The New Horizons has crossed Neptune's orbit the day of Voyager's anniversary for the same milestone. To celebrate that, NASA has released the best map we have till now of Triton.
The best resolution is 600 metre per pixel, this is all we know of Triton. The contrast is a bit higher of what human eye would see.
I was thinking about something interesting for the TODO list (that I haven't found) that might be usefull for SE: It would be possible to generate an hybrid between procedural generation and non-procedural one? I mean for example, immagine we want to render Triton at it best, so we put the NASA map in the program, but there's a problem; there are only few patches of terrain detail and there are many different resolutions, so the render would look strange and flying over the surface would be very uninteresting. So maybe we could create an algorith that with map as base work it starts to generate mountains, craters ect... at a lower scale than the resolution of the image, with the geological features to immitate what is seen in the highest resolution parts of the image or simply showing what is espected to show at that level. This generation of procedural landscapes would have a condition: all the "noise" on the surface has to generate the image at large distances (so the colours and contours are fitted and linked to those of the real image). It would be like asking to the program to immagine the surface aspect of parts of the planet/moon where we have little or no information based on that little information and on what is expected to exist in the patches where no image has been ever taken (like the north hemisphere of Triton). For example, the Moon has a quite good map on SE, but it gets boring at surface level and extremely low orbit (something that dosen't happens with procedural surfaces) so the programm could immagine based on the lunar map a lot of features to explore on the surface to see hills with higher resolution (even if we don't have knoledge of those things) with more complex shaders ect... all fitting the whole panorama of the region's image.
Could that be possible?
Edited by FastFourierTransform - Monday, 25.08.2014, 12:39