Where will it go? Well the intent is to explore the gray area between Cad and 3D visualization programs. Exploration presented on a detailed level for editing and clean modeling in both formats. It sounds basic in passing, but presents many winding roads when put into concepts and practice.

Segment Introduction
The Work Flow

The typical Architect & Cad format to CG Viz applications is the norm if used. This methodology usually locks both parties into separate software formats and perceptions of Freudian envy.
The exploration of 3D editing here is of AutoCAD - Architectural Desktop and Viz/MAx. The AutoDesk Suite of products may have some advantages, not the discussion here! Never the less, the workflow illustrates the use of many CAD to Viz and combination software.
We will explore in this series modeling in AutoCAD - ADT for optimal results in Viz - Max. Understanding the how the AEC, Multi-View blocks, blocks, Solids and 3D Face components can be used and manipulated before entering the 3D Viz worlds. Whether you work with a CAD, CG Viz or both types of applications it may be advantages to understand what may be going on in each or other types of applications. Allowing for greater flexibility of accessing components, then being able to quickly, definitively clean and properly order complex CG Viz scenes.

This segment will look at some basics related to ADT/CAD and 3D programs. Then move to light application in calculated lighting solutions, Global Illumination (GI). The use of the term GI is used very broadly here as a catchall term. A narrower view for GI is for the Photon Based lighting engines-Vray, Mental Ray, Brazil vs. the Radiosity based of Max and Lightscape. Anyhow, the GI based and even fakiosity lighting has the greatest affect on "clean Modeling". Understanding how these GI engines work is the key to proper and efficient modeling.

Starting with making a model and developing it in ADT. The dwg file drawing is of this real world space. In the future, we will be adding some proposed cabinetry and a new 1/4 round window. (Change order -- For illustrative purposes only!)

We will start by looking at the AEC walls. The ADT advantage is flexible & multi-fold. Generation of Construction line type drawings and 3D modeling combined in AEC objects. The CG Viz disadvantage is generation of redundant coincidental faces.

AEC WALLS
COMPONENTS

ADT uses definable boundaries to create the lines and 3D objects of walls / AEC objects. Walls have many adjustable parameters that control the boundary definitions, in the style of the wall. Most important to our discussion here is the entity display dialog; function relating to materials and elements in Max.
A Viz/Max compatible Multi-sub material applied by default to all walls. Viz has a default material/s with individual maps and materials that correspond by color to the boundaries and or components, of most AEC entities. The default multi-sub materials have sub materials ranging of order to 154+ materials. These sub materials can be applied or altered; however, it is usually more practical to create and assign one with just the number needed.
To do this you need to assign a new color. The colors in AutoCAD are ordered by numbers; red=1, yellow=2, green=3..... to a defined black at 255. Each number represents a place or slot in the multi-sub material: therefore red is the first multi-sub material, yellow second..... These numbers also directly correspond to element selection and function of editing in the max platform.

In this diagram, a default or no end cap style is applied. If not needed for line drawings, it is just as easy to edit the vertices of the wall components than having to eliminate narrow faces of a returned drywall for example.
The AEC generation of 3D objects creates faces based on corner vertices. The diagram shows with orange and violet shading the co-existing faces that typically generate with normals facing each other in max; "co-incidental faces". Viz and Max have overcome some of the problems with processing these by ignoring them. However, it takes processing time to do this. In Lightscape, Mental Ray and others light energy is still processed, on these surfaces. A large scene can spend hours, needlessly processing useless information. Unneeded faces and geometry should be eliminated. Doing this in the CG Viz application is much easier. By using the entity display properties in ADT, you can make selection of these surfaces simple. The multi-sub material function is very useful for exporting into Lightscape from max. Both are directly related.

ADT ENTITY DISPLAY
Adding Usable CG Viz Controls

There are many wall styles available in ADT. Using StudX .625 GWB (both sides) style is a good for both CAD aspects and 3D walls with different materials on each side. To make good use of this wall style we will change it's entity display characteristics.

The co-existent (co-incidental) faces also are a real problem with all Global Illumination (GI) rendering. Use the glass edge image for an example. Lightscape, Radiosity, Mental Ray (Photo Based) GI all will equally light the glazing frame on both sides of the glass. To overcome this a dense; mesh sampling level, or photon number works to better the lighting and shadow, however at the cost of processing time.

CALCULATED ILLUMINATION
Radiosity - GI Explanation& Modeling

Lightscape is a good place to see a lot of what goes on with Radiosity or mesh based GI. All objects, blocks in Lightscape, become faces that combine to create a surface mesh of an object. Polygons, 4 sided surface or sum of 2 triangulated surfaces, are ideal. Keeping in mind that a square is the optimal shape and narrow long rectangles are problematic.
The images below show the parameter driven receiving meshing that Lightscape uses to apply illumination and shadow (the lack of illumination). It is polygonal in nature. The objects surface mesh will be divided by the receiving mesh, if larger than the parameters set (meshing size). The ceiling is one large polygon, divided into smaller square receiving polygons, based on the meshing parameters. The window mull, right side, shows the interaction between actual mesh size (width) and the receiving mesh divisions (height). By the x,y dimension of any given surface plane.
Looking at the walls receiving mesh there is a rectangular division. Further refinement based on a triangular refinement, being complementary to the face of the wall surface as define by the lower window vertex. Continuing refinement allows for a more accurate illumination gradient, where needed. You can see a basic gradient example by the violet highlighted polygon mesh segment. This is Shadow leak, caused by the lack of illumination of the wall, transferring out in the polygon. Decreasing the meshing size will minimize the artifact and in the process dramatically increasing the time to calculate light distribution for this surface.
This is a very simplified explanation of a solid and flexible Lightscape program.

LIGHTSCAPE EXAMPLE

Back to the glass edges, using the basic Lightscape meshing for radiosity we will look at the window frame & glass. The example on the left demonstrates the meshing and illumination gradient. The glazing material was change to opaque. The radiosity solution calculated, after hiding the glazing panel for rendering. Notice the fairly even gradient from illumination to darkness, not a realistic or accurate illumination. Changing the radiosity, meshing can help. An optimal method is model the frame splitting this surface plane into separate meshes. The vertical edges, vertex to vertex, match the edges of the glazing solid. The example on the right shows the results. Even not being modeled as described, edge to edge, some overlapping occurred. The Lightscape parameters and calculation time were, identical for both. The gradient well lit on the illuminated side and dark where it was not. The surface mesh does have a positive effect on radiosity illumination, of course when done correctly.

MAX RADIOSITY EXAMPLE

The Max Platform uses a mesh similar to Lightscape. However, it is a triangle based mesh, similar to faces-tri/s as opposed to polygons-quads of Lightscape. The triangle meshing can form squares & rectangles. More often, it creates a mesh that is more adaptive to irregular surfaces and not so perfect modeling. Overall more straight forward and simpler to control.
The example below on the right shows a 1'0" meshing size and mesh. The Model is identical to the Lightscape example. The typical, rendering of the glazing object. Notice the difference of the illumination gradients. The mesh is fixed using a global size and does not refine it's shape to the shadows as Lightscape does, even more reason to use clean modeling techniques. You can very easily override the global mesh size on the object level, allowing adjustments for improving local radiosity solutions.





Looking at the window frame notice the gradient in the left image, above. Similar to the Lightscape example, but it has a curve to the distribution. This is a tri vs. quad meshing difference. Still is not correct. By splitting the face of the window frame surface, we get accurate results, right image.
Clean modeling can overcome the shadow leaks and artifacts. You can see these in the highlighted and marked areas. A continuous roof slab causes an extended shadow into the room, based on the mesh size and layout. Looking closer, the shadows above the window are triangular and directly correspond to the subdividing mesh. A smaller meshing size will help. Regathering will help a great deal. Both will add to the radisoity calculation time. Clean modeling of these surfaces can keep the calculation times close to, one directly out of ADT, with better results.

Photon Based GI EXAMPLE

The mental Ray rendering engine plug-in for Max6 illustrative of the photon based GI. Again, the same model and set up.
The way a photon based lighting calculation is quite different from the mesh based. It is usually mesh independent, an equal opportunity illuminator. Surface properties, materials and rendering engine specific shaders, have the greatest control. With this comes speed and loss of physical correctness. Although very close to being, correct. Samples, sampling size, photon size, the numbers of photons, and energy will render higher quality results.
Clean modeling may not be quite as critical, but why store lighting information buried behind visible faces.





Photon based GI applies indirect light by intersecting a photon, with a face, leaving light (energy) and bouncing light energy back into the scene. You can see the disco ball effect in the left image above. The photon parameters are quite poor, but illustrate how the energy is applied. Look at the circled areas; the photon energy equally applied across the glazing frame at the opaque window panel (hidden for rendering). The arrows show a material property having an effect on the light energy. The high-energy sun/sky photons outside of the window reflected back out and not applied uniformly to the frame. Being caused by a high angle of incidence and bouncing off the reflective glass surface.
Usually with a photon, based GI is a final or re-gathering of light energy. As with Mental Ray, final gather by itself or in conjunction with the photon mapping (distribution) of light. The right image has fairly low final gathering sample range. You can see, the additional illumination applied by the final gather process, areas affected by the photon mapping, and sampling effects (smoothing) of the final gather. The arrows point out the material shader effects at the glazing frames and the parentheses show these effects on the exterior soffit.

POLYGON MODEL
Lighting & Image Map Based Thinking

The mesh-based radiosity uses a scalable mesh for light calculations. Why not model with a polygonal mesh to conform to lighting characteristics. Better yet, control the materials. You then have, at the time of modeling, verified polygon dimensions and aspects for editing a bitmap image to fit.



As shown above, an image (bump/displacement MR) mapped to individual polygons. The actual polygons are visible in the wireframe. An obvious problem is the rectangular polys above and below the opening. An advantage is the ability to verify the dimensions and ratio aspect of an edited image to fit in ADT.
The wireframe displacement image shows how a MR displacement actually creates a rendering mesh for the "physical" polygon mesh. This kind of displacement being more efficient than having loads of vertices "physically" displaced in max and of a higher quality.

WRAPPING IT UP
Why Lighting Examples?

Understanding how software applies lighting is the key to successful modeling. Using your understanding of the mesh parameters to build a scene, allows for quick accurate and unadjusted results from your render.
Further knowledge of how to take advantage of cross platform information and workflow is critical. Having verified aspect ratios, from ADT, for an image map will make editing in Photoshop a one-stop effort. Building in, entity display functionality makes selection; mesh editing and material application seamless when brought into the Max platform.
These are some of the fundamental basics for working in a single platform or cross platform modeling. This is the foundation for further tutorials. Not very hands but that will change............