Oukay, I’m home now and waiting for my overheated microwave dinner to cool down ...

Just a tad of background, I'm a civil engineer and a land surveyor by schooling and career. Over last 10 years I've been punching and nagging and kicking and screaming and yelling about 3D, and every now and then I would get to do some of it. Right now, I work at a civil engineering and geology firm doing 3d modeling of terrains and various engineering designs. I managed to talk the company into creating a data visualization department, and they made me a 'director'....

I've also done a fair amount of GIS (Geographic Information systems) and GPS (global positioning) work.

I'd like to show some of the newer things, but I'm limited by copyright and proprietary data issues on current projects... Plus a lot of it is raw modeling, and not that many 'petty pictures' yet.

Back to the posted project….

Initial data ws given to me in various formats. Engineering drawings come in AutoCAD, in 2D, with a lot of elevations annotated in a form of text, and some by symbols, and the rest I filled in on my own, using my grading design experience. Lot of keying in of numbers. Some information with respect to environmental constraints came in different GIS formats, and was imported in CAD. Existing ground is a mix of higher precision aerial survey (giving me 1-foot contours and about a million and a half of triangles to handle) and less detailed stuff beyond that came from USGS DEM’s (digital elevation models)

The idea was to make a precise model. When I say precise, I'm talking everything is in calculated place, to within couple of tenths of a foot. No eyeballing and approximations.

This enabled us to do some of the engineering analysis on the project. For example, check street gradients and utility locations, estimate volume of soil to be moved and shuffled around. Determine precisely which trees can be saved which will have to be removed… On this project the client had to pay a $10,000 environmental impact mitigation fee per each mature oak tree to be removed. There were 3000 (three thousand) trees on the site that were inventoried!

Most of the initial calculations and input of 3D line work was done in AutoCAD with help of some of AutoCAD add-ons and few minor scripts and routines I wrote. Then I created TIN type meshes representing existing ground and a proposed grading. On the meshes I delineated locations of existing trees, tree rows t be planted, some utility locations, constructed curbs, gutters, building pads…

After Autocad I took he meshes into 3D max and 3D Viz. This is where I made few rudimentary house models (the client didn’t want to spend the time and money building more detailed house models, which was a bummer. They wanted ‘boxes’. I couldn’t talk them into more detail. In Max I also planed the new trees. It was important to mimic the treeline sizes and coloring f the actual trees to be planted, so the planning commission gets a ‘feel’ for the colorfulness of the
planting on site, but on a site of this size (700 acres) and over a 1000 lots, it was cost prohibitive to do realistic trees.

Mind you this was in the days Photoshop just came out with version 5, AutoCAD was a version 14, Max was a 3.1, Bryce was still 4. I had a computer at the office that had a dual processor, a ‘screaming’ 1 gigahertz each, and 1 gig of ram. If I remember right, the year was 2001, there abouts.

So, in max I planted the houses and the trees, then took the model into Bryce for texturing and rendering.  I liked Bryce scene and material setups better then max.

The original image was 6000x4000 pixels, or larger. I had to render things in pieces. It was too much for the computer to render trees and the houses and the terrain etc all at once, so I ended up rendering trees separate from the terrain, separate from the houses/roofs etc, rendering each layer on a bluescreen background. Then I assembled all the layers in Photoshop and did post work. Adding some drop shadows, doing some smudging and brush-stroking in the areas where we wanted things softened up, adjusting coloring and the usual things.

After the first review the client and the landscape architect decided that it looks little too sterile without cars in the streets, so I found some simplistic cars and rendered 4-5 basic models in 5-6 colors each – again on a blue screen at relatively high resolution – bigger then the flea size you see on the image. Then I made use of Painter (I think it was Painter 4 – before Corel bought it), and used the are images to build a few image nozzles (hoses) to draw them around the site. When drawing the cars, I had to pay close attention to the right mix of colors, if you look out in the street, after a while you’ll notice most cars are white, then there is a specific mix of which colors are more predominant from others. Also, you want to be careful about the density of cars. You don’t want to communicate that your project will create a traffic jam, and you don’t want to have it look deserted.

I opted against placing cars in 3D and rendering against another blue screen, but at the moment I don’t remember the reason why.

In addition to this image, called the illustrative plan, I had to do a few dozen renderings and a couple animations to answer planning commission and public hearing questions like… If I drive down this street on my bicycle about 15 mph, will I still see ‘nature’ or will I be staring at stark houses and concrete? If I stand on this intersection across the freeway from the project, will I be looking at a bare, 70-foot high wall, or will it be vegetated. How far will my line of sight be approaching a specific intersection, do I have enough time to stop if a kid runs out on the street, or do we need to lower a speed limit. Will we see nothing but rooftops instead of a quaint valley from out next-door high dollar horse property?

What we were able to offer, in addition to pretty pictures answering these questions is a numerical proof that they’re not looking at an ‘artistic interpretation’ – which in their minds never ends up looking that way when actually built - but that they are looking at a graphic representation of an accurate mathematical model. In the minds of public and city planners, it really adds to the believability of the presentation.

If I remember right the whole project took about 5-6 weeks worth of man-hours to produce, from modeling to rendering to post-work to assembling presentations, animations and printed materials.

I’m not sure what else to day about it. If you have specific questions, please ask away. I’m off to dig up few more images from the project, so you can see some details, and some WIP shots and experiments, and wireframes.   
Heh, I think my dinner is plenty cold by now....  :m_laugh:

Thnks for reading! Hugz, Connie

Hi, my namez: "NO, Bad Kitteh, NO!"  Whaz yurs?
BadKittehCo Store  BadKittehCo Freebies and product support