still be animated quite exsily. F LoL 1, Set up the bones (or seg You now have three separate 2. Your main task is to animate ments) on the right lez hierarchies: two legs and a the legs and feet. Leave the so that the toes are at the top standard skeleton with no body our of the shot for this of the hierarchy. The order of legs. part, because it won't be the hierarchy is as follows: needed until later. Position toes parent the heel, then the the legs in the extended posi- shin, and then the thigh. Do tion, with the right leg for= the same for the left leg, ward. To adjust the positions of the legs, move the toes first, and then work your way up the chain from there: The ion of the legs deter ies the stride length. Also, be sure that the tops of the legs are not too far apart, where the “hips” would be.chapter 9: walking and locomotion Ea \ 3 - }. Now you can animate the 4. Jump forward to flame 13— 5, Align the passing position. rest of the walk, On the right the beginning of the next Go to frame 7, which is leg, rotate the toe and move step. Move the left foot into halfway through the first step, the foot so it contacts the the extended position and and rotate the right shin at ground. From this point, the rotate (don't move) the toe the ankle to make the leg foot remains stable until the of the right foot to the almost vertical. Rotate the next step is taken, Notice proper position. left leg so the foot is some how you can rotate the joints what vertical and position the of the leg without affecting Go back through the first thigh by rotating the knee. the position of the foot. The step and align the rest of ‘The tops of the right and left feet stay put and no slipping the keys. legs should be about equal, occurs, with the left slightly lower. \ 7. Create a key near frame 10, 9. You now have the first step. where the left foot swings By repeating this procedure, forward. Because the hierar- you can continue and create \. Go to frame 4 and get the chy is backward, positioning, as many more steps as you recoil position. Again, rotate the free leg may be a little want, Animate three or four the right leg at the ankle so tricky. more steps. When you're fin~ that the knee is bent, Rotate ished, you should have a and align the left foot so that 8. Play through the first step nice pair of legs walking the leg is free. and tweak the keys. across the floor.digital character animation 2 0. Now that you have the legs 1. Move to the next set of keys, and feet, all you have to do is at the recoil position (fiame 13. 14 place the body above them 3), Position the hips above and rock it back and forth so the legs, but lower. If you that the hips of the body skeleton match up with the tops of the legs animated the feet and legs properly, you should naturally see how to position the hips On the first frame, position the body s0 it is above the legs, with the hips aligned with the legs. Rotate the spine to give it a natural stance. Continue through the rest of 15. Animate the arms, As in the the shot, positioning the hips previous example, you need above the legs and rotating to have them drag slightly them from right to left along behind the action of the legs with the tops of the legs. ‘When the right leg is for award, the right arm is back, Play back the animation to and vice versa. make sure that the character remains centered above the 16. Finally, render the animation legs. and tweak as necessary. 12. mM Ny NI Next is the passing position, In this position, the left side is slightly lower at the hip, rotate the hips slightly in this direction. =chapter 9: walking and locomotion Beyond Walking All these exercises have shown how to create a basic walk. There are plenty of other ways to get around besides a walk. Characters may also ‘un, skip, sneak, shuflle, and tiptoe, among many walking variations. Within these gaits, each character's personality will also come through in the walk, Many animators think the walk is one of the key parts of a character’ personality. John Wayne's walk is completely different than Groucho Marx's, for example. Knowing the character means knowing the walk. With that in mind, lets analyze a few other types ‘of locomotion. The Run ‘Arun is more than just a fast and highly exaggerated walk. Instead of continuous falling, i's best to view a run as continuous leaping, with the body tending to lean forward a lot more. In a walk, one foot is always ‘on the ground; in a run, there are times when both feet are airborne. The stride length also increases, making this distance longer than the feet could normally reach in a walk. The timing of a run is faster than a walk, and can get down to a few frames per step. If you're animating an extremely fast run, motion blur is id to keep the feet and legs from strobing. Take a look ning character and sec the differences between a walk and a run. [ On the CD are four animations that will show you this run from four different views. These files are called RUNI AVI through RUN4.AVI (Mac users should load RUNT. MOV through RUNA.MOV.) Watch these animations for reference This figure illustrates the “contact posi tion.” Notice that the body leans for- ward, and the legs are farther apart than they are during a walk,digita character animation 2 Alter contact is made, the forward leg absorbs the shock of This fique illustrates a position similar to the passing the body at the recoil position. As in walking, this point is tion, because itis halfway through the step. At this point the lowest position in the cycle. Because the body is moving the cycle, the grounded foot pushes the body upward, faster, momentum is increased, causing the bent leg to be even more exaggerated. The body is over the planted foot to help balance the weight ‘The body is now aitborne and is at the highest point ‘The body lands and the next step starts. As in the walk, in the cycle second half of the cycle should be a mirror of the first. The body is over the planted foot to help distribute the weight,chapter 9: walking and locomotion Eee The Skip Another gait that is completely different from walking is the skip. In a skip, the character takes off and lands on the same leg, and then switches, legs for the next step. Even though the foot pattern is changed, the con~ ‘cepts of weight and balance do not. The character's hips, shoulders, and spine all move in concert to keep the skeleton in balance, be This skip, animated by Angie Jones, As in the walk, the weight is trans- The foot pushes off, sending the char= shows the basic foot and body move- fered to the forward foot (or in this acter into a small leap or hop. The ‘ents. The skip starts much like the case, the forward toe). ‘opposite foot moves forward as the Wak, withthe feet apart. A skip hap- hips rotate. ‘pens primarily on the toes, however. *¢ Inthe middle of the leap, the back leg ‘The character lands, but an the The forward foot touches down, and moves forward sightly. The forward leg rear foot, with the forward foot the cycle repeats as a mirror of the remains high and tucks under the body. staying high. first step,ae digital character animation 2 The Sneak Another favorite walk is the standard sneak. If you analyze it, this is variation of the basic walk, but the character is walking on his toes. is a good example of how the basic walk can be modified to show acter and mood. The main goal of a sneak is to walk quietly, ‘This sneak, animated by Angie Jones, The recoil position. Notice how the feet The passing position. In the sneak, the starts much ike the standard walk, stay very close to the ground to help planted leg does not extend nearly as bout with the character on his toes maintain balance much as in the walk ‘This makes the character slightly off center, so he extends his arms to regain balance. Before the passing leg touches down, The cycle is finished and ready for the the character leans over at the waist next leg. to help maintain balance.chapter 9: walking and locomotion Animating Four-Legged Walks Ifyou think mastering a two-legged walk is difficult, you'll find a nur-legged walk doubly troublesome. First, the anatomy of most four- legged animals (as well as birds and dinosaurs) is quite different from mate anatomy, <¢ a horse, for example. The horse actually walks on its toes. Whereas human heel actually touches the ground, the horse’s “heel” is far the ground—where the human knee would be. The horse’ knee isactually even higher up.as are the thigh and hip. The front legs are ilar to human arms. Again, the horse walks on his fingers and his “wrist is far above the ground. create a realistic four-legged walk, you need to study the movements ‘of real animals. Take a video camera to the 200, oF rent a nature docu- ntary and go through the animals movements, frame by frame. addition to normal walking, a four-legged animal has several different ts.The animal varies the timing and rhythm of its steps as it moves, rand faster. These gaits are the walk, ot, canter, and gallop. Some imals, however, may do only a subset of these. The elephant, for ple, always walks—it never changes its gait It simply walks slower faster. the walk, the horse’ legs behave very much like the arms and legs of hhuman—if the right rear leg is back, the right front leg is forward, the opposite happening on the left. This positioning changes as the des change. By the time the horse has reached a full gallop, the front are in sync, going forward and back nearly in unison. The back legs rate the same Way.pay digital character animation 2 EL AL AL ef The walk, BE AL hh 1h The gallop, Another way to view a four-legged walk is in a more cartoonish way. Think of the old vaudeville act where two guys get into a tattered old horse suit. On stage, the horse literally walks like two people stitched gether. You animate the walk like a double-cwo-legged walk, forcing. you w have different joint constraints and body construction The back legs on this dog are not realistic; they bend the same way human legs do. Still, the cartoon nature of its design enables the animator to move the dog this way and get away with it A cartoonish four-legged character is much simpler to animate, because you can set up its walk Ike a double-two- legged walk.chapter 9: walking and locomotion Conclusion Walking is a very technical animation task. involving weighs, balance, and motion. As you master the syntax of walks and runs, move on to _adding the subtleties of character. Each of your characters will walk a bit ferently, depending on its personality. Like a musician who learns his les and then learns to improvise, a walk isa fundamental building ck for animators to improvise upon.anBojeig pue uolewiuy jeisesFacial Animation and Dialogue Armes the face is one of the most challenging and rewarding tasks you will encounter as an animator. We humans are all experts in the subtleties of expression and emotion, We know instine- tively when a bad actor is on the scre —rypically, because his emo- tions seem forced rather than genuine. The same goes for animation; great facial animation looks seamless and does not draw axcention to itself, whereas bad animation seems curiously “off” for some unknown reason. Animating the face requires a good eye and a thorough knowledge of acting and emotion. While you are in the learning process, keep a keen eye on people, their face background in classic films—particularly those of the silent comedians, and how they express themselves, A good who had to express themselves purely through facial expressions and body language—can prove invaluable to the animator. Additionally, good software tools are a must. You need a py ‘can build and animate complex shapes with a minimum of fs. A ige that swell-modeled head that is flexible and animates simpy will save you a couple of bottles of aspirin down the road.38 digital character animation 2 Methods for Animating the Face ‘Over the years, many methods have been used to animate the face. E tuter graphics, some people rugged and pulled the vertices of face, frame by agonizing frame, co get animation. Later, other people deformed the face directly through the used of bones, clusters, and late tices. All these methods have their place, but they all have a number of weaknesses, most important of which is the lack of control over the subtle details of the shape of the face. ‘Over time, most people learned that the best way to control these subtle details is in a modeling program, not a deformation module. The method that takes advantage of the modeling program is called ‘morphing. In most productions, morphing has become the de facto standard method for animating faces. Morphing Morphing can create very complex facial animation. Although other methods ace still used, morphing is very powerfull and fainly easy to control. Much of the facial animation exercises in this book concentrate ‘on the use of morphing, Morphing changes the shape of the base object into the shape of other, target, objects. Setting up your character for animacion is simply a mat- tet of using your programs modeler to work up a library of the appro- priate facial poses. This has distinct advantages, because each shape can be refined to perfection in the modeling program, rather than at anima~ tion time Morphing works for both patch and polygonal models. Some packages force you to create separate objects for the shapes, whereas some enable you to create shapes by using one model. Most software packages require that the models being morphied contain the same number of vertices in the same order, a task easily accomplished by modeling a sin- gle, stock, expressionless face, copying it, and reworking it into the many ‘expressions and ficial poses you need. There are two basic methods for creating morphs: single and muluple target morphing Single-Target Morphing A single-target morphing system allows only transition between two discrete shapes. This forces the animator to model all the possible poses ahead of time, which discourages experimentation You also can't dochapter 10: facial animation and dialogue ey] simple overlapping actions, such as closing the mouth while opening the yes. This results in animation that seems stiff at best. Most decent soft- ware these days offer the more-sophisticated feature of multiple-target morphing, Multiple-Target Morphing Multiple-taxget morphing (also known as shape weights and blend shape) enables mixing multiple shapes. It enables you to model individual shapes for parts of the face that smile, open the jaw, raise the eyebrows, and so on. You can then combine these shapes for a pose that is a smil- ing face, with the jaw 50% open and the eyebrows raised 309%. This makes manipulating the face quite easy. Each morph target moves only part of the face, with the facial expression derived by mixing all the parts. You can combine as many shapes as you want, All you realiy need. to animate are the relative weights of the shapes, making it much easier to create the perfect pose for a given moment and alo requiring that fewer poses be modeled. Most packages these days even allow you to attach these weights to virtual sliders so that you can poses, which makes generating poses at animation time downright easy ‘dial in” your Modeling Morph Targets for Animation The best way to model morph targets for animation is to study the underlying anatomy of the fice. Recall from Chapter 5,""Modeling Heads for Animation,” how the muscles move the face. In a multiple- target morph, because poses can be mixed, all that needs to be modeled are the extremes of the individual muscles. Because each muscle is con- nected to the face in a specific place, flexing just that muscle changes the shape of only part of the face. If you model that exact shape, you have an anatomically correct animation when all the shapes (all the muscles) are mixed together. Ideally, you could model targets for every muscle in the face, which, amounts to several dozen targets. For ease of use, this book outlines a dozen or so of the most useful shapes. Although some of these poses combine several muscles, the theory is the same: Each slider control. only part of the face. When modeling for a morph system, be carefial to move only those vertices that need to be affected by any particular muscle. If you accidentally move a vertex on the ear, for example, a character's ear may change shape every time the jaw opens. Also be sure to make smooth transitions from the right to left sides of the face. IF youry digital character animation 2 model a right smile, make sure that the smile transitions smoothly to the left side of the face, and vice versa. Ifthe same part of the face is manip= ulated by more than one slider, you may get unwanted results, such asa top lip that goes bonkers when both smile sliders are moved ‘Also remember that these poses are extremes, so be sure to model them as suich. Try to imagine the biggest smile, the saddest frown, and so on, ‘Try looking in the mirror and use your own face as reference. Lower-Face Poses Lower-face poses are centered around the mouth, although some, such as the smile and sneer, affect the eyes a bit, Lower-face poses are usually Used for lip syne, but can also be a part of many other expressions, The first pose is the neutral face Model it with the eyes open and the jaw slightly slack, leaving the mouth ‘open, All the other poses will be derived from this basic face. ‘The next pose i the open jaw. This is modeled by selecting The pursed li, or oooh shape, is the product of the obicu- the vertices of the lower face, and then moving them lars ors muscle. When modeling this shape, be sure to down and slightly back to simulate the effect of the jaw maintain volume in the lps, which get slightly thicker as rotating open. the muscle contractschapter so: facial animation and dialogue ev Smiles are primarly the action of the zygomatic major muscle. In 2 smile, the corner of the ‘mouth is pulled up into the cheek, which tends to puff out. In an extreme smile, the cheek igeates an arc, which cuts into the bottom edge of the eye. Tis frown is a combination of the trianguiaris and depressor labilinferoris muscles, Basically, these "muscles work to pull the corner of the mouth doven The see’ s the product ofthe levator labi superioris muscle. t pulls the lip up toward the edge of the nose. I also affects the skin along the side of the nose al the way up to the corner of the eye242 digital character animation 2 Upper-Face Poses Ic has often been stid that the eyes are the mirror to the soul. Although many animators may concentrate on the mouth and lip syne, true emotion is communicated through the eyes, brows, and upper face. Getting good upper-fice animation is eritically i facial animation. The corrugator pulls the brow down Left and right eyebrows are manipulated by the frontals. Each brow is modeled {and in foward the bridge of the nose. seperately to get maximum control. In an extreme position, the character looks angry, although this shape is used for many other emotions If your character has eyelids as part of the face, a left and right blink also need to be modeled ;portant to any Barnes: All these face shapes are located on the CD.chapter 1o: facial animation and dialogue ery Upper-Face Poses for Stylized Characters Some charicters are modeled without fleshy eyebrows and lids. These character can be still be manipulated to get upper-f e poses. Typically, a character may be modeled with simple spheres for eyes and a hemi: sphere for a lid. By rotating the lids to different angles, you can still get the effect of a brow Other Facial Poses In addition to the poses mentioned in the last sections, any number of ‘astom poses can be modeled to get just the expression you need. If the sliders don't ve you the face you need, the best advice is to get the face as close as you can, fx it, copy the new head to make another target, and use the n sv target as the face for that particular moment. Nit all characters have fleshy eyes, They are simply placed inside this to make cartoonish eyes These eyes are made from simple characters head, spheres with the lids as hemispheres, Ifthe lids are rotated outward, you get Rotate them inward, you get another ine expression expression. In this case, the lids take the place of the brows.digital character animation z ‘This pose simply closes the lips. it can be helpful in dalogue This pose tucks the bottom lip under the front teeth to get or when tweaking pose the sound of the letter F (is used in dislogue. In addition to custom poses, there are a few more stock poses you can model to complete your inventory. These are not mandatory, but they do come in handy. Assigning Targets After the targets are modeled, getting them to animate is simply a ‘matter of assigning them as morph targets. How this is done depends 6on your software, so consult your manual on the exace keystrokes. When assigning, itis a good idea to keep the targets grouped logically within the interface. With 10-20 sliders to contend with, it’s much more effi- cient if the right smile slider is located next to the left smile slider, for example, Keep lower- and upper-face targets together for the same reason. Manipulating the Face Afier the targets are assigned, the fun can begin. Most software provides sliders, which make manipulating a face quite easy. Ifthe targets were modeted correctly, manipulating the face is as simple as moving sliders. As fun as it can be, you still need to understand some basics of how facial expressions are manipulated and how facial expressions convey emotion, Pushing Poses Most software packages enable you to manipulate a slider beyond the range of 100%. This means that your big smile can be even bigger if you move it above 100% to, for example, 125%, When a vertex is mor- phed, the software simply moves it along a path to its new location: sochapter 10: facial animation and dialogue ae ‘when a brow raises, for example, the vertices representing the brow move along a line to their new position. If the slider is at 50%, the ver= tex travels halfway to the new position. At 125%, itis beyond the posi~ tion, but along the same line. This technique can be very handy, but you will find that there is a point past 100% where the surface of the fice will begin to break up. or the pose will simply look unnatural. If your poses are already pushed, this may happen just above 100%, Still, this lite technique is great for get~ ting a litle bit more out of your character. Negative Sliders Another trick is to push the sliders below zero into negative territory. This simply causes the vertices to move in the opposite direction, but along the same fine as before. This can enable some of your poses to do double duty. For example, if you push the jaw slider below zero, the mouth closes, which may eliminate the need for a closed mouth pose. Another good example is the center brow: When positive, the slider indicates anger. When it is pushed into negative territory, it raises the brow, which creates a worried expression. The right and lefé brow can also be used in such a manner. Pushing the blink sliders into negative territory opens the eyes wide, Some poses, however, do not behave well when pushed into negative territory. A negative sn and frowns are also too specific to work well when pushed into negative territory. The center brow also works along the same lines. Positive When the jaw slider is moved into positive territory, the jaw values show anger, and negative values show worry. ‘opens. Negative values make the jaw close. will not make a frown, for example. SneersAnger: The eyes are open, but the brows are down. The Disgust: The whole face is tightened, withthe eyes nar- mouth is usually open with the lips tensed and teeth bared. rowed and center brow lowered slighty, The mouth is The jaw may be lowered digital character animation 2 Creating Facial Expressions The reason you manipulate the sliders is to create facial expressions. Expressions are much like the body poses that an animator uses to help. a chatacter act and carry off a scene. When building an expression, always ask yourself what the character is feeling at the moment. This helps guide you to create the correct face. Basic Expressions ‘The face can make an infinite number of expressions; luckily, they fill into some broad categories: anger, disgust, fear, joy, sadness, and surprise, Til detail them here, but you must recognize that they are extreme expresions only—primary colors among the many hues of expression. You could never animate with these six alone, but understanding these basic “faces” can take you a long way toward understanding the under- lying mechanics of expression. closed and the upper lip pulled into a sneerchapter to: facial animation and dialogue Fear The mouth is wide open and pulled back at the lower Joy: The mouth is pulled upward into a smile, exposing ‘corners. The jaw is dropped, and the eyes are wide, with the upper teeth, forcing the cheeks up. The brows are the brows raised. Usually relaxed. Sadness: The mouth is pulled down at the lower corners Surprise: The eyes are wide open, and the brows are and may expose the lower teeth. The eyes squint, and when raised but not furrowed. The mouth is relaxed, and the ying, may be closed. The brow is raised only in the middle. jaw is slack.digital character anima The Brows and Expression ‘Remember that the upper part of the face is the most important part of an expression. One way to change the meaning of an expression quickly is to manipulate the brow. The position of the brow probably conveys more information about emotion than any other part of the face. When brows are lowered, it usually indicates dark emotions—anger, for exam~ ple. When raised, the face becomes more open. ‘Changing the brow changes emotion. With brows With the same mouth, but brows lowered, he raised, the character looks happy. looks evil Symmetry and Expression ‘The expressions we've looked at so far have been fairly symmetrical. When animating the face, its always best to introduce asymmetry into the poses to make things look more natural. Just as when posing the body, the dreaded twins can creep into facial animation as well, making your face look stale and boring. Still, you need to be careful not to make things too asymmetrical or it will change the meaning of the expression. When the face pose becomes extremely asymmetrical, the expressions become less clear. Ifa character flashes a crooked smile, for example, it usually means he's not sincere. If one eyebrow is raised, it can indicate curiosity or insight. Of course, if your character needs to express such. emotions, this knowledge can work to your advantage.chapter 10: animation and dialogue This symmetrical face is nice, but not Lowering the left smile @ bit introduces very interesting, asymmetry and makes the face more natural Complex Expressions Many expressions are not as distinet as the ones indicated previously. Of course, the palette of human emotions is broad with many subtle hues. ‘Many times, facial expressions are asymmetrical, and many combine various attributes of the basic expressions, A smiling mouth ‘with lowered brows might indicate evil, for example mes they When creating these expressions, remember the basics learned from the fundamental expressions. Because the eyes are the most important co conveying emotion, start with the eyes. Much as with the hips in the body, if you get the eyes right, the rest of the pose falls into place. If you ever get confused as to what the expression should be, look in che mir- ror and try to mimic the expression. Ba Lower it too much, however, and the meaning of the expression changes.al character animation 2 Many complex expressions can be made simply by manipulating the sliders,chapter 10: facial animation and dialogue Animating the Head and Face When the face is set up and you understand how to pose a face to get expressions, you are ready to animate. Those who are learning facial ani- mation should not start with learning to animate dialogue, but with the nple act of conveying emotion. You do not need to have a character speak for it to be expressive, There are plenty of examples of characters who express themselves quite well without dialogue. Animating the Head Before you animate the face itself, you need to study the head and how it moves. A perfectly animated face on a perfectly still head looks horri- ble. Head motions are necessary to accent and emphasize the facial ani- mation. Ifa character is talking, his head nods or bobs to accent certain lines of dialogue. Ifa character is curious, he may cock his head to one side, If the character is disapproving, the head may shake as ifto say “nol” Ha character turns his head, it usually suggests a shift of attention or focus. (As was mentioned in Chapter 8, “Basics of Character Animation,” the head always turns along an atc. When the head turns from left to right, it usually dips in the middle of the turn.) Also remember the principles of animation. If a character turns his head to the right, he anticipates it with a slight turn to the left. The character may overshoot the carn slightly and sete in. All the principles of ani- mation apply equally co the head and face as they do to the body. Eye Direction Because the eyes are the most important part of the face, where the eyes are pointed is a very important bit of information, Be sure to have your character’ eyes firmly fixed on the subject at hand. If your character is talking to someone, but the eyes are looking off into the distance, your character will appear as if its staring into space and not paying atten- tion, Of course, this can be used as an effect, but typically you want your characters to be looking at the people they're talking to. Another little tidbit of human interaction is that even when a person moves his or her head slightly—to accentuate a part of speech, for example—the eyes tend to remain fixed on their target.digital character animation 2 “Look At” Functions Most packages have functions that enable you to constrain an object's direction so that its pointed at another, Some packages call this function “look at” contoller, because it forces the object to always face or look at the object to which it is constrained, When such a controller is applied to the pupils af the eye, for example. the eyes automatically rotate to track an object. ‘One good example is a fly buzzing around the room. By constraining the pupils of the eyes so they always look at the fly, you can make the fly go all over the place and the character’ eyes always follow it. This example is extreme, but the principle applies even to a situation as simple as two people talking face to face, each chat on the other. You simply have each character’ eyes look at the other character's face. eyes fixed ‘A much more flexible way to use the same function is to create an invisible null or dummy object to help define your chatacters focus es look at this invisible object. You can then move the object anywhere in the scene and rest assured your char Simply make your character’ e acter’ eyes will be pointed exactly where you want them, In this scene, the character’ pupils are fived to the red When the sphere moves down, the pupils fallow, remaining sphere with a “look at” controller. fixed on the sphere. Making the sphere imisible enables you to use it to guide the eyes in any shot. Simply place it where you want your character to look, and the eyes will folow,chapter to: facial animation and dialogue EZ Pointy Objects Another way to keep the eye on target is to stick pointy objects in your character's eyes, No, I don’t want you to maim your characters. By pointy objects, ! simply mean an invisible object used as a guide to assist in point- ing the eyes. Usually, you use a thin cone or a cylinder that has a transpar- ent texture applied to it so that it doesn't show up when rendering. To create such a guide, simply model a long, thin cone or a cylinder and attach it to the pupil via a hierarchy or a constraint. The guides will then point out into space like searchlights, showing the direction of the charac ter's gaze. Ifthe pupil turns, so will the guide, giving you a very good idea as to where the eyes are pointed and helping you aim them at their target. Blinks Blinks go a long way to adding life to your character. Generally, you should have your character blink every few seconds just to show it alive ‘The timing of the blink itself depends on your character’ personality and mood. A really fast blink might ake 4 frames, and a normal blink will take 6-8. A 20-frame blink may mean that the character is very sleepy or very dumb, In this wireframe view, you see how these simple cones help if you turn off the cones to rendeving (or apply an invisible position the eyes. texture to them), they disappear after renderingeenn5chapter 1o: facial animation and dialogue Ea Blinks and Squash Another trick, particularly with c: toonish characters, is to squash the eyes as they blink. This will help exaggerate the motion and bring the blink to life. The squash can be done in any number of the methods mentioned in Chapters 5 and 8. Many people simply scale the eyes, but some use lattices around the eye hierarchy to get better control of the shape. To get a litle bit more bounce to the blink, another tick is to ove shoot the shape at the end of the blink. Let’ say you have a blink of 8 frames. At the bottom of the blink on frame 4, the eyes should be squashed, At the end of the blink on frame 8, make the eyes sightly taller, and then settle in to the final pose on frame 10. This also works for more realistic eyes; simply move the shape sliders into negative values so that the eye gets slightly wider for 2 frames at the end of the blink, Blinks and Head Turns During a head turn, most people tend to blink. Adding a blink in your character's head curns will make him seem more alive, Another litle tid- bit is that characters tend to look in the direction of the head turn, so it’s best to lead the turn with the eyes. ff the head is turning left, the eyes will turn left before the head turns. OD 2 ©8 08 23:0 To get added life from a blink, squash the shape of the eyes at the bottom of the blink and then overshoot the final shape bya few frames $8588 The eyes usually point in the direction of the head turn, and a blink in the middle adds some life.digital character a Blinks and Eye Direction One problem that often crops up is a character with a lazy eye, This is an eye that seems to float rather than appear to be locked on a specific target. Some of these problems are poor eye direction, and others can be fixed with a simple blink. When people change the direction of their gaze, itis almost always accompanied by a blink. If your character looks from left to right, add a blink as the eyes change direction. The blink will help the eyes look less “floaty.” ‘There are times wher: a character's eyes need to travel without blinking, One good example is when a character is reading, In that case, the eyes go from left to right without a blink. Remember, like the head, the eyes move along arcs, dipping slightly in the middle of a tum, Overlap in Facial Animation One of the big reasons that multiple target morphing is used in facial animation is because it gives you the ability to overlap the actions of the face. Just as with the body, actions do not happen all at once in facial animation. character may need to smile as he’s talking, for instance, or pethaps change the position of the brows. ‘One tip is to let one side of the face lead the other. If a character smiles, the left side of the face may begin the smile a frame or two ahead of the right, Also, as is the case with head turns, the eyes tend to take the Jead ix facial animation. Ifa change of mood happens, the change usual- ly begins with a change in the expression of the eyes. To understand how facial animation works, do a simple test where a character changes his emotion through the simple use of expression Make a character go from happy to sad, curious to shocked, or worried to fearful, for example. Be sure to use blinks, as well as head and body motion, to accentuate the shift. Animating Lip-Sync Animating lip syne can really frighten the beginning animator—and righty so because its one of the most difficule techniques for an animator to mase ter. Live-action people have it easy; they just point the cameras at the actors and ask them to speak. Re-creating natural lip movements in animation,chapter 1o: facial animation and dialogue however, requires a great deal of time, patience, and analysis. In addition to getting the mouth shapes and positions right, you must be concerned with the acting and body motions associated with the dialogue. As with any hard task, however, animating lip sync is grounded in some simple techniques. Practice those techniques and you'll be on your way to mastering lip sync. ‘As you have seen, facial animation is a lot more than just moving the mouth. When animating a shot, it is very important to create facial anima tion in layers. Do not do it all at once; begin with the dialogue. When a character speaks dialogue, the shape and position of the mouth is perhaps 10-20% of the total effect. After the character is speaking with a neutral face, layer in the eyes and brows and then the rest of the facial animation, moving onto each as you finish the first. This keeps the facial animation from being mushy. Equally as important to the audience is the movement of the body and head. For the purposes of this book, let's start the process with the mouth, Recording Dialogue In animation, dialogue is almost always recorded before the characters are drawn. Dialogue looks more natural when the animator follows the natur~ al rhythms of speech. Voice actors have difficulty matching previously ani- mated dialogue while trying to make the dialogue sound natural, which is why recording the speech before animation begins is essential. Directing a voice session is an art in itself, but the voice actors must know the scene and the setting, This knowledge enables the actors to inflect different tones and pauses into their speech. It also is useful for timing purposes if the actors act out character movements as they dialogue. If the voice actor does this, it might be a good idea to videotape the voice session. You can chen play the tape to provide additional refer- ence for your animation of both dialogue and gestures. Another method that may prove usefitl is to give the actors a storyboard so they can visual- ize the shot as they record the dialogue. record the Afier the dialogue is recorded, the animator is responsible for breaking down the track, frame by frame, into individual phonemes (the most basic sounds of human speech) to be animated. The easiest way to picture a phoneme is to think of each discrete sound that makes up a word. The word “funny,” for example, has four phonemes: the f sound, the uh sound, an n sound, and a long e. Reading a dialogue track can be a tedious cask. The dialogue must be hroken down frame by frame and written by hand onto exposure sheets. Computer animators have the advantage of using digital audio software to read their tracks, and some packages are now attaining the capability of performing voice recognition and can actually break down the track automaticallyal character animation 2 The exposure sheet is wed to break down the dialogue track, frame by frame. Here a characte is saying, "Give ita shot” (highlighted in white). The exposure sheet can also contain other information. To the lft of the dialogue i$ 8 column in which to write notes on how you want your characters to act— to blink or bob the head, for example. To the right of the dialogue are notes indicating which mouth shapes to use. The Eight Basic Mouth Positions Before reading the track, however, you must first understand how the mouth moves when it speaks. There are dozens of different mouth shapes made dur- ing the course of normal speech. Animators usually boil these down to a handfl of standard shapes that are used repeatedly: Depending on the style of animation, some animators get away with as few as three or four shapes, and some may use dozens. For most situations, you ean get away with approx imately eight basic mouth positions. These eight positions usually provide ade~ quate coverage and give you the ability ro animate most dialogue effectively. ‘To really see how these positions work, watch yourself in the mirror while you talk. Make the sounds used by each position. If you talk naturally, you'll begin to see how the shapes work and how they all fit together into a coatin= xnous stream. The shapes and the rules that govern them are certainly not strict. Different accents and speech patterns may cause you to substitute ‘one shape for the other in order to achieve a more convincing look. ‘You may notice that some of these positions are not in the standard library of shapes you modeled as morph targets. They can, however, be created at anima- tion time by mixing the appropriate sliders. One good example is the sound of the word “Oh,” which is created by mixing together an open jaw with the ‘oooh sound. In faet, for speech, most of the grunt work is done by manipulat~ ing only these two shapes, with a possible fif thrown in when needed. Other sliders—such as smile, frown, and sneer—are used mainly to add character to the face.chapter ro: f - Position A is the closed mouth used for consonants made by the lips, specifically the M, 8, and P sounds. Typically, this can be made by pushing the open’ jaw pose into negative territory to close the mouth. In this posi- tion, the lips are usually their normal width, For added real ism, you could mix in an ‘additional shape to get the lips slightly pursed for sounds fol owing an oooh sound, such asin the word “room,” Position E has the mouth wide open in an elliptical shape. ‘This isthe position used for the vowel O, as in the word flow.” It is created by mixing together an open jaw and the ‘00h sound. Sometimes, par- ticularly when the sound is at the end of a word, you can ‘overlap this shape with the ‘one in Position F to close the mouth, acial animation and dialogue Ea & & Position B has the mouth Position Cis used for the Position D is used primarily ‘open with the teeth closed. wide open vowels, such as for the vowel E, but it can This position is a common Aandi. Its essentially the also be used on occasion shape and is used for con- samme as the fundamental for C, K, or N during fast sonants made within the shape for an open jaw. speech mouth, specifically sounds made by CD, G, K,N,R, 5, TH, ¥, and Z, All these sounds can also be made with the teeth slightly ‘open, particularly in fast speech, Position F has the mouth Position G has the mouth _Position H has the bottom lip smaller, but more pursed. _wide open with the tongue tucked under the teeth to Position F is used for the __against the teeth. This posi-_-make the sound of the let- ‘000h sound, as in food, and tion is reserved for the letter ters For V. In highly pro- for the vowel U. Its one of, but can also be used forD nounced speech, this shape the fundamental mout or TH sounds, particularly is necessary, but the shape shapes, when preceded by A or | it could also be replaced with 's essentially an open jaw Position B for more casual or with the tongue moved up rapid speech. This shape is against the top teeth. if ‘one of the extra shapes speech is particularly rapid, modeled previously this shape may not be nec: essary; substitute Position B.tal character animation 2 Reading the Track Now that you understand the basic mouth positions, it’s time to break down the track. If you have animator's exposure sheet paper, use it. Otherwise, get a pad of lined paper on which to write your track, using one line per frame. (If you prefer, you can create a spreadsheet for this purpose and do it all digitally’) Load the dialogue into a sound editing program. A number of sound editing packages are available, and you should choose one that enables you to display the time in frames, as well as select and play portions of the track. The ability to label sections in the editor is also handy. ‘The first thing you should do is match your sound editing program's time base to the time base you're animating—30, 25, or 24fps, for example. After your time base is set, selecting a snippet consisting of a few frames of dialogue should enable you to listen to the dialogue a word or phoneme at a time and read cach phoneme’s exact length on the editor's data window. The visual readout of the dialogue gives you clues as to where the words start and stop. Work your way through the track and write down each phoneme as it occurs on your exposure sheet, frame by frame. This is a tedious bus necessary chore. Some packages give you the ability to play back audio in syne with the animation, This feature is particularly helpful because you may be able to skip the step of reading the track and simply eyeball the syne, Sell, its always a good idea to read the track methodically before animating so that you completely understand the track and know exactly where all the sounds occur. ‘When reading the track, be sure to represent the sounds accurately In human speech, most consonants are short and usually don't take up more than 1 or 2 frames. Vowels, however, can be any length, Ifa person is shouting, for instance, you may have vowels topping 30 frames in length. In these cases, it is important that you don't simply hold the mouth in the exact same position for more than a second—it would Jook unnatural. Instead, create two slightly differem mouth positions and keep the mouth moving between them so that the character looks alive,chapter 10: facial animation and dialogue a Exercise #2: Reading a Track Let’s take a line of dialogue and d it for a imation On the CD-ROM. choose the audio file called DIALOGUE. WAY if you use a PC, o choose DIA LOGUE if you use a Mac Load the appropriate file into your favorite sound editing program. The dia- logue ‘Oh no, they're on to me” At 30fps, the dialogue measures about 20 frames. After you the load the file into your sound editor, this dialogue file looks some thing like the image co the left The first sound in the file is 4 gasp that runs from frames 22 through 30.Exercise: continued The next big hunk of sound is the word “Oh,” which runs from frames 42 through 49, ‘After that is the word “No.” Work through the entire track, writing down the positions of each phoneme The following is a graphic representation of where the phonemes fill. digital character animation 2 Cer Sr Animating Dialogue Now that you've read the track, you're ready to begin animating the dialogue. Dialogue is slightly different from lip syne, because lip syne simply involves the lips. Dialogue, however, involves the whole character. ‘When animating characters, be sure co get the character's entire body into the acting, Mouth or Body First? With the body so important to dialogue, one of the questions you might first. In cel anima. have is whether to animate the mouth or the bo tion, animators are forced to draw the mouths last, because it makes no sense to draw mouths on a character until the animation of the body is wn, In stop motion, the mouths are done at the same time as the body. In CG, it’s really not that big an issue, because any part of the ani mation can be tweaked independently of the others.chapter 10: facial animation and dialogue Some animators simply do the mouth first just to get the tedious task out of the way. It also is easier to get the mouth animated first om a still head rather than on one that is moving. Another good reason to do the facial first is to get the shythm of the track. At a certain point, you no longer hear the words and instead hear the “beats” in the track. This enables you to animate the body much better, because you know where the major accents are in the track. Other animators like to concentrate on the body first and then get the mouths, Many times this is done for characters who have extremely broad gestures—characters who “talk with their hands” Both approach es work equally as well, and because you can always go back and tweak the body and the lips independently, the line between the two methods is pretty muck a gtey area. For this particular animation, let's stare with the lips and then move on to the eyes, head, and body. Animating the Mouth Ifthe track is read properly, the phonemes and their location are pretty much known, In the track you just read, for example, you know there is, a gasp at frame 22 and the word “Oh” at frame 42. One important trick that works to your benefit is to always try to open the mouth quickly and close it slowly. Vowels Vowels are those points in speech where the mouth opens. When ani- mating a vowel, you need nwo positions. The first position is the accent pose, when the vowel is first uttered. The second position is the cushion pose, which happens toward the middle-to-end of the vowel sound. The accent usually has the mouth open wider than the cushion. One good way to do this is to animate the jaw so it closes slightly as the vowel progresses. For fast vowels of only 2 ffames, this may not be much of an issue, but this rule of thumb applies to anything above 4 frames. In the dialogue track you read, for example, the first word “Oh” takes approximately 7 fiames to complete. In this time period, the mouth needs to travel from wide open (accent) to the lips closed and pursed into an oooh sound (cushion). At frame 42, you open the jaw and then animate its value lower through frame 49 as the jaw closes. Simultaneously, you add the oooh pose, which rises in value from somewhere around frame 42 through 49.For reference, on the CD, there is an animation of just the mouths {for the dialogue read previously, called DIALMOUTH. AVI, , For reference, on the CD, there is an animation of the mouths and eyes called DIALEYES.AVI digital character animation 2 Consonants Consonants are those points where the mouth closes. With the posible exception of a long M, F.orV sound, most consonants are only a few frames in length; some can be less than 1 frame long. With this in mind, make sure you leave each position on the screen long enough for the audience to read it. At 24 or 30 fps, consonants must be on the screen for at least 2 frames in order to be read. If the consonant is too short, steal time from a vowel or combine two consonants into one. Eyes and Dialogue After the basic lip sync is accomplished, the eyes are next on the list. When animating eyes with dialogue, be sure you understand where the character needs to be looking. Ask yourself the question, “Who is the character talking to?" Try to keep the eyes focused on the subject at hand. Of course, there are places where a character may need to look away. People who are nervous tend to give darting glances. A dishonest per- son’s eyes may be somewhat shifty. Don't be affaid to change the shape of the eyes and brows as the dialogue requires. A character whose eyes remain the same shape throughout a line of dialogue appears lifeless. Blinks are also very important. They accompany most major head motions, so if the head turns or bobs to accent a phrase, you need to blink the eyes as well. Dead spots in the dialogue are also good places to sprinkle in a blink or two. Blinks can also be storytelling tools; shy or lying people tend to blink a lot, for example. Head Motion and Dialogue The head moves quite a bit when people talk. The head bobs, nods, and shakes to emphasize certain words in a line of dialogue. When speaking oud sounds, the head usually raises to help open the throat. This is helpfull when animating the loud sounds or accents in speech. When animating an accent where the head raises, it is always a good idea to anticipate the motion by lowering the head three or four frames before the accent and then popping up the head on the accented syllable. This is also known as a head bob, and can be accompanied by a blink. To get more action into the head bob, you can also get the body into the action. As the head moves down in anticipation of the accent, raise the shoulders a bit. As the head pops up, lower the shoulders Taken to extreme, this type of motion is the same as used in the classic cartoon “take.”chapter to: facial animation and dialogue pr In the track you just read, good place for a head bob might be as the character says, “Oh no.” As he says “Oh,” lower the head, popping it up a at the start of the word “no” You can overshoot this pose a bit and then settle into a more relaxed pose by the end of the word “no.” ‘On the CO, there is an animation of the mouths, eyes, and head | motion called DIALHEAD AVI Body Language and Dialogue ‘When talking, many people use their hands to clarify and emphasize the ‘major points of their speech. Getting this part of the animation correct isa lesson in acting. If you want to see how net to animate, watch some really nervous or first-time actors. They usually are very self-conscious, stuffing their hands in their pockets, wringing them nervously, or hang ing their hands loose at their sides. tn teal life, body language precedes the dialogue by anywhere from a few frames to as many as 20, Generally, slow, dim-witted character has more time between gestures and dialogue than a sharp, quick character. Speedy Gonzales has considerably less lead time on his gestures than Forrest Gump. Someone giving a long, boring speech is much slower than a fire-and-brimstone evangelist. You should also make an effort to ensure thar your gestures fit che dia logue smoothly. The first gesture every animator learns is the ubiquitous finger-point for emphasis, followed soon after by the fist pounding in the palm. These gestures certainly have their place, but within a much larger palette. Simply watching people in their natural habitat is always your best reference. One of the better ways 10 animate the body for use with dialogue is With pose-to-pose animation, Listen to the track to the point where you understand the major “beats” of the track—where a character gives, emphasis to a word, pauses, or otherwise changes cadence. Because the body moves a few framies ahead of the dialogue, place your major poses slightly ahead of these beats. When the poses syne up with the dialogue, allow the computer to in-between these poses, and then go back on a second pass to tweak, Now that you've gotten some advice as to how to proceed, animate your character speaking the dialogue you trick-read, Assign the mouth shapes, and chen get the eyes, head, and body moving in sync. Render a test, and then go back and fine-tune the animation until you are, For reference, the final test on the pike CD is called DIALFINAL AVI266 | digital character animation 2 FPF RF PEELE ES ZZEE get the enchapter 10: facial animation and dialogue Conclusion This chapter presented the fundamentals of dialogue animation, The fundamentals are eayy to understand and grasp, bur getting your jons to look good requires some more work. Now itis time for you to practice, practice, and practice some more. You should record some tracks of your own and animate some more characters, If you are shy in front of a microphone, there are many ays to get dialogue besides hiring a voice actor. One way is to find some dialogue worth using it an old movie. Pick a scene from your favorite movie and animate a character speaking the dialogue. Of course, if you want to distribute such a film commercially. you run into copyright problems and need to get all sorts of permissions. For practice use only, however, it's not an issue. Another way is to record a normal everyday conversa- tion and animate that. Wherever you obtain your dialogue, remember to have fun when animatingsBulpeay pa}sab6nsSuggested Readings Anatomy Books Anatomy Coloring Book by Wynn Kapit, Lawrence M. Eon ISBN: 0064550168 Great for learning all about the muscles, skeleton, and their connections, Ie a hands-on exercise that helps you learn how the body moyes and works, It also has every part of the body separated, so you don’t see a shoulder area with al! the muscles at once. Bottom line: Get this book and study it! You will become a betrer animator and create better skele- tal setups Anatomy for the Artist by Jeno Barcsay ISBN: 07060716900 Amazing drawings of both the deep-layer muscle and superficial-layer muscles broken down with the skeleton, Lots of drawings that show how the skeleton moves underneath the skin in motion, Melloni’s Student Atlas of Human Anatomy by June L. Melloni (Editor), Ida G. Dox, H. Paul Melloni, Biagio John Melloni ISBN: 1850707707 This book offers pages of each muscle, its origin of attachment, the insertion on the skeleton, and its action. You can easily use this book to set up complex relationships between your skeleton and your deformation tools to create the muscle movement that happens when Joints flex.digital character animation 2 Albinus on Anatomy: With 80 Original Albinus Plates by Robert Beverly Hale, Terence Coyle ISBN: 048625836X_ Great reference, plus just a beautifial book of illustrated plates. Orders the muscles into three layers, Each section starts with a skeleton, and then a first, second, and thitd order of muscles. Color Adas of Anatomy: A Phorographie Seudy of the Human Body by Johannes W. Rohen, Chihiro Yokochi, Elke Lutjen-Drecoll ISBN: 0683304925 If you need a reference for realistic textures of skin, bones, and insides...look no further. There are many color photos of cadavers and body parts. We bought it mostly for the texture information, but there is alo plenty of other good information in this one. Facial Expression by Gary Faigin Watson-Guptill ISBN 0823012685, ‘This book delves deeply into the structure and musculature of the face as it applies to creating facial expressions. This book is a must for those: wanting to create realistic facial anima Animation Books The Mllusion of Life - Disney Animation by Fank Thomas and Ollie Johnston Hyperion ISBN: 078686070 Written by the ultimate Disney insiders—two of the original "Nine ‘Old Men"—this bible of animation has become a classic It explains Disney animation from the ground up and is very complete. There are original sketches, fipbooks and explainations of how memorable movie sequences were made. Though it’s a book written by pencil animators, the techniques described are those used by every animator—including those who animate with pixelsappendix a: suggested readings Zartoon Animation by Preston Blair Walter Foster ISBN: 1560100842 Cartoon Animation is without 2 doubt one of the time honored classics of animation. This is a reprint that has been expanded and repackaged into a smaller format book that can aetually fit on your bookshelf. For years, this was the ONLY book on animation, and Preston Blair is one of the masters, having animated for both Tex Avery and Disney durin the golden age ‘The Animators Workbook by Tony White Watson Guptill ISBN 082300 An award-winning animator offers a complete course on the principles and techniques of drawn animation, covering every aspectof the process. The book covers all of the basics of animation, from dialogue to walk- ing to squash & stretch and much more, This book has become a classic and it is a great book far any animator, tegardless of medium, Creating 3D Animation ~ The Aardman Book of Fidmaking by Peter Lord & Brian Sibley Harry N Abrams ISBN 0810919966 This book expl puppets. Ie was written by the people at Aardman, who have done some of the best clay animation ever, While it is grounded in stop motion techniques, many of these can be translated to computer animation. »s how to do real world 3D animation using cl Tinning For Animation bby Hatol Whitaker and John Hales Focal Press ISBN:924051310X An excellent book that covers the subject of timing. The book deals with timing of characters as well as inanimate objects and effects such, as fire Computer Facil Animation by Frederic Parke, Keith Waters AK Peters ISBN 1568810148 This book, while very technical, was written by one of the pioneers of facial animation. It has a number of techniques that still apply todaNumbers & Symbols 3D animation Gales, 8 we a character staging, 160 A A Bg’ He. 74 aceent pose, 263 aceenories and clothing, 12-13, setion, line of, staging characters, 155-156 action, secondary, movement snd, adding ‘ie Lops, 29 pup npr. fh realism to handy 1-87 shouts, 154-155, ‘scight through motion, 177-179 advanced deformation tools, 136 anatomy Ips. 52 lew 33 tees 54,8091 shoulders 51 sertcbre. 52 wget as facial expression, 246 ie, 168 cnimaig Madu, 262°265 coin tage 244 Index igang sie, 244-208 sg ma a smophing, 298-239 reading the rah 260-262 flor sacks, 202-203 head nk of eyes, 253-256 dination ep 251-254 ein 256 morement. 179 ‘ukipaon, 190-193 nt 180185 ys, 198.19 and dag, 182-183 a, 199-200 ea and flow eb, 199-198 ronda) ation, 1419S dint and nts, 181-182 stands, 185-190 shipping. 231 sneaking, 222 nding character rt sented po tions, 200-201 dendoping gol ene of 169-170 sve mich ws, 168 pres of 166168 nes 70175 pe chee, pst and cry, 176d suggesig wig ah, 177-179 walking Fourlogo, 233-234 ‘ein en gn, 240-213 timing, 210 anticipation, movement and, 190-195, appeal of characters, 4 plications limitations effect om des decsions + ‘metaball implementations, 44 support for modeling hase 22 suming and eunes, 1704175 lope she, 175176 sth ad taj, 176-177 ares, movement and, 180-181 rnin with na sacks, 18 Toke in walking, 209 articulated points, cha tists, approsc +s ect meth deformation, 133-136 anges, 244 wrighted, mesh deformtion, asymimetrical ey fdiences rods appeal of characters. 4 sudo, animating Kp-tyne, 256-257 mouth positions, 258-259 reaing the track, 260-262 frowning logue, 297-288 automatic stretching, squash-andestretch, 190 Avery, Tex, 107 Brapline carves, 24,55 balance, posing characters and, 180- 181 base, time, See Framenper-second beveling, 26-27 Biézier curves. See linear curvesbiceps 54 DiRail operation, 36 Blend operation, 39 blend shape morphing. See ‘mulkiple-target morphing blended wrists, modeling hands with, #182 blends, NURBS, 43-44, 189 like logue and, 264 eye 253-255 Blob Modeler, metaballs and, 44 bobs, head. 264 bodies hea cing 1241S proportions, 8 when to animate 262-263 body language, dialogue and, 265-266 body skeletons, 129-131 bones, See skeletons Booleans i polygonal modeling, 33- au boundary surfhees operation, 36 broken hierarchies, walking and, 211- 212, 226-228 ‘brows ‘cxpresons, 28 ‘modeling, 110-111 posing, 202-243, boilding segiente characters, 120-121 skeletons, 124-132 c ‘Gleulated iaverse Kinematics UK), 127-128 camera placement and staging, 19- tat cardinals, 24,34 cartoon gloves, advantages of, 11 cel animation inating moun and bodes, 262 cycles, 198 to 19 chains, inverse kimematies (IK), 125- ro ‘hamfers, NURBS surfaces, ‘ya tt changing, curves. 173, ee shapes 0 porta emotiom, 910) ‘Choplin, Chart character animation advanced deformation took, Ig seletonn, 128-132 elaen,135.18 Shee asigament, 133-134 coelopes 16-134 berries and 18 lances, 140-181 manipula rachis, 118-19, 2328 soephing 142 pve and join reson, 1196122 tegen chante 169 ‘pe defrmion,t-142 tere ype 137-139 ‘sgh signet 134136 charactor design, & {approaching aan ais categorie, 67 clothing and accesories, 12-13 tvetng simple characters exercise, 1 digiuingseubeues. 18-19 ce she apd se, 9-10 feet 10 and 1 teal and body proportions 9 ela 15 sh Spy 16 photographing tur 7 Svcching on paper 4 eames cules mo computes, 17-19 ing mags erences 15 character modeling, 21 tui 17 st, 34 on 6-82, 654 segmenting 5-58 shtao 0,58 hand 69 dln reason, 84-87 xan ie of 0 fest of 71-73 rom pont 75-78 mtb, #384 icone, 7675 with NURBS afc 78-82 aad 89 ‘cme fom 102 Soong od, HIS oye, 106-117 Beis, 4.95 ten fs, 98-94 pated, 98 ‘goal, 103-106 pagons, 34-95 fete of 9092 toh, 112 tongs 13 ‘op dal deg 95-102 meth 486 [NURBS method ond 43-44 chamfer 44 ‘ames om sui, 42 maton by sui, 4 files, 44 inning, 43 Tein 0 cues om sic, 4 patch method, 34 ating paces fom cares 3638 elses 35 ‘aching pte tse, $940 needed pes 41 polygonal method Boole, 33.34 ede lop, 29-30 len 26°28 polo, 26.28 ating sine 25 uhdsion suf, 30.33 serie, 26.27 statis, 25 sure pes ‘sing and mating, 25 NURBS, 24 ah, 28.24 olga 22.28 charncters posing felon, 150.184 powetonpore, 6 aighhoad, 147 poet, 152 i 153 th lu sak, 148-130 sample segmented, 120-121 single-skinned, hierarchies, 122 squashing-and-stetching. 187-190 sexing ‘camer placement, 159-161 ne af aon, 198-156 on than one, 157-158 soe hte, 154-155 standing from seated postions, 28 201 seized, upper fice poss, solame, 186 cheeks, muscle movement, 92 choice of lenses, effects upon anima- tion composition, 160-161 choosing. Kimemates pes for skeletons, 128 surfaces for character modeling, 22 topologies, 9-102 clavicle, St claws, adding to fingers, 85 clay, modeling with, 15-16, 98, ‘Clay Stadio Pro, metaballs and, 44 lipping corners, polygons, 31-32 6clothing and accessories, 12-13, clusters evribed, 142 rumeric weighting 138 collapse operation, 26-28 ‘combining pose-to-pose and straight ahead animation methods, 7 communication among, characters, staging, 157-158 complex expressions, 249-250 ‘complex heads, 102 ‘composition of animations, effects of lenses upon, 160-161 computer animation holds 199 See ao animation computers timing on. 168 tearing septs imo, 17-19 connecting. ‘character and skeletons with ‘herrchies, 118-128 heads to bodies, 4-115 constraint, inverse kinematics (IK), 7 corner clipping, polygons, 31-32 Ccorrugator muscle, 91-92 Coyote, Wile E.. 169 creasing skin, adding to hands, 87 creating ed fp. 29 cys 107 foal exprenions, 46 trois, 248 ‘omiplex exprsions, 249-250 amery ad, 248 sets kinematic (IK) cans, 125 patches from curves 8-38 eas hands, 84-87 short, 154-155, simple segmented chancters, 2-121 seletons, 128-192 eight through motion, 177-179 crowds, staging, 158 ‘Beale See linear cures cresting on NURBS ses 42 ites fom, 3638 curve surfice operation, 38 cing. 173 {in sfiwate animation, 170-47 near 173 Jong to on NURBS sutaces. 43 posc-to-pone asimaion, 174-175 step. 173 types 20-24, 435 63 cushion po clothing and accessories cut/connect operation, 28 cycles movement and, 198-199 walling, 213-228 ‘ylinders, building simple ‘segmented characters, 120-121 D Dally Duck, 7 decisions in design, 45 Aefaak positions, joints, 127 Aeformations “yg surface, NURBS srfices 4 slobal. 130 lace. 190 mesh, skeletons and ‘svoned donation ras 136 wisi leony, 128-132 user, 135, 141 it agent 123-134 lonkpe 134-138 iis and chancter cnimation, 118, aie 290180 mailing hie 118.119, 123-128 oping. 42 ots joe rotation, 119 Segoe hare, 140 spn efoto, 141182 Sa pe, £37139 ih ments 14-156 seletons, 188-189) Sspnh-anbsercech, 188-198 toe, 136 deforming. pateher 139 polygons, 137-138 See ao modeling posing degree of curves, 23 deltoid, 54 depressor labilinferioris ‘muscle, 90 designing characters, 8 ‘poaching a as ats, 4-5 categories rel, 62 cothing and accessories, 12-13 Aiguizng sculptures, 18-19 ye shape and size. 9-10 feet, 10 fad 1 head and body proportions 8-9 in cay, 15-16 horographing seulptures, 17 Shetching on oper. 14 transfering sculpts nto computes 7-19 ung images as references, 15 dialogue animating. 262 ‘ody hone, 265-266 comonsnt, 264 ye, 264 ead sin ad, 264 mouths 263-264 seueh 268 reconing ip-sync. 25 timing and, 166 digitizing sculptures, 18-19 cligitizing tablets, eracing images a references, 15 Aicect as . leformation, 133-134 jection of eyes inks, 256 deag, movement ans 182-183, drawing ‘efernces for design harites, 15 See ako shetbing ears, pol edges ke loops, 2M i polygonal modeling, 26-28 siting curves, 173, tye shapes portray emotions 3 sfoetors constas, 127 ‘manipulating vere Kine: (K) chain, 126-127 hang eye shapes port, Es aricters saying, 1572158 ‘envelopes: mesh deformation, 134-138 ‘exercites im charactor design, simple characters, 19 ‘exponire sheets and timing. o7= 168 ‘expressions cll, 246-247 ‘rows, 288 complex expressions 249-250 seaging, 157-158 symmetry and 248 extensors, 54 ‘external eyes, modeling, 107-109 extruding patch modeling, 36 7 polygonal modeling 20-27eyebrows riodeling 10-111 ssowement, 92 See ao brows eyeids, 107 eyes symmetrical, 10 bk, 283-256 dlilogue and, 254 direction of movement, 251-253 sternal, 107-109 internal 16-107 modding, 106-111 souscle movenent, 92 posing, 242-243, shapes and ies, 9-10 F few suman, 251 “ging tage, 244 Wins pes 255-256 at fl expressions Aioge, 262-268 ton fee mame, seas Fiery, 256.259 ‘ening sade, 244-245 modeling orp ae a 219-241, 245 oping. 238-259 vein 256 reading te tack, 260-262 expressions, 246-247 comple cxprsions, 249-250 suaging, 137158 yrcry and, 248 movement of, 92.95 acs of, 90-92 sie and placement of evs on seructue of, 0192 topalagy and modeling, 95-102 fear as facial expression, 207 feet ‘ulin, 151 Jcepiny on gonind when walking, 210213 Proportions, 10 role ia walking, 207 formu, 53 Sila, 53 field of view (FOV), 160 filles NURBS sures, 44 ‘operations. 9? humerus ‘lms, cening Seveloping ood verse, 169-170, Tow uch rouse, 168 fone thing a tie, 169 HOC OF 16168, fingers ‘esi, 71-73, nucle, 85-86 al 85 number to we for sharieters, 70 webbing beewsen, 87 first-degree curves, 4,34 EK. See forward kinematics ‘esh, See skin Aexibiiy oF hands and fingers, 71-73 ‘exible skin, head modeling, 94.95, exors, $4 our sacks animating, 202-203 posing charac with, 148-150 focusing cameras, character staging, 160 follow-through, movement and, 195- 198 footstep generators, 211 force, movernent and, 182-183 forearm, $1 forward kinematies (FK), 118, 124, 8 animating walk eyes with ‘oor-legged walking, 233-234 four-sided patches, FOV. See field of view frames-per-tecond and frontalis muscle, 91-92 fanctions, 252 G ‘aaocner, 4 fenerators, fottep. 211 sestures, dialogue and, 268-266 lobal deformation, ‘quatheand-stretch, 19 loves, advantages of, 11 Genes animes 4 goals, jim, 127 Gonzales, Speedy, 265, tides, forward Kinematics walking and, 212-213 H ands modeling. 69 ding velo 6, 84-87 exumining statue of. 70-71 Jesh fom polygon, 75.78 mcabal, 83-04 segrented, 74.78 dk NURBS sis, 78-82 proportions of, 11 skeletons, 132 ‘nash spines, 41 heads animating Winks of es 253-256 ition of ep movin, 251-258 ig 256 bob. 264 had and body proportions 8-9 modeling, 99 comple fms, 102 conmacing hed, 114.125 ey, 106-111 este skin, 94.98 peo fates, 93-94 aude aed, 95 olga 108.106 polos 94-95 metre of 0-92 teth, 115-412 toms, 103 eel ad fel, 98-102 movement dialogue and. 264 posture in walling, 209 28 character averse walking and, 211 ranipulting forward nena (ER), 118, 124 nese inert (IK), 118,124 28 single-skinned characters, 122 hips anatomy. 52 tabince, 151 imation and, 118 2, 226-228 ners hierarchies, 211 posing with lou sacks, 148 role i walking, 207-209 hholds, movernent and, 199.200 horses, walking, 233-234 humerus, StIK, See inverse Kinematics images as references for designing characters, 15 inserting, 26 Interactions among characters, stag- ing, 157-158 {internal eyes, modeling, 106-107 fnverse hierarchies, walling and, 211- 212, 226-228 fnverse Kinematics (IK), 118 ssmating walk eyees, 222-224 chains, 125 manipting, 126.127 consis, 127 described, 124-128 real-time and eaeulted, 127-128 walking and, 210 ‘wen toe, 128, aos, musele movement, 92 joins articulated poins, chains, 126 sffecors ‘mst 127 foe kinrma (1K) cis, 126 po rest postions and gals, 127 sou 125 totaiony and pivots, 119-122 setinglmiatons, 127 Jones, Angie, 231-232 Jumping, animating Dour sacks for, 202-208 K Keep Tighe, 35 “Keeping feet on ground when walk- Ing, 210-213 kinematics ‘choosing type 4 use, 128 forward simating wale yes with, 212 orb 118, 124 oh ides, wig ad, 22 rimatng wl cles ith 222224 hain, 125 tans 127 Sear, 1128 manipalting chains, 126-127 relsime aed clade, 127-128 walking end, 210 knowing your characters, 4 Inwebles, adding to hands, 88-86 L ing See dag language of body dialogue and, 265- 266 lathe operations, 36 ladssieus, 54 latices deformation, 190) described, 140-181 lens anatomy of 53 balance 131 posing with Hour sacs, 148 role in walking. 207 lense, effects upon animation com position, 160-161 levator lab superioris muscle, 90 levator palpebrae muscle, 91 ids, modeling, 107 lim soins. ering 27 ‘offre fees on design ‘decions. 4 line of action, staging characters, 135-156 linear curves, 24, 34, 173 Tinks, 214 lip-syne, animating, 256 ‘mouth poisons, 258-259 recording lege, 297-258 lips, muscle movement, 92 lofting, 36, 39 Took at fanetions, 252 oops, edge, i polygonal ‘modeling, 29-30 lower-face poses, modeling for ani- ‘mation, 240-241 ‘naling design decisions, -5 smanipolating face, ers, 24-245, Iieatchies, 118, 1246128 mapping, sticky, 47 apt texture punting character features, 94 segmenting characters, 36 Marx, Groucho, 205 matching and mixing modeling surface types, 25 ‘mesh deformation, skeletons, 91 “uvanced deformation tools, 136 boiling skeletons, 128-132 clusters, 135,141 lie asignment, 133-144 envelopes, 5-138 Iierarchics and charac ‘imation, 118 Ince. 140-141 smanspolaing hierarchies, 118-119, 123-128 morphing. 142 Phot and join rotation, 119-122 segmented characte, 140 spline deformation. 141-142 surf spe, 137-139 weight sigaments 13136 setaballs siting, 40-46 models hands 83-84 problems with, 46 ‘MetaReyes, metaballs and, 44 ‘mixing and matching modeling sar- face types, 25 ‘modeling characters, 21 “baie 49, maces, 54 ani, 60-82, 65.67 segneing, 55-58 shen $0, 53 eyebrows, D-II hn 69 dig rea 84-87 examining sre of, 70.71 Aesiiy of. 71-73 fom pigons, 75-78 cabal, 83-84 sth NURBS sues, 79-82 hea 99 ‘cpl fms, 102 cnet bd, 1411S eyes, 106-111 ese sie, 94-95 pte fates, 93-94 path acd, 98 polgoml 105-106 polo, 94.95 rte of, 9092 tech, EAI ramus 113 tapiogy ad fl dling, 95.102 smetaballs inaing, 146 oben with, 46 morph ange for sniation, 239 buerfie pe, 240-241 ppt pose, 242-249 reeks, 11S NURBS metho Wend 43-44 chamfer, 4 nes ae, 42 ‘Afmaon by sui, 44 filer, 44foie #3 tang 10 ces on sie, 43 patch method, 4 ring pat fo cae, 36-34 rel ues, 35 ‘aching paces ee, 39-40 Atnesded pes 4 polygonal method. 26 ‘olan, 33-34 ge ops, 29.30 es, 26:28 polgons 2628 ‘eating snl, 29 suison suf 30.33 series 26.27 sateen 25 surface types ud muching 25 NURBS, 26 cone lower pres, 240-241 snub, 289 single, 238 upper ft pois, 242-248 motion, 179 snccipation, 190-193. das 180-181 cresting weight through, 177-179 eyes, 198-199 fingers, 71 fone and drag, 182-188 rok, 199-200 ‘overlap and follow-through, 195-198 overshoot, 193-198 secondary action, 194-1 slowins and slow-ous, 181-182 squash-and-aretch, 185-190 thumb, 72-73 mouth pole 0. 97-98 posing, 240-241 positions lipsync, 258-259 ome, 263) when to animate 262-263, ‘camer, character waging, 156-161 creating weight through, polygons ins 253-256 siron of 281.285 ‘ls, 10 fceu 95 ead dilge and, 264 smulijoinced object, 183 snuck 92 terminology wed in, 179 utiption, 1906198 ‘180181 sys, 198-199 fone ed dag, 182-183 a, 199-200 swe ad fila, 195-198 eves, 193-198 senda aon, 194-198 slosh nd soucouts 181-182 squashoand seh, 185-190 ‘mulijointed objects, movement, 183 multiple target morphing, 142,239 muscles faces 10-92 adelng 54 ‘necks, Connecting t0 bodies, Tes negative sliders, 248, Non-Unitorm Relational 1s. See NURBS surfaces ie weighting, clusters, 135, NURES ‘ends, 139 curves, 37 modeling Hod, 4-44 chan 44 ating curves on gis, 12 ‘Lfomaton by oui, 44 fil, 44 Sing, 43 ‘ing ees on si, 4 speraton in, 37-38 nol hand, 78.82 rodlng seamls bodies, 64.67 ‘sbjects muljinted, 183 pointy. am eyes. 253 operations 'NURDS modeling, 37-38, patch modeing, 36-38 pessoal modcing 26.28 orbiculais ors muscle, 91 ‘oribiculars ocul muscle, 91 ‘overlap ttl anion, 2 movement and 195-198 overshoot, movement and, 193-198 Pp DPalating character Features, ‘animated texture maps, 94 paper, sketching characters om, 14 partially segmented bodies, ‘modeling, 38 paste-on features, head mo. 4 patch modeling, 36 head, 95 morphing, 238 pches from curves, 3638 regula surices, 38, seiching patches tagether 39-40 surfiees, 23-24 three-side patches. 41 patches creating from curves, 6-38 determing 139 paths in software animation, 176-177 pectoralis major, S4 Photographing sculpeares, 17 pivots, joint rotation and, 119-122 placement of eyes on faces, 9 placement of cameras and aging, 159-161, points, articulared, chains, 126 pointy objects in eyes, 253, poles a the es, 98-99 the mouth, 97-98 a the tp ofthe hend 6-97 polygonal modeling, 26 Booleans, 33-34 ges 26028 edge lps, 29°30 wes, 109-106 morphing, 238 polygons, 26,28 simple. 29 subdivision wrices, 30-33, surfaces, 22-23 vertices, 26-27 polygons seforning, 137-138 head modeling, 8495 dn polygonal modeling, 6-28.modeling In, 75-78 seams odes, 60-62 subdividing, 1-33, portraying emotions by changing eye shapes, 9°10) pose test, 174 pose-to-pose animation, 146, 174-175, poses chincter, fone, 150.151 pasetpoe, 146 sight ahead, 147 symmetry 152 eight, 153 with lsc, 148-550 cushion, 253, pushing ders 244 ‘quashand-srech, 187 See alo deforming. modeling positions of rest, joints, 127 Potato Head, Mr, 9,93 preconceptions of characters, 7 Program lineation effects on design decisions, 4 programs retaball implementations in, 44 suppore fr moog characters, 22 timing sed aus, T0175 ape shee, 178-176 ath day 17617 proportions feet to Dane 1 heads and bodies 8-9 puppets, modeling with, 93 Pushing poses, sliders, 244 QR ‘Guadraceps, 51 radial head. 8 radius, $1 Rankin-Bass Christmas specials, 93 reactions among chacecce staging, 157-158 reading the track, lip-ryne, 260-262 real-time inverse kinematics (IK), 27-128 realises adeling o hands, 84-87 character designs, 6-7 poles, at the mouth Aaloguelip-syne, 257-258 reading the rack, 200-262 references, drawings a5, 15 polygons regular surfaces in patch modeling, 38, rest postions joints, 17 revolve operations, 36 fisorius muscle, 90 Roadrunner, 169 root, 125 rotations, joint ps 119.122 Radolph the Red-Nosed Reindeer, 98 ruled operations, 36 ‘unning, animating, 29-280 s ‘Sacks of flour, 302-203 psig characters with, 148-15) walking, 215-216 sadness as facial expression, 247 Santa Claus Is Coming 19 ‘Tow, 93 satorius, 54 scaling, equash-and-stretch, 188 seanning images as references, 15 seapala, St Sealey, des sculptures sharers in clay. 15-16 sand, ing characters with, 16 Aligiizng. 18-19 photographing, 17 traniersing into computers 17-19 seamless characters, modeling, 60 with NURBS surfices, 61-67 swith polygons, 60462 seamless wrists, modeling hands with, 78.80 seated characters, standing from, 200- 201 second-degree curves, 14, 6 secondary action, movement and, 194-195 segmented bodies, 85-58 segmented characters ‘ual siropls, 120-121 mesh deformation, 140 segmented hands, 74.75 Kimematics types fr shletons, 128 suas for character modeling, 2 topologies, 9-102 setting joint limitations, 127 shape weights morphing. See multi= ple-target morphis shapes of eyes, 9-10 shin, 83 shoulders stony 31 balance, 151 posing With four sack, 148 sole in walk silhowetter, 151-185 imple characters, 19, 120-12 single-skined characters, Iilerarchies, 122 Single-target morphing, 238 sizes of eyes, 9 selecons ‘teh deformation sd dfvation tly, 136 fling ston, 128-132 fen 185, 140 is sgrment, 1134 cemelpes ISH 138 iris a hacer awimaton, 18 ais, 140141 manipaating hehe 118-119, 123-128 smrplang, 142 adj tion, 119-122 famened characte, 140 pln defensin, 151-182 sure: pe, 137-139 eile sgnmon, 14136 modeling 30.53 sushi eich deformation, 188 er on paper, 14 See ako drawing skin, 48 reasing aling to hans, 87 head modeling, 4-9 skinning, 36 skipping, animating, 231 sliders rmanspulating fice, 26-235 negate, 245 slowing and slow-outs, movement and, 181-182 sneaking, animating, 252 Softimage, metaballs and, 44 sottware limitations, ect on devin decom metal implemeneanons i, 4 support for modeling characters, tung an 170 ues, T0175 ape het, 175-176 ths god ajo 176-177 sounds, animating lip-sync, 256-257 mouth postions, 258-2 reading the tack, 260 reconing dialogue, 257rygomatic major spheres bullingsianple segmented ‘haractes 120-121 metals 45 spines anacony S2 bubnee, 151 posing with flag sacks, 138 posure in walking, 209 tole in waking, 207-208 spline deformation, M1142 squash-andestreteh, 126, 185-1 th, blinks and, 255 ig characte ‘camera placement, 89-161 lige of action, 158-156 more than ome, 157-158 strong silhouettes, 154-155 standing characters from seated pe ions, 200-201 step curves, 173 sticky mapping, 47 stitching patches together, 39-40, stop motion, animating, mouths and bodies, 262 straight-abead animation, 147 stretching, automatic, ‘squash-and-stretch, 190 stylized characters esigns,7 upper-fce poses 243, subdividing polygons, 31-33 subdivision surfaces in polygons! ‘modeling, 30-33, supinator, 54 surface types snes deformation, 137-138 ‘modeling charcte, mising and mating, 25 NURBS, 24 path, 24-24 polyol, 2 regula in patch modeling, 35-38 subuvsion in polygonal modeling. 30-53 surfaces, NURBS Blend, 43-44 char, 44 creating curves 06,42 Seformstion by surice, 44 filler. lotting to curves on, 43 modeling Inn, 782 seams bee, 64-67 srprize as facial expression, 47 sweeping, boad details, 98 symmetry expressions, 288 ing characters, 152 muscle ‘able, dy references, 18 talons, adding to Bingers, 85 targets sgning. 244 morph lower poses 240-241 model jor animation, 239.240 epee poi, 202.203 teeth, modeling, 111-112 tendons, adding vo hands, 86 testing skeletons betore production, 138 tests, pose, 174 texture maps pining character faces, 4 serpnenting characters, 56 third-degree curves, 35 three dimensions. See 3D. three-sded patches, 41 thumbs lexbiity of, 72-73 tibia, 53 time base, See frames-per-sscond antapaton, 190-193 fae 10181 ‘yc 198-199 dveaping god seme of 169-170 foc and di, 182-188 Tracing imager a Is, 199-200 sda and fl, 195-198 vent, 193-198 sendy ion, 194-195 loin nd soot, 181-182 quashnd-seth, 185-190 logue, 166 expire sheets, 167-168, ames-per-second, 167 fon computers, 168 fone thing a a ie, 169 proces of 166-168 sofieare an 170 canes, 17061 ‘ap sheets, 175-176 sand rset, 1760 soggesting weight with, (77 walking, 210 179 topologies choosing, 99-102 ead ad fice, 95-102 ‘Tay Stony. 13 lacing images as references, 18 tracks, reading, lip-syac, 210-262 trajectory in software animation, 176- "7 transferring sculptures into ‘computers, 17-19 1g See maninalating ‘riangularis masele, 90 of heads, blinks and, 255 wins, 182 ‘upper-face poses, modeling for ani- mation, 242-243 vertebrae, 52 vertices in polygonal modeling, 6-27 videos, timing, 166 Uevelping good ers of 169-170, how much to use, 168 one thing a aie, 169 proce of 166-168 ‘volume, characters, 186 wz walking snimanng ‘ves 2034228 fouled, 235-234 deping jet om grvend, 210-213 four sacks 215-216 hoses, 233-238 mechanics 206 sms, 209 feet and eg, 207 ips, ping and sides, 207-208 pot hla pine, 209 Wayne, John, 205 webbing, adding berween fingers, 87 weight igaments, mesh deformation, 134+ 136 smameri, cheers, 139, posing characters and, 153, suggesting with timing, 177-179) West, Mac, 205 ‘wrinkles, adding to hands, 87 wrists ended, 81-82 seamless, 78.80 zygomatic major muscle, 90informiT Www. informit.com YOUR GUIDE TO [IT REFERENCE New Riders has partnered with informlT.com to bring technical information to your desktop. Drawing from New Riders authors and reviewers to provide additional information on topics of interest to you, InformlT.com provides free, in-depth information you won't find anywhere else. DNaisd Keep your edge with thousands of free articles, in-depth features, interviews, and IT reference recommendations—all written by experts you know and trust. Online Books Answers in an instant from InformIT Online Books’ 600+ fully POWERED By searchable online books. Safari ras Review online sample chapters, author biographies, and customer rankings and choose exactly the right book from 2 selection of ever 5,000 titles. New Riders wwmnewriders.comPak iapit te | CHARACTER AUREL Tatanen 2 All-New, Rewritten, and Updated Version of the Groundbreaking Bestseller [digital] Character Animation The cross-platform CD-ROM Before any CG artist can become a great animator, he or she wean includes: needs to have a strong understanding of the fundamentals of . the art behind the animation. [digital] Character Animation 2 1B Multiple walk cycles Volume I: Essential Techniques provides the foundation needed (AVI or QuickTime) erated characters in ¢o create professional-quality, compute Lip sync examples 2D and 3D. This classic book covers conventional techniques, (Avior Quicktime) such as walk cycles and lip synch, and shows you how to apply these techniques to your CG characters. This non-software- @ Models of 3D characters specific book provides examples that can be done in various programs, along with a discussion of tools and techniques, common to all major animation programs. George Maestri has worked asa writer, director, and producer in both traditional and computer animation [digital] Chasacter Animation 2 Volume I: Essential Techniques for sich companies a Nickelodeon”, Warner includes all new character animation information: Brothers”, Disney®, Comedy Central, Film Romany MGM™, ABC®, CRS*, and FOX*, among others. He | Basic design concepts—how to think about your ‘was the original animation producer on the Comedy cies ae Deters ou! Bean AD cnsats I ‘Ceneal series South Prk” and was nominated fora IH New techniques for modeling your character's head, Cable ACE award for writing onthe Nickelodeon body, and hands series Reko's Modem Lif, He has writen sever Brand new information on skeletons and mesh deformation books and numerous articles on computer animation for magazines such as Digital Focused discussion on the fundamentals of character Magic, Computer Graphics animation, locomotion, and dialogue Word, DV, and New Malis, ‘Whether you're creating work for television, advertising, movies, multimedia, or gaming, [digital] Character Animation 2 Voinme I: Essential ‘Techniques shows you how to bring your imagination to life. This first volume of the two-volume /digital] Character Animation techniques collection—the art behind the animation—comes from the industry authority: George Maestri $50.00 USA / $74.95 CAN £46.49 Net UK (inc of Vat) ISBN 1-56205-090-0 New Category: Design/Graphics—

You might also like

• Course Syllabus: What You Will Learn in This Course

  

  No ratings yet

  Course Syllabus: What You Will Learn in This Course

  11 pages
• Mastering 3D Animation 2nd Edition Peter Ratner Latest PDF 2025

  

  No ratings yet

  Mastering 3D Animation 2nd Edition Peter Ratner Latest PDF 2025

  74 pages
• Mastering 3D Animation 2nd Edition Peter Ratner Instant Download

  

  100% (9)

  Mastering 3D Animation 2nd Edition Peter Ratner Instant Download

  71 pages
• Lecture 1-Animation

  

  No ratings yet

  Lecture 1-Animation

  23 pages
• Mastering 3D Animation

  

  100% (6)

  Mastering 3D Animation

  369 pages
• Tricks To Animating Characters With A Computer: Course M A N ' '

  

  No ratings yet

  Tricks To Animating Characters With A Computer: Course M A N ' '

  3 pages
• 3D Animation and Modeling

  

  No ratings yet

  3D Animation and Modeling

  54 pages
• SLqm6 CE-1 Notes

  

  No ratings yet

  SLqm6 CE-1 Notes

  6 pages
• 10 1 1 1 1695 PDF

  

  No ratings yet

  10 1 1 1 1695 PDF

  26 pages
• Character Animation in 3D Focal Press Visual Effects and Animation Series 1st Edition Edition Steve Roberts Full Digital Chapters

  

  100% (2)

  Character Animation in 3D Focal Press Visual Effects and Animation Series 1st Edition Edition Steve Roberts Full Digital Chapters

  92 pages
• Animation - 03

  

  No ratings yet

  Animation - 03

  10 pages
• MST Powerpoint

  

  No ratings yet

  MST Powerpoint

  46 pages
• Animation (Compatibility Mode)

  

  No ratings yet

  Animation (Compatibility Mode)

  35 pages
• Topic 3 (5) - Animation

  

  No ratings yet

  Topic 3 (5) - Animation

  39 pages
• 5 Types of Animation

  

  No ratings yet

  5 Types of Animation

  3 pages
• 3 D Human Modeling and Animation Second Edition Peter Ratner Online Version

  

  No ratings yet

  3 D Human Modeling and Animation Second Edition Peter Ratner Online Version

  83 pages
• Mesmer XSI Illuminated Character 2004 Edition PDF

  

  100% (1)

  Mesmer XSI Illuminated Character 2004 Edition PDF

  390 pages
• Character Building in Blender

  

  No ratings yet

  Character Building in Blender

  49 pages
• 3D Studio MAX 3 Fundamentals

  

  No ratings yet

  3D Studio MAX 3 Fundamentals

  97 pages
• Unit 3

  

  No ratings yet

  Unit 3

  59 pages
• Ch. 09 Computer Animation

  

  No ratings yet

  Ch. 09 Computer Animation

  11 pages
• MC 8 PDF

  

  No ratings yet

  MC 8 PDF

  49 pages
• 2D 3D Anim

  

  No ratings yet

  2D 3D Anim

  7 pages
• Fundamentals: Character Animation

  

  No ratings yet

  Fundamentals: Character Animation

  11 pages
• ARVc 4

  

  No ratings yet

  ARVc 4

  46 pages
• Submitted TO: Submitted By: Mca Department Chintu Gandhi Roll No 157 Class-3 Sem

  

  No ratings yet

  Submitted TO: Submitted By: Mca Department Chintu Gandhi Roll No 157 Class-3 Sem

  10 pages
• MB Tutorials1

  

  No ratings yet

  MB Tutorials1

  226 pages
• 3D Animation Reoprt - 216304026

  

  No ratings yet

  3D Animation Reoprt - 216304026

  13 pages
• Module 6 Introduction To Character Animation

  

  No ratings yet

  Module 6 Introduction To Character Animation

  11 pages
• 20160307090354lecture 1 Introduction To Animation

  

  100% (1)

  20160307090354lecture 1 Introduction To Animation

  56 pages
• Foundation Flash Cartoon Animation 1st Edition Tim Jones Newest Edition 2025

  

  No ratings yet

  Foundation Flash Cartoon Animation 1st Edition Tim Jones Newest Edition 2025

  126 pages
• Manual 3dsmax 01

  

  No ratings yet

  Manual 3dsmax 01

  782 pages
• Render Engine Cutting Edge Techniques in 3d Modeling and Animation 47iqwiw9xcm4

  

  No ratings yet

  Render Engine Cutting Edge Techniques in 3d Modeling and Animation 47iqwiw9xcm4

  7 pages
• Animation Principles: Prepared by Razia Nisar Noorani

  

  No ratings yet

  Animation Principles: Prepared by Razia Nisar Noorani

  60 pages
• Module 1 Fundamentals of 3D Animation

  

  No ratings yet

  Module 1 Fundamentals of 3D Animation

  20 pages
• Computer Animation

  

  No ratings yet

  Computer Animation

  42 pages
• MMO Animation Techniques

  

  No ratings yet

  MMO Animation Techniques

  20 pages
• Module - 6

  

  No ratings yet

  Module - 6

  10 pages
• Digi Notes Unit5

  

  No ratings yet

  Digi Notes Unit5

  20 pages
• Animation

  

  No ratings yet

  Animation

  7 pages
• Basics of Traditional Drawing

  

  No ratings yet

  Basics of Traditional Drawing

  54 pages
• The 3D Animators Toolkit - From Concept To Creation 01

  

  No ratings yet

  The 3D Animators Toolkit - From Concept To Creation 01

  72 pages
• Debabrata Saha Ca1 Ms

  

  No ratings yet

  Debabrata Saha Ca1 Ms

  9 pages
• Virtual Human Animation

  

  No ratings yet

  Virtual Human Animation

  9 pages
• GameProg Lect5 3D Modeling

  

  No ratings yet

  GameProg Lect5 3D Modeling

  76 pages
• Character Animation in 3D Focal Press Visual Effects and Animation Series 1st Edition Edition Steve Roberts Get PDF

  

  No ratings yet

  Character Animation in 3D Focal Press Visual Effects and Animation Series 1st Edition Edition Steve Roberts Get PDF

  152 pages
• Maya Rig

  

  100% (2)

  Maya Rig

  329 pages
• Character Animation Guide 2023-24

  

  No ratings yet

  Character Animation Guide 2023-24

  9 pages
• Topic 7

  

  No ratings yet

  Topic 7

  41 pages
• 2D Animation-2nd Q-W4

  

  No ratings yet

  2D Animation-2nd Q-W4

  8 pages
• Foundations of Drawing Cartoon Characters For Animation

  

  No ratings yet

  Foundations of Drawing Cartoon Characters For Animation

  119 pages
• Character Animation Crash

  

  100% (10)

  Character Animation Crash

  241 pages
• Art of Rigging Vol I Book - CG Toolkit

  

  100% (14)

  Art of Rigging Vol I Book - CG Toolkit

  129 pages
• Preston Blair Cartoon Animation (1) (041 080)

  

  100% (1)

  Preston Blair Cartoon Animation (1) (041 080)

  40 pages
• 3D Total's Introduction To Rigging 3ds Max

  

  No ratings yet

  3D Total's Introduction To Rigging 3ds Max

  106 pages
• How To Draw Clothing For Manga Learn To Draw Amazing Outfits and

  

  96% (24)

  How To Draw Clothing For Manga Learn To Draw Amazing Outfits and

  348 pages
• (Ebook) Dynamic Reference Sheets ENG - Compressed

  

  93% (59)

  (Ebook) Dynamic Reference Sheets ENG - Compressed

  104 pages
• Perspective Sketching Freehand and Digital Drawing Techniques For Artists & Designers PDF

  

  97% (76)

  Perspective Sketching Freehand and Digital Drawing Techniques For Artists & Designers PDF

  227 pages
• World History - From The Ancient World To The Digital Age (PDFDrive)

  

  100% (34)

  World History - From The Ancient World To The Digital Age (PDFDrive)

  418 pages
• Character Design

  

  88% (66)

  Character Design

  113 pages
• Figure It Out Drawing Essential Poses The Beginner's Guide To The Natural-Looking Figure by Christopher Hart

  

  94% (18)

  Figure It Out Drawing Essential Poses The Beginner's Guide To The Natural-Looking Figure by Christopher Hart

  148 pages
• The Art of Princess Mononoke English

  

  36% (14)

  The Art of Princess Mononoke English

  115 pages
• Anatomy For Artists Drawing Form Pose The Ultimate Guide To Drawing Anatomy in Perspective and Pose (Tom Fox)

  

  100% (20)

  Anatomy For Artists Drawing Form Pose The Ultimate Guide To Drawing Anatomy in Perspective and Pose (Tom Fox)

  297 pages
• Draw Buildings and Cities in 15 Minutes - Amaze Your Friends With Your Drawing Skills

  

  94% (31)

  Draw Buildings and Cities in 15 Minutes - Amaze Your Friends With Your Drawing Skills

  201 pages
• Mythology Book PDF

  

  94% (33)

  Mythology Book PDF

  354 pages
• FORCE - Dynamic Life Drawing - 10th Anniversary Edition (PDFDrive)

  

  97% (29)

  FORCE - Dynamic Life Drawing - 10th Anniversary Edition (PDFDrive)

  234 pages
• Anime Academy! How To Draw Manga Your Complete Guide To Drawing Anime Characters From Heads, Anatomy, and Clothing, To Color Illustrations!-1

  

  85% (33)

  Anime Academy! How To Draw Manga Your Complete Guide To Drawing Anime Characters From Heads, Anatomy, and Clothing, To Color Illustrations!-1

  134 pages
• Art Explained (Art Ebook DK)

  

  100% (27)

  Art Explained (Art Ebook DK)

  108 pages
• Fearless Drawing

  

  97% (58)

  Fearless Drawing

  131 pages
• Portrait Drawing - The Quick Guide To Mastering Technique and Style

  

  100% (20)

  Portrait Drawing - The Quick Guide To Mastering Technique and Style

  115 pages
• Ultimate Guide To Manga Art - 1st Edition 2022

  

  100% (15)

  Ultimate Guide To Manga Art - 1st Edition 2022

  134 pages
• Xiao Weichun - Human Anatomy For Artists

  

  100% (24)

  Xiao Weichun - Human Anatomy For Artists

  328 pages
• Animation Human Body Structure Special Training in Performance Techniques

  

  97% (33)

  Animation Human Body Structure Special Training in Performance Techniques

  776 pages
• Art of Drawing Complete Course

  

  100% (35)

  Art of Drawing Complete Course

  260 pages
• Figure Drawing For Artists Making Every Mark Count-xBOOKS PDF

  

  95% (161)

  Figure Drawing For Artists Making Every Mark Count-xBOOKS PDF

  568 pages
• Game&Animation Human Body Sturcture Hand Drawing TechniqueXiao Weichun

  

  100% (15)

  Game&Animation Human Body Sturcture Hand Drawing TechniqueXiao Weichun

  514 pages
• CDQ Character Design Quarterly

  

  100% (12)

  CDQ Character Design Quarterly

  104 pages
• Imaginefx How To Draw and Paint Anatomy 2010

  

  98% (130)

  Imaginefx How To Draw and Paint Anatomy 2010

  118 pages
• 3DTotal Com LTD Dynamic Characters 2010

  

  93% (72)

  3DTotal Com LTD Dynamic Characters 2010

  54 pages
• Sketching From The Imagination PDF

  

  98% (110)

  Sketching From The Imagination PDF

  321 pages
• Patrick J Jones - The Anatomy of Style

  

  96% (144)

  Patrick J Jones - The Anatomy of Style

  101 pages
• Inventions A Visual Encyclopedia PDF

  

  95% (40)

  Inventions A Visual Encyclopedia PDF

  306 pages
• Japanese Mythology

  

  93% (44)

  Japanese Mythology

  144 pages
• Michael Hampton Figure Drawing Design An Text

  

  93% (14)

  Michael Hampton Figure Drawing Design An Text

  241 pages
• Fundamentals of Character Design

  

  100% (20)

  Fundamentals of Character Design

  306 pages
• Dynamic Web Forms Professional Projects - Premier Press PDF

  

  No ratings yet

  Dynamic Web Forms Professional Projects - Premier Press PDF

  537 pages
• Structural Stiffness Analysis Guide

  

  No ratings yet

  Structural Stiffness Analysis Guide

  3 pages
• Civil/Architecture Workbook: - For Cadpoint Use Only

  

  No ratings yet

  Civil/Architecture Workbook: - For Cadpoint Use Only

  1 page
• SYNERGY CAD CENTRE Mechanical Practical Book

  

  No ratings yet

  SYNERGY CAD CENTRE Mechanical Practical Book

  6 pages
• Material Costing

  

  No ratings yet

  Material Costing

  1 page
• Directly Authorized by Autodesk, Altair, Dassault Systemes & PTC University - The Manufactures of The Software's Itself On Which We Train On

  

  No ratings yet

  Directly Authorized by Autodesk, Altair, Dassault Systemes & PTC University - The Manufactures of The Software's Itself On Which We Train On

  1 page
• Bending

  

  No ratings yet

  Bending

  11 pages
• What Is Catia (Computer Aided Three Dimensional Interactive Application)

  

  No ratings yet

  What Is Catia (Computer Aided Three Dimensional Interactive Application)

  3 pages
• Gear Modeling with CATIA V5

  

  No ratings yet

  Gear Modeling with CATIA V5

  4 pages
• Report

  

  No ratings yet

  Report

  7 pages
• Visual Lisp

  

  80% (5)

  Visual Lisp

  183 pages