Rendering Saturn's Rings

• Create a 1-dimensional transparency map. I obtained the raw numerical data for mine from Bjorn Jonsson's planetary rendering site, and digested it using an IDL routine to create a map that would have correct geometry when wrapped with spherical geometry around an axis located a pre-determined distance (I chose 2 Saturn radii) above the ring plane. The rendering of the rings here was done using a 1 x 12,000 pixel mask- because it's one dimensional it can be very high resolution without using much memory. The nature of spherical mapping means that only about 1/4 of those pixels can be used for the actual rings- the rest are blank. Here is my 1 x 12,000 pixel transparency mask, in .BMP format- download using "Save Link As".
• Create a plane larger than Saturn's rings, and center the texture axis the predetermined distance above the center of the plane. Apply the mask as a transparency mask, with spherical mapping around an axis perpendicular to the ring plane.
• That's all that's required for the lit side of the rings- if the ring color is something light, both the rings and their shadow on the planet will look right. There might be a problem with the rings being too dark when the sun is a small angle above them- fixing this requires some of the techniques required for the dark side of the rings.
• The dark side is more complex because, like most rendering programs, Imagine can't handle the diffuse scattering that illuminates the dark side of the rings. A different model is needed, that in addition to the transparency map has a brightness map, with similar values and identical projection to the transparency map, but set to zero where the rings are completely transparent, where the transparency map's value is 255. Here is my 1 x 12,000 pixel dark-side brightness mask, in .BMP format- download using "Save Link As".
• Because the dark side of the rings has a brightness map, it will glow in the dark where Saturn's shadow crosses it, and we don't want that. My solution is to use the Imagine slice function to cut out the part of the ring that's in shadow (using a cylindrical slicer), and color that part black (while maintaining the transparency map, so the dark ring will still block light from the planet where it's thick). The resulting ring model will only work for a single sun angle, but I haven't yet found a better solution.
• The same "artificial shadow" will work on the lit side of the rings, and allows addition of extra brightness to the lit rings to make them brighter when the sun angle is low.

Return to John Spencer's Home Page