Symmetry

Think of how many objects around you have symmetry. Chairs, couches, lights, cars, glasses, and pretty much every animal that every walked the earth. When so many things around us are symmetrical on at least one plane, then why not have a way of literally cutting your work in half by modeling only half of the model and copying it?

Symmetry objects help by automating the process. They help because we don't have to constantly be copying our current progress, flipping it, and repeating this step every time we want to see how the full model looks.

Whenever possible, use a symmetry modifier of some type to make models quicker and more exact.

Remember that even if some objects are not exactly symmetrical, an elephant with one broken tusk for example, .·´¯`(>▂<)´¯`·. you may still want to use a symmetry tool for the face and just attach the tusks as seperate objects.

Here is a simple sphere where a brush tool has been used to stretch out one side. The effect has been mirrored on the other side of the center axis (It's important to note the symmetry is across the modifier axis, not the world axis).

Instances

Similar to symmetry objects every 3d package will have some way to make instances. An instance is basically a copy of an object that automatically updates itself when the original is editetd with only their position in the world being the difference between the two. The word "copy" usually refers to a totally new object that does not. Make sure you're paying attention to whether or not a "copy" or an "instance" has been made when duplicating objects for this very reason.

Instancing objects is not only a great time saver when it comes to modeling similar objects but can produce smaller files and (sometimes) faster render speeds at the same time.

In the model below, instances were used to duplicate each post after just one was made, greatly reducing the amount of time it took to create this model.

Notice how not only is the model being duplicated but also the properties of the model such as the texture are perfectly duplicated (In some cases this may be undesireable and is discussed further in the "texturing tips" section).

Deformers

Deformers (sometimes "parametric" deformers) are used to distort the shape of the polygons in a model in an exact, mathematical way. This is often easier than pushing and pulling each and every point yourself. One drawback to using a deformer hower if precisely that you can not push and pull every point.

Imagine a twisted wrought iron fence for example. Effectively each bar on the fence is nothing more than an elongated rectangle twisted and bent in on itself. This can be easily accomplished with the help of a few deformers used together.

Various deformers and their effects on a subdivided cube. From upper left to lower right; "spherical", "shear", "taper", "twist", "ripple (forumla), and "explosion". These will all differ between programs.

The effects of mesh density

One principle to always keep in mind when using deformers is to make sure that you have enough polygons in your deformed object to actually see the deformation. In the example below you can see the effect of not having enough divisions in the appropriate direction for a bend deformer (vertical cuts would offer no benefit).

Examine the image below.

Above we see a simple "bend" deformer. Pictured are three cubes with the deformer applied, and with the same settings on each deformer (a "strength" of 55, giving a bend of about 55 degrees). The cube on the left currently has the modifier itself turned off so that you can see the original object. The middle cube has the deformer activated, but has too few edge loops to follow the deformation accurately. The cube on the right has had divisions added so that it now follows the deformer properly.

There are quite a few deformers in every program. So many that it is actually best for a new user to just try them out themselves. Most of them will be very obvious. A "bend" deformer will bend an object, or part of it, while a "twist" deformer will twist the object and so on. Some will be slightly more obtuse, like those that are dependent on object interaction. The "formula" deformer typically allows the user to type in mathematical formulas to achieve various effects while Cinema 4D's "Collision" deformer for instance requires you to link it to another object to simulate dents caused by that object.

One deformer that may not be as immediately apparent in how to use but exists in every program is the FFD, or "Free Form Deformer".

For instance the turtle shell below offers the following conundrum. WIP section If we have a simple torus we But if we manually move some of the points to create our curved underside then

Axis position

The last thing to remember when using deformers is that in almost every 3D program the axis / origin position of the deformer is important, maybe moreso than any other kind of object besides animation joints.

Below is an image of three vertical squares. They are all under the effect of a single deformer. Examine how the distance of each square from the center of the deformer changes the way the deformer affects it.

See how each square follows the curve of the deformer no matter how far away they are? The "bend" deformer specifically will take into account the position of the deformer axis, the highest limit of the deformer, and the lowest limit of the deformer (here high and low limits are the top and bottom line of the deformer box). From there is stretches any object affected to match that curvature. In the case of the bend deformer you can use the axis position to change the sixe of the object. If the object height exceeded the high and low limits of the deformer than this could be used to decide over what length the entire bend takes place.

Remember though, every deformer will act differently, some experimentation may be required to obtain the desired effect.

Subdivision surfaces

To subdivide a mesh is to divide certain parts of it, typically splitting a single quad into four, to either add usable vertices or to smooth the surface.

In the image above we see a normal cube on the left. The second object is the same cube after your typical subdivision operation. Note that for each face we now have four faces due to dividing the face along both the width and the length. Here the resulting verts have had their positions averaged creating a smooth shape. Subdividing it further results in the third shape having sixteen faces. At 3 stages of subdivision the last shape has created sixty-four faces from what used to be a single face (and therefore has 384 polygons total).

You must always be careful not to make use of this kind of smoothing when it is not needed. One mistake I've seen new students make countless times is to apply a subdivision effect too early in the modeling process in the hope that it will make the model more complex. Let's establish one rule to always consider if you are going to apply a subdivision modifier. In fact let me put it in quotation marks to make it sound like it's an important quote by a famous individual.

"Increasing the number of polygons in a model, without justifiable reason, does not necessarily increase its complexity."

-Some Smart Guy

In effect, a subdivision modifier should *only* be applied if you are going to utilize one of two of the following techniques.

When to use subdivision surfaces

The first instance is if you intend to coninue modeling the newly created geometry as an editable object. This can be hazardous however since a subdivision function typically subdivides ALL polygons on a mesh object. If you are in need of more geometry for a particular area during normal modeling it can be beneficial to use either a cutting tool or a loop-creation tool to make sure you are not adding more polygons than you actually need.

A valid instance would be if you had modeled, say, clothing and KNEW that you wanted to add folds and creases across the entire model once the basic shape is constructed. In this case you know that surfaces changes will have to be made across the entire model and subdivision might be called for.

The same might be said for a car body that might make use of various divots, dips, and angles in the surface.

This is further discussed in the "wasted polygons" area of the "Modeling Tips (Structure)" section.

Edge loops for form

There's a common problem that occurs when using smoothing subdivisions on objects that have hard edges. The edges you want to retain as "sharp" are dulled and sometimes lost completely. Let's look at an example of a simple box again, but this time we'll make a few extra cuts along the edges, make note of them in the initial shape on the left.

Always keep the tools outlined on this page in the back of your mind. When possible you should be using things like 3d paths and deformers to make an object since they will allow you to edit the object with fewer actions than a plain mesh object will.