When it comes to global communications, geography has been kind to Fiji. The island nation is located on a major submarine cable route between Australia and the United States and, as a result, has access to plenty of bandwidth. Enough, in fact, to share with neighboring Tonga and Samoa.
Papua New Guinea, on the other hand, is comparitively remote and poorly connected.
But a new generation of communication satellites is changing that.
The ring of satellites – 12 have been launched so far – orbits the Earth at about 5,000 miles, much lower than geosynchronous satellites, which orbit at 22,200 miles. As they pass over ground stations, the signals they transmit are handed off from one station to the next, with the lower orbits providing faster transmission times. The idea is to provide broadband Internet service to emerging markets around the world, primarily in the southern hemisphere.
The owner, O3b Networks, is named for the estimated “other 3 billion” people in the world who lack dependable Internet connections. This page from the company’s website explains the technology behind the satellite system. Here is a map of the coverage area.
O3b’s satellites are launched four at a time atop Russian-built Soyuz ST-B rockets at the European Space Agency’s Guiana Space Centre.
Here is a video of a December 2014 launch.
This is fascinating stuff.
But, back to our graphic
I’m often amazed at what you can turn up with a bit of persistent Googling. But no amount of Internet searching yields actual 2D diagrams or blueprints of the O3b satellite. The best I can find, from various sources, are low-resolution images of satellite construction and preparation for launch.
You need good reference material to create an accurate 3D model. In this case that would mean diagrams or blueprints and high-resolution images. We lack both, so our model won’t include as much detail as we might like. However, because the satellite image will be only a small part of the graphic, that should be OK.
We select a representative group of the low-resolution images and use Photoshop’s measuring tools to estimate the satellite’s overall dimensions. Then, in Maya, we begin with a box and reshape it to fit.
A single dish antenna, made from a spherical NURBS surface and a polygon support structure, is duplicated to create the satellite’s antenna array.
Some detail – as much as we can glean from the images – is added. We have to be careful here, though, since it’s better to leave details out as opposed to making them up.
Here is a rendering of the unshaded model, before any textures are applied.
The rendered surfaces will need to convey a sense of realism. We’ll create a digital image to simulate the various textures – metal, foil, ceramic – and map it to the model.
Unwrap and map
Actually, this process, called UV mapping, works the other way around. First, we create the map, and then the image.
The positions of vertices, edges, and surfaces of our 3D model are all located using XYZ coordinates in 3D space. UV coordinates, on the other hand, locate points on the model’s 2D surface.
It’s like a wrapped gift. The gift box exists in 3D space. But the wrapping paper, when removed and flattened, is a 2D surface. The pattern on the paper can be described and mapped with a 2D coordinate system.
When I first learned 3D modeling, UV mapping (for some reason) was the hardest part for me to get my head around. But really, there’s nothing to it. The process, which can be tedious, is a vital step in the creation of a believeable model. Time spent making a clean and usable UV map is never wasted.
Our low-resolution satellite model is simple to unwrap in Maya’s UV editor.
With the map loaded into Photoshop, we can add additional layers and, using a combination of imagery and painting tools, create the foil and other surfaces of the satellite.
A similar grayscale image will serve as our bump map, which will provide the illusion of small displacements in the model’s surface.
When the image and bump images are applied to the model’s surface, they line up perfectly because they are mapped to the model’s UV surface coordinates.
Bright directional lighting simulates illumination provided by the Sun, and a second fill light simulates reflected Earthlight. An environment image of the Earth, rendered in Terragen, is added to the scene to provide subtle reflections in the solar panels and antenna dishes.
Time to put everything together.