Character Creation - Part One: The Concept and Wire Planning
Getting to Know the Concept
I’m gonna keep harping on this because I think it is possibly the most important part of the process for producing a complete character mesh. I begin with a very basic idea: for me I have had this panda/cat looking thing sketched in various forms for almost two years now. It’s about time I actually sat down and modeled him in one form or another.
I know a few things about this panda from past character development. His name is Raoul. He is a security guard later in life. He has an insatiable hunger for cookies. He is endlessly curious. He is incredibly cute as a kid. His ears are rather large (from later sketches). I’m sure there are some other things I’m missing, but that’s a pretty round description, especially if I’m just in it to model him.
Exploring the Character
So the above is the initial ground work to get to know the character. Now I get into the exploratory sketches to see how my final version of Raoul will come together. I have already decided this model will be as simple as possible and still animable so the detail will be sparse at best.
I begin with volume sketches in different poses to see how Raoul’s body reacts. You will notice that the majority of this sketches are just quick and dirty and there is no real commitment any form yet. Buried in amongst the sketches you may find a single front-side ortho… 
I have always thought in orthographics and rarely have I ever ventured into 3D, especially when thinking of characters. This is a bad habit on my part that I should really think about breaking in the near future.
So you say: “A set of really scratchy, completely pointless sketches.” Scratchy, yes! Pointless - oh, far from it. I have spent time with this concept now and these sketches begin to unfold the characters attitude toward life as well as answering a lot of questions about what types of major deformations can be expected.
As you can see I have opted to do a younger version of Raoul who isn’t as tall as a lot of the older sketches. I’ll get to an older Raoul later on. The next sketches are mesh planning. Yes. I do mesh planning on paper. In fact I do almost all of my planning on paper for almost everything I do that involves CG things.
A few things to note in the mesh planning stage: 1) every limb that is extruded from a mesh will be accompanied by at least two poles, 2) any extrusion in general will have four poles, 3) any change in edge loop flow will result in a three-pole with a neighboring five-pole or six-pole, depending on the situation. I use different colored high-lighters to get across the main points. In green are the major edge loops (and probably the final edge-loops will not differ from these), in pink are the expected poles bot front and back, and in orange is axillary loops that aren’t as important.
I did wind up going back to my rough sketches page for the final loops because I liked my single orthographic view so much. I apologize for the blurriness of the image, but it has enough clarity in it that I will be able to use it while modeling.
Tomorrow I move on to modeling the form of the character and ironing out the final wires for the ears and head.
Revitalizing Your Creativity - Before, During, and After Modeling
Often I have had trouble focusing on a single given mesh as I have had no real attachment except as a hobby. The result was that almost nothing I ever did got finished. In part it was because I did not plan and in part it was because I wasn’t really excited about the mesh after the first day. And that meant that I had a pile of unfinished meshes on my hard drive waiting for a little TLC. Not only did I lose interest, but I also lost my creative spark.
Before
I would find myself sitting in front of the computer practically pulling my hair out trying to decide on something to model that both interested me and was a feasible thing. In the event I would decide on something invariably it went unfinished because I lost interest
When you find yourself in this position one method is to come up with a completely clean idea, one that has never been thought of before. Simply tell yourself a lie and build upon it. Then pick an aspect of the lie and plan it out and model it. For example: chickens are actually robots that are trying to peaceably remove war from Earth. To extend on this we tell more lies: to achieve this they steal weapons and hoard them while they wait to gain the mental power needed to destroy the weapons. Suddenly we have a huge number of options for story elements, characters and locations. We have chickens that are thieving and cunning; smart enough to steal from the highest security armories but too dumb to actually do anything but hoard weapons. Locations, characters, etc. abound. Choose one and plan accordingly.
Note: Finding something you want to model may be easy, but planning to model it is the crucial part. Without a plan your approach will be unorganized and in most cases you will be unable to complete what you started.
Another method for focusing your modeling prowess is to collect images that inspire you. I have a folder with a few megabytes of images in it. Every time I happen across an image that I’m drawn to I save it. During periods of time I am searching for another thing to model I look through these images first and then re-find the authors on the internet and look at more of a given author’s work. (Finding things is made easier if you have a file for an author’s work with their name on it.)
During
Staying interested during the modeling process has always been a challenge for me. My mind moves so quickly from one thing to the next I often lose myself in a stream-of-consciousness search of the web. Granted I find some really interesting things, but I have strayed from my initial purpose which was modeling something. To keep your focus and creativity on your current model it is best to have a plan for how you will approach your model. If you have a plan you will be able to look at it, see what needs to be done (if the concept has been treated properly), and do it.
To help yourself focus your creativity on a model show it with other people. They might just tell you “neat…” or tell you they don’t care. Keep telling people until you find someone that cares (within reason). This really helps in those times when you just can’t bring yourself to model. Generally the best people to show a model to are the ones who want to talk about it when they see it and you won’t just find these people on the internet. Look for local groups or just people in your area you can sit and chat with.
Get feedback from “people who know what they are doing.” Someone who models for a job probably won’t respond to emails. I’m not saying they are cold or they don’t care-I’m saying they do it for a living and they need a break some times. The best people to get feedback from are those that model as a hobby as well. Internet forums are extremely powerful in this way.
After
This is a lot like Before from above. In fact most times it will be exactly like what is up there. You finished a project and you’re ready for a new one so now what? During a previous project (assuming you finished it) you probably had ideas. If you’re like me the ideas had nothing to do with the project you were working on (or they were simply variations on a theme).
I always have a pen or a pencil on hand everywhere I go as a kind of security blanket just in case an idea hits me. If the idea doesn’t hit me too hard I write it down and I keep a stack of ideas just lying around. First off, writing something reinforces it. Second, you now have a record of the idea that will (hopefully) outlast your memory (in case yours is as short as mine).
Tweaking the Pro Way
Tweaking is something that most of the modeling industry takes for granted. They’ve all done it and it works well for them. This method of modifying a model works for any type of model but especially well for organic models. In order to do this effectively you must know what your model should look like inside and out.
Once you have this knowledge in hand you will know what the shape of your mesh should look like basically and in depth. The other side of this is knowing how to move the view rapidly in your chosen modeling program.
Basically all you do is when you have blocked your mesh and are ready to begin editing and adding detail you start orbiting your camera around your mesh. The first thing that doesn’t look right that catches your eye you fix. Was rinse and repeat. Keep your reference handy while doing this to make sure you are being as true as possible to it.
Getting to Know the Concept
Getting to know the concept you will be modeling is essential to finish a model quickly and efficiently. Many things are involved in this but the primary piece is planning. If the concept is yours you are at liberty to make things up (but keep in mind you want to keep the integrity of the concept), but if not you must only fill in the pieces that cannot be seen in the original concept and make sure that they fit the overall style of the concept. One ill placed thing within a concept that has not been designed to fit the overall style (or thingness) the whole thing goes to pieces.
Sketching
Begin simply by sketching (I don’t care if ‘you can’t draw,’ it doesn’t matter) the shape of the concept on a piece of paper. Draw it from as many different angles as you can. If it’s pose-able draw it from different poses. All of this should be done with as little detail as possible using shapes (circles, ovoids, triangles, etc.) to achieve the desired silhouette.
Always keep an eye to the original reference image so as to keep yourself as close as possible to the original. Once you feel comfortable with the basic shape of the concept move onto sketching the detail you can see. This is important because if you do not know the detail you will not be able to model it. Work the detail areas over like you did the basic shape until you have a firm grasp of them.
Breaking it Down
Another way to handle the detail is to break it down into its component parts. Just draw the pieces without their surroundings; and even the pieces of the pieces so that you can become familiar enough to model them accurately.
Planning - It Helps Everything
Having a plan, even if you deviate from it, helps to achieve goals. The only reason to plan out what you are doing is to clarify the end result and how you plan on getting there. So, first things first, what is the end goal? It could be broad and sweeping such as “I want to take over the world!” to as simple as “I want to shower every day for a month.” In this blog it goes like this: “I want to produce a pro quality mesh so some game company will hire me!” Actually that’s two goals… We’ll go with “I want to produce a pro quality mesh!”
Defining Feasible Steps
If you’ve never modeled before there are some steps in between now and modeling the pro quality mesh. Just realize that no goal is infeasible simply because the steps to getting there are feasible. If you have the dedication to keep a list and a plan for achieving that list you can model pro quality models. But it’s gonna take time. To tell you otherwise or to omit this I would be misleading you.
Things Not Known
We begin with a list of things we do not know that are required in order to achieve the pro mesh. First and foremost we need to know a modeling tool. Any tool with a decent set of modeling tools will do. There are a few out there such as Blender (which has the advantage of being completely free), Silo, Modo, etc.
The next thing we do not know is how the pro modelers even model. Because we want to do what they are doing we need to know this. This may come in the form of looking at tutorials from all over the internet and doing so continuously for about a year. The more time put in at this stage the better the outcome will be. This is simply a function of finding the right information (which is the hard part) and then applying it in each and every single mesh you model. Within this single point on the list there are dozens of sub-points such as mesh/poly/edge flow, poles, maintaining volume, modeling for deformation, adding detail, etc., etc.
Do recognize that more items will be added to this list of things we don’t know as we run into them. And we may find that some of these items are a moot point further down the line. But with this list in hand we now know what we don’t know… As it were.
Getting from Point to Point
Next we plan out a chronological path from point to point on our list. This is our final plan taking shape. We will include things like how long we think it will take us (in hours) to get to each point and how much time we will spend daily on each. When points are taken care of we cross them off and move onto the next. Some points (the above: “getting to know a tool”) may take weeks or months to get all of the idiosyncrasies of the program and so this point may not be crossed off until much further down the road.
A lot of the points in the plan may be taken care of at the same time. We would more than likely begin modeling characters following a given process and addressing each of the modeling points as we get to them.
Finish!
There will come a point in time when you get ready to cross off that final point in your plan. When this comes along sit back and think about what it took to get to that point. It might be time for a little celebration!
KISS - Is There a Simpler Way?
We’ve all heard this: “Keep it simple, stupid.” And if you’re like me you ignored it. At least at the time. I personally tried paying attention to this little maxim when I recently began to model a character. The denser the mesh became the harder it was to manage and it went down-hill from there at an appalling rate.
Whenever you plan to execute a given task always ask before you start “is there an easier way?” Don’t be stupid with this and throw it in people’s faces if they expect you to do it a certain way. Apply it to personal projects and it will save you hours of pain.
Once you’ve asked the question you can answer it with your first reaction. The first reaction may not always be the right one, but it is a place to start.
Pros and Cons
Once you have a set of answers get a piece of paper, fold it in half length-wise, and write the pros on one half and the cons on the other. Do this for each answer to the question above. On the back of the sheet of paper you will write one or two questions concerning what you will need to know in order to take this course of action.
Odds have it that you won’t do this for every thing you ever run across, but at least weigh the pros and cons before diving in and you will save yourself a world of troubles.
Simplify
Simplifying something is not necessarily less of a thing (in my case a mesh) so much as it is a reorganization of a thing into a sensible pattern. For me all it took was for me to back up and plan out what I was doing and then execute according to plan. Things went much smoother.
Consider your options, make sure you have at least three, and then decide based upon the knowledge you possess and the pros vs. cons of each path. This process of writing things won’t always be necessary as it might, with enough use, become an internal habit that just happens in the background.
Character Modeling - Three Steps to a Better Character Quickly and Efficiently
The Concept
The concept is the keystone on which your entire model will be built. If it is only a rough concept or a single perspective image and not a character sheet make sure that some study and some care is put into the parts of the model that cannot be seen so that when it comes time to model those areas you do not get stuck trying to figure out what went where or how exactly a detail was to be modeled. Take the time to get familiar with the concept as you will be spending more than an hour with it.
In fact, if you are feeling adventurous, sketch the concept a few times in different poses to get a feel for its volume and silhouette. Also sketch close-ups of potential trouble spots to work out all the problems before you touch the computer. This will save you immense amounts of time.
A side note on volume and silhouette for those not familiar: volume is the space the character takes up. Literally it is the form that the character takes. Silhouette is the shape or contour of the character when placed in high contrast.
If you drew the concept image and feel you are ready to model it make sure that you know exactly what will happen on all sides of the character model and that all of the little details are known. If you miss even a little piece it can slow you down.
Blocking/Sculpting
In general I do not use 3D sculpting programs because the results are always bad. These are available for those that choose sculpting over blocking. The following is the way I approach blocking in the volume and proportions of a character.
I’ll begin by saying that it does not matter how you translate the concept into 3D so long as it resembles the concept closely. Specifically to be considered at this point of character modeling is the overall of the character matches the concept. Generally I use primitives and extruded circles to define the characters I work on.
When I have a character sheet I simply place it into the background of my modeling program and block in using primitives and cylinders.
When all I have is a perspective view (which is most of the time) I build a stick figure to generate the skeletal structure that the volume of the character will fit on. Once the skeleton is in place I begin to add the volume on top of it using the skeleton as the mainstay of where and the reference as to how. That is, the contours provided by the reference can be accurately traced onto the skeleton (also from the reference).
Doing this greatly simplifies the process. Instead of focusing on both lengths and proportions at the same time I only have to focus on each in its turn. I find it is helpful to break a difficult process down into its parts in order that I might simplify it so that my path to the finish is easier.
Retopology
When the above object is finished (all of the major shapes and details are accounted for) we begin to trace over it with a new mesh. A major item of interest is mesh flow and how it affects joints-especially for meshes bound for animation. Generally speaking the flow around a joint should be as smooth as possible and without poles. This is mostly so that during rendering no pinches in the mesh show up.
When plotting a mesh on a previously modeled surface keep in mind things like muscle flow, etc. so that you can hold true to “form follows function.” Which in this case is mesh flow follows form follows function.
This is not to say that the configuration of the faces at the joints is smooth, just the flow of the faces. When retopologizing the joints I begin from the torso and work outward so that the number of edge loops remain consistent.
A note on hands and feet: simpler is better. This doesn’t mean skimp on verts to make it simpler, just use as little as possible to get the shape of the hand in a clean manner. Also, hands are always tough, so be prepared to model them at least twice. It’s good for you!
Next in line is the torso. Getting a good transition from the extremities to the torso is difficult in organic modeling and it may take a few tries. Generally I hide the resulting poles on the front or back of the torso (under the arms is also popular).
When you finish this process you should have in your possession a very well modeled character with clean polygonal flow and edge-loops. And with a little practice it should happen relatively quickly!
More Sculpting (Optional)
The reason blocking isn’t mentioned here is that for the ultra detailed mesh one simply requires sculpting. Only so much detail can be acquired with a mesh. If you need even more detail then sculpting is the way to go. Though I suppose one could retopologize over an already retopologized mesh.
This process could be applied to various aspects of character modeling instead of taking the whole character at once-such as building up the body of a character first and then the clothing and then the accessories.
I will gladly answer any questions that might come up! Just ask in the comments.
Recent Comments