3D Modeling

Creating a 3D model is at the heart of the imaging process. It’s similar to contructing a real building, but it also borrows techniques from film and stage production.

3D modeling is the process of using specialized software to create digital objects that have three dimensions, and can be manipulated, viewed, animated and moved around like real 3D objects. The big advantage of using a computer to render 3D models is that you can build a model once, and render from it an unlimited number of times. You can also edit, revise, add to, and subtract from the model and re-render. The editable nature of 3D stands in marked contrast to traditional hand renderings, which are always one-off creations with either no, or very limited ability for revision.

Since the focus of this guide is architectural visualization, we’ll restrict the discussion to the types of modeling used when building architectural projects.

One quick note: In order to make this part of the guide less abstract, we’ll use a sample building to illustrate the process, the final 3D rendering of which is shown in Figure 7.1. We’ll also restrict the discussion to the 3D program that our shop uses, 3D Studio Max 2010, hereafter referred to as 3DS Max.

[insert Figure 7.1]

The first order of business is to import the data. For the sample project, this took the form of 2D CAD files drafted in AutoCAD 2010. It is worthwhile to open up the CAD files in their native software and do a little prep work. The principle things we do are: 1) make sure everything is on the same flat plane (sometimes objects are drafted at different height, like an “exploded” parts view), and 2) any layers that we don’t need to import (for example any symbols or text) are turned off.

After we’ve imported the CAD files into 3DS Max, they look like Figure 7.2:

CAD files imported into 3D Studio Max

Figure 7.2: CAD files imported into 3DS Max.

Notice that all the CAD pieces came into Max on one flat plane, just like a drawing sheet. We’ll now move each of the CAD pieces into their proper 3D orientation. We’ll put the floor plan and site plan on the ground plane; we’ll rotate elevations into the 3D location they represent, so the front elevation goes to the front plane, left to left, back to back, etc. Finally we move the roof plan to a location at the top of the walls so it is floating about 9′ off the ground plane. When we’re finished, the file looks like Figure 7.3.

CAD elevations moved and rotated to be in their proper orientation

Figure 7.3: CAD elevations moved and rotated to be in their proper orientation.

We’re now ready to start building the 3D model. The method our modelers use is pretty simple: we trace over the elevations, creating multiple 2D objects, to which are added thicknesses, and presto: a 3D building. You might wonder why we rebuild what’s already there in CAD data. The reason is our 2D objects must be built in a certain way, obeying a set of rules. The CAD files are built under an incompatible set of rules, and if we try to add thickness to these CAD objects, nothing happens, and so hence, we rebuild.

We don’t have to rebuild everything, however. Figure 7.4 shows the state of the 3D model after we’ve created all the objects that are custom to this job. This includes the walls, gutters, roof, porch, driveway, sidewalk, and curbs. Everything else – the windows, doors, railings, trim pieces – are all based upon stock items that we can grab from our 3D model library.

[insert Figure 7.4]

Our 3D library contains just over 77,000 models. We have pre-built windows, doors, door hardware, columns, porch lights, stairs and stair rails, brackets, balconies – almost any building part you can imagine. We also have pre-built plants – small trees, large trees, shrubs, flowers, groundcover, vines and grass. And when it’s time to populate a scene with a little street life, we have hundreds of 3D people, cars, park benches, street lights, mailboxes and bus stops from which to choose. The point of all of this is to emphasize the modeling approach: we divide the work into two groups: that which must be custom built for the project, and that which can be created from pre-built objects.

So back to our house model – we’ll start by bringing in windows from our library. The windows must be modified to fit the opening, but this edit is quick and much easier than building windows from scratch. We go through the same process for all the remaining parts that can be made from modified library objects. When the model is complete, we add a set of placeholder colors corresponding to the broad categories of materials present in the building, for example light gray for the walls, gray for the roof, dark gray for the windows. We add a light to represent the sun at about 11:00 am on a cloudless day- this is flattering light, not too harsh, and the angle is great to ‘read’ the various 3D surfaces. We render the scene; the result is shown in Figure 7.5.

[insert Figure 7.5]

The neutral colors in this rendering allow us to review the 3d model for errors without being distracted by the actual materials. After we’re satisfied the model is correct, we send a rendering to the client for their review and comments. When the clients sign off on a complete, correct model, it’s time to progress to the next phase: Texturing.

Menu