![]() |
|||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||
![]() |
|||||||||||||||||||||||||||
![]() |
Radiolaria are holoplanktonic protozoa, a form of plankton found throughout the worlds oceans. One of the most interesting aspects of these organisms is that they possess a skeleton composed of amorphous silica. Their skeletons take on a diverse set of forms from ornate geodesic-like polyhedral lattices, spike encrusted spheres, to other complex geometric designs. For more information and images of these amazing organisms be sure to check out Radiolaria.org. This image came as a result of my interest in the look of electron micrographs and the amazing geometrical forms that these organisms exhibit. By the time I started on this image I had pretty well figured out the materials and lighting needed to replicate the SEM (scanning electron Microscope) look. I have no doubt that it can be improved but it's a fair approximation. The real challenge with this scene was how to best construct the fairly ornate morphology of the subject. After picking through countless images on the web and in books I felt that I had better pick something fairly simple. There are thousands of more complex forms but I needed to first figure out how this type of geometry could be built in Max without having to build each lattice section by hand. By choosing one of the spherical based forms I was hoping that I might be able to use a geosphere primitive and go from there. As it turned out I tried several other strategies including isosurfaces before coming back to poly modeling. In the end a simple geosphere, some tweaking and meshsmooth fit the bill. Here you can
see the poly mesh used for the image prior to being refined at render
time. The final technique to get to this stage ended up being fairly easy.
The process of figuring out how to get there took a bit of experimentation
though. |
||||||||||||||||||||||||||
![]() |
|||||||||||||||||||||||||||
Next, select edges that delineate hexagonal sets of faces. You'll notice that the geosphere can't be divided up evenly into hexagons; there will be several left over tris and a few pentagons. While these rogues lead to some uneven lattice cells they tend to blend into the rest of the form in the end. Another option would be to just select all of the tris and go to the next step, but it too creates its own look. With the edges selected use Chamfer to add a width to form the basis of the cells. The width will have a lot to do with the final ratio of openness/lattice. You may want to try this a few times to get a feel of how it will look. Select all the interior faces and delete them. And, yes this can take awhile on a complex mesh. For the next step you'll need to grab a copy of Andrey Zmievski's invaluable Solidify plugin. If you haven't used it before you really owe it to yourself to play with it. Essentially it extrudes a face or selection of faces based on their respective normal. It's a quick way to add thickness to a mesh and that exactly what is called for with this object. I've extruded the surface inward about the same amount as the width. This will give a fairly round cross section once the Meshsmooth has been applied. Lastly apply MeshSmooth and crank up the iterations till you get the desired smoothness. Naturally,
the final image involved tweaking around the mesh a little bit to add
both the spiky bits and some randomness to the size and locations of the
cells. The technique certainly worked well for this form but as you can
well imagine a similar approached could be used to provide a variety of
forms that would be difficult to produce through other means. 04/27/03
|