Introduction: Here’s something that was news to me when I first got started. There is more than one way to model in Lightwave. I’m not talking about the difference between your style and mine. I mean, within Lightwave Modeler, there are no less than four different approaches (each of which has associated tools and techniques) that you can use when beginning a modeling task. These four approaches are: Polygonal modeling Spline Patch modeling Metaball modeling Subpatch modeling I always knew that you could create splines in Lightwave (although how to model with them was a mystery to me). I knew Lightwave had a subdivision mode. And that you could create metaballs. It simply never crossed my mind that these weren’t simply tools, but rather distinct approaches to the entire modeling process. For those to whom this is a new concept, let me first say you’re not alone. It may not make sense now, but hang in there. It can be a hard thing to wrap your head around. Here’s an analogy that might make more sense. If you wanted to paint your house, you could approach the job in one of three ways. You could use paintbrushes, you could use a paint roller, or you could use a motorized pressure sprayer. Each approach uses different tools and requires different techniques and skills. In all likelihood you probably wouldn’t paint the whole house with just one approach. You’d use the right approach for the right job. You wouldn’t bust out the sprayer to paint the area around a light switch. And you wouldn’t use the brush to tackle the ceiling. Rather, the light switch is best handled with the brush, and the ceiling calls for the roller. It’s the same with the different modeling approaches. In any project you will likely use a combination of the different approaches to meet your specific needs. How will you know which is the best approach for any given situation? Practice. Seriously. I wish I could give you a simple guideline, but honestly that’s something you’ll have to learn as you go. Don’t dispair, though. It’s not entirely without logic. For example, if you wanted to make a box, you could laboriously do so out of spline patches. But you probably wouldn’t, as the same box could be made with a single click using the polygonal approach. So there is some amount of logic that goes into it. But the more you practice with the various approaches, the more you’ll get a feel for when to use one over another. In the end, it mainly comes down to a matter of choice. I might use polygonal modeling for something and the guy next to me could easily argue that the subpatch approach would be faster/better/etc. He could be right. But if I’m getting the job done and I’m happy with the results, that’s all that matters. Okay. I’m going to briefly cover the various modeling approaches, but I want to spend the majority of my time on subpatch modeling. This is the method you’ll likely use for modeling characters and other organic shapes. But before we get into that, let’s look at Polygonal Modeling, Spline Patching, and Metaballs. Polygonal Modeling. Polygonal modeling is perhaps the oldest approach and, interestingly enough, is the easiest to get started with. Under the polygonal approach, all that matters is the end result. You don’t have to worry about polygons with more than 4 points (as you do in Subpatch modeling). You don’t have to hassle with spline patches. You just get the geometry out there and go for it. Cut, slice, stretch, boolean the heck out of it. It doesn’t matter. As long as it looks good, you’re gold. Take a look at the image below. Holy smokes! Look at that mesh! It’s a mess! But look at the preview window. Doesn’t look too bad, huh?. If you’re new to modeling, you’ll likely end up with objects that look messy in wireframe mode. That’s because common operations such as Triple, Julienne, and any Boolean function will create a ton of extra geometry. As you gain more experience, you’ll learn to model more efficiently which will make editing your models easier and also speed up your renders. But for now, if your mesh is a mess, don’t get discouraged. That’s normal. Spline Patching. Spline Patch modeling is an approach that begins by creating 3D outlines of your object. You typically draw your splines over background reference images (which you can get here: http://www.suurland.com/ for vehicles and here: http://www.fineart.sk/ for anatomy). Your splines are usually drawn so that they share four edges, which allows a roughly “square” area to be identified. You then “fill” these areas with patches of polygons. It’s a great way to make complex shapes and is an excellent way to get the basic form of an object you wish to convert to a subpatches. If you look closely at the image above, you’ll notice that every spline acts as an edge. Taken together, these edges define a series of square quadrants (in that they have a top, bottom, left, and right side). The notable exceptions are the nose and tail, which define triangular areas. Spline Patching is great because you can see your object materialize in a 3D wireframe form right before your eyes. I have to admit, that’s pretty satisfying. But it can be a laborious process, and patching those splines can be problematic if you haven’t set everything up properly. For a great tutorial using Spline Patches, check out Gerald Abraham’s “LightWave 8 Vehicle Modeling” video available at Kurv Studios. Metaballs. Metaball modeling. Okay, I’ll be real honest. I’ve never done a complete model using Metaballs. And I’ll bet that you won’t either once you’ve tried out the other approaches. That doesn’t mean that Metaballs are useless. It’s just that they’re unwieldy and it’s difficult to get a really detailed model with them. Still, they’re a lot of fun to play with. Exactly what is a metaball? It’s hard to explain. But perhaps the best way to understand them is to think of the individual blobs in a lava lamp. Each has its own mass, but under the right circumstances, it can meld with the other blobs to create interesting organic shapes. Each metaball has both size, and influence. These settings determine how the metaballs interacts with each other. Take a look at the image below. This little “doughboy” character was built in about 2 minutes. Doing something like this with polygons, splines, or even subpatches would take considerably longer. So don’t rule out metaballs. Just don’t try to model a car engine with them. Polygonal Modeling, Spline Patching, and Metaball Modeling are all viable approaches to the modeling process I would encourage you to play with each and get a feel for them. When you start to realize the advantages and disadvantages to each, you’ll be able to make solid decisions that will result in more productive modeling. There’s still one more approach we haven’t talked about yet. Subpatch Modeling (AKA Subdivision Surface Modeling). When you press the Tab key, Modeler turns any polygons with three or four points into subpatch objects. Subpatch objects can be used to create organic objects quickly and easily, but they are also excellent for inorganic models as well. If you like, you can follow along with me. Create a plain box (Fig.1) and hit the Tab key. Modeler rounds all the edges, making something that looks like an odd shaped ball (Fig.2). Figure 1 Figure 2 When you turn on Subpatch mode (tab key), Modeler looks at all the 3 and 4-point polygons and attempts to smooth out any hard angles. It’s similar to the idea used in the Smoothing function in the Surfaces Editor. However, instead of simply giving the appearance of smoothing, Subpatch mode is actually adding geometry. The level of geometry being added is determined by the Patch Division setting in the General Options (o key) panel. (Fig.3) In Figure 2, above, the Patch Division was set to 2. As you can see, the hard edges of the box have definitely been rounded, but it still has sharp edges which can be seen along the outside profile of the object.  It looks as if we created a sphere with too few polys. We can remedy this by increasing the Patch Division level to something like 6 or 8. In Figure 4 below, the Patch Division was set to 10. Notice that, while the edges of the box are still visible, the corners have been smoothed out tremendously. I haven’t done anything to my model. I’ve just adjusted my Patch Division. Figure 3 Figure 4 Patch Division is something that you can adjust at any time. Depending on the speed of your computer’s video card, you may want to dial this way up or way down as you work What’s important to keep in mind is that subdivision models don’t need a lot of extra geometry to look smooth. This is an important thing to remember. It will be tempting as you model to continue adding detail with more geometry. But before you do, you may want to crank up your Patch Division setting and see how things look from there. Often, a little bit of geometry coupled with a higher Patch Division will go a long, long way. Just remember that higher Patch Divisions require more power from your video card and will also take longer to render. If you find that Modeler is not as responsive as you would like, dial your Patch Division down. But be sure to bring it back up now and again to ensure you’re not bogging your objects down with unnecessary geometry. You control the shape of subpatch objects by manipulating the proximity of geometry. In other words, the closer together your polys are, the area to smooth will be smaller and you’ll end up with a sharper edge. If your polys are further apart, you’ll have a larger area to smooth and you’ll get a softer edge. Let’s try this with the box. If you’re still viewing the object in subpatch mode, click the tab key so that you see the regular box again (as it was in Figure 1). I’ve set my Patch Division level to 6, and my Surface Editor’s smoothing is off. Now, press Shift-K to activate the Knife tool. We’re going to make two vertical slices in the object approximately 1/3 of the way in from either side. Place your mouse in the Top viewport above the box. Click and drag down to create the first slice. If you hold down the Control key before you click, you can constrain the cut so that it is perfectly straight. If you’ve got a mouse with a middle button or a middle wheel (which will usually double as a middle button), you can simply click it and drag versus holding the Control key down. This will constrain your cut so that it’s perfectly straight. Just a side note, but using the middle mouse button (or the Control key) will also constrain the Move tool, the Extrude tool, the Drag tool, and just about every other modify tool. After you’ve made the first cut, hit the Enter key to set the cut, then press Shift-K again to activate the Knife tool, and make the second cut. You should end up with something similar to Figure 5. Figure 5. Now that you’ve made a few cuts in your box, let’s hit the Tab key again to enter Subpatch mode. You should see something like Figure 6 below: Figure 6. The last time you hit the tab key, your box turned into a sphere (Figure 2 and 4). Now you have something like a cylinder. Let’s take a look at what’s happening: Figure 7. Take a look at the middle poly in Figure 7 above. It looks so friendly and unassuming just sitting there, right? You bet. But let me tell you, behind the scenes, this poly (and actually the entire band of middle polys you created by cutting the box) is doing a lot of work. Let’s examine this in more detail. One of the first things to note is that the cuts you made did not create any new angles. Even though you added geometry, you did not change the object’s shape. That means that the subpatch mode won’t have to smooth across any new edges since these polys are perfectly level. This doesn’t mean that smoothing isn’t happening, however. On the contrary, since there isn’t any change, this band of polys acts to constrain the smoothing through the entire middle of the object. Take a look at Figure 8. Figure 8. Notice that the middle band of polys is relatively unchanged between Figure 7 and Figure 8. Smoothing is obviously happening. There’s no denying it. But the middle band of polys has the least amount of change. The real areas of change are on the left and right sides of the middle polys, and on the edges where the top of the box meets the front and back sides. See Figure 9. Figure 9. Now, remember what I said earlier about smoothing happening based on the proximity of your geometry? By placing cuts about a third of the way in on either side, you’ve constrained the smoothing that happens over the middle of the object. When there were no cuts, smoothing took place over the entire surface, which is why you ended up with a ball-like shape before. But with the cuts now constraining the smoothing across the middle of the object, you’re getting something like a cyllinder, as the middle section stays somewhat flat in comparison to the ends. The other thing to take a close look at is how much smoothing is happening on each of the sides. Check out Figure 9 again. By creating cuts, the polys to the left and right of the middle ones (indicated by arrows) now have less area to smooth as they hit the edge and go over the sides. This smaller area to smooth results in sharper angles. They’re still well rounded, but compare them to the front edge that’s circled in Figure 9. The front edge is almost completely rounded in the subpatch mode. This is because there is no additional geometry to constrain the smoothing process along this side of the object. Sure, there is additional geometry from the cut you made, but it runs in a parallel band of polys, creating three vertical streams. Smoothing will only be inhibited by geometry that runs perpendicular to it. I know that might sound confusing, so let’s use an analogy. You can think of the smoothing process like the flow of water over a waterfall. In Figure 9, the three bands of polygons running over the top and down the front can be seen as the waterfall.  If you wanted to limit the flow of water (the rounding effect) before it went over the falls (the polygon’s edge), you would have to put the dam up so that it ran against the flow of water. It would have to be perpendicular to the flow of the water. Look at Figure 9. There is no dam to limit the flow of water over the front edge. So the maximum amount of smoothing will occur. Conversely, if you look at the flow of water (smoothing) from the top to either side, there is a blocker there, at about 1/3 the way in from each side. These are the cuts you made. They limit the amount of rounding that happens from the top as it goes over the left and right sides (indicated by the arrows). So how do you affect how much smoothing is going to occur? You do this by adjusting the proximity of the geometry. You do this by placing a “dam” closer or further away from the edge. Let’s do this to see it first hand. Take a look at Figure 10. Figure 10. I’ve gone back to the regular polygon mode by hitting the Tab key. I then switched to Points mode (Control-G) and in the Top viewport, I used my right mouse button to draw out a lasso around the middle points. Now I want you to press the h key (keep in mind it’s case sensitive, so don’t press the H key). This turns on the Stretch modify tool. I don’t like to call it stretch, since it’s really more of a resize in only two dimensions, but you call it what you like. Anyway, make sure that Mode:Action Center Selection is set by pressing Shift-F8. Place your mouse in the Top view and hold down the Control key (or the middle mouse button). Click and drag to the right so that your points move away from the center and towards the edge. You should end up with somethign similar to Figure 11. Figure 11. Drop the Stretch tool and deselect the points. Then hit the Tab key to go back into Subpatch mode. You should see something that looks like Figure 12. Figure 12. Notice that the edges are now even sharper than they were before. By reducing the area that Modleler has to smooth, you’ve limited the area of the rounding, resulting in an even sharper edge. Figure 13 shows the subpatch object before and after moving the middle poly’s points. Figure 13. Here you can clearly see the effect that geometry position has on smoothing. By carefully placing your geometry, you can easily adjust the shape of your object. This allows you to get wonderfully smooth edges with very little geometry. The key to effective subpatch modeling lies in the fundamental understanding of these principles.  Rounding occurs from edge to edge, and rounding is inhibited by geometry which runs perpendicular to the flow of polygons.  Of course, understanding this and practically applying it are two separate things.  But I hope this tutorial has provided you with a basic understanding of the concepts. From here, I would encourage you to experiment and get a feel for working with subpatches on your own. If you’re interested in a more in-depth study on these topics, I would recommend my forthcoming book on Modeling, available soon from Wordware Publishing.