Hard Light Productions Forums
Archived Boards => The Archive => Stellar Assault => Topic started by: Thaeris on January 05, 2011, 01:27:51 pm
-
Over the past few days, I've been digging up information on NASA FTPs. But, why?...
(http://amaqbw.bay.livefilestore.com/y1ppcRMn3sJMjsFnEcwAffsF67JqwUk_4eS5cItPvkT1KhRJU7Kj9QfpT3fiY7_YX9lrpHOsCiGZup5pAoDWp0mw7RdS9SjG7u8/MarsTest004.jpg?psid=1)
^ That's why. Our project is going to, if possible, use actual NASA terrain and elevation data to produce landscape models for Venus and Mars. Right now, the only really decent data I've got is for Mars via the MOLA satellite - the above image is from the lowest resolution available, where the data is interpreted by a free program from the NASA Planetary Data System. This program is called NASAview.
The problem at hand now is that I need to determine how to get elevation data, which a greyscale image doesn't readily show - instead, any image from Mars via the data is dynamic - the program will automatically adjust the shade of any selected region's max, min, and mean elevation data and spit out that information within 255 (254, technically) shades of grey. This is great, because for the most part we'll use these maps to generate 3D terrain from the greyscale image... but we still need to know the maximum and minimum altitudes of the region in question for this to work.
Fortunately, despite all of those silly politicians cutting education and the space program (you know, actually important things?), NASA is still made of win, so much so that they'll answer my questions about how to use the available data:
Dear [Thaeris],
The MOLA PEDR data set (Precision Experiment Data Record) is the source of the elevation values shown in the MOLA topographic maps. The PEDR data are not images but binary tables containing the along-track altimetry measurements recorded by the MOLA instrument. The data are online here:
http://pds-geosciences.wustl.edu/missions/mgs/pedr.html
These data files are in a custom binary format which is rather complicated. You would have to write special software to read them. An example of such software is given in the SOFTWARE directory of the archive.
Many users prefer not to tackle the daunting job of reading, interpreting, and mapping the altimetry measurements. A tool is available that will do this work for you. It is the MOLA Query Tool, part of our Mars Orbital Data Explorer (Mars ODE) service. ODE is primarily a service for selecting and downloading Mars orbital data from many instruments, but for MOLA in particular a special function is provided in response to the great interest from our users in this data set. With the MOLA Query Tool you may select from the data base of altimetry measurements by latitude, longitude, altitude, time, orbit, or product ID. Results can be returned in the form of ASCII tables, CSV tables, shapefiles, or binned images.
The ODE home page is here: http://ode.rsl.wustl.edu/mars. Look for "MOLA PEDR Query" under "Additional Tools". If you have questions about using this tool for your particular project, the best person to ask is ODE developer Keith Bennett, [email protected].
Good luck,
Susan Slavney
So, yes. this is entirely doable. The next issue, which unfortunately is more pressing... is Venus. The information from our last Venus mission, Magellan, is also available. Magellan's information, unfortunately, is early 90's vintage, and it shows. I've got a bit more research to do before I can put together precisely what I need to do or how I need to do it, but it will get there. Assumably, that might involve some more emails with technical questions, but ones does what they must.
Regardless, feel free to ask any questions if you have any. The NASA PDS data is freely available to the public, so if you think something would be of use to a project you're working on, have a look! However, you'll most certainly want to konw what you're looking for. And if you find it, be prepared to do the research to make use of the information you obtain. Again, I'll be happy to share any information I've learned so far.
-
didn't Herra already do this or something
-
he did sort of
our needs are a bit more specific and a lot more detailed.
In most science fiction stories, Mars is only half-way colonised and barely terraformed.
In SAFSO, its Fully Terraformed and the Population is large and healthy.
The same also goes for venus. The challenge is that with Venus specifically, there are several "near orbit" missions where spacecraft and fighters are moving about at Spaceshuttle altitude, and as a result we need a high resolution source to create a large "Terraformed Venus" skybox in similar vein to the existing earth skybox.
-
Furthermore, we do intend to have atmospheric missions, so the actual terrain is highly important as well.
-
Furthermore, we do intend to have atmospheric missions, so the actual terrain is highly important as well.
For Mars moreso, again, for venus there are so many non-descript islands there, that it would be easy to create a random island terrain in max and call it Venus.
I'd like to add that I am still on the fence for doing low altitude atmo missions only because Freespace does terrain so very poorly.
-
This is indeed understandable, but I think it's worth a shot - otherwise I wouldn't be doing this!
:)
-
ah got it, terrain models
well the big issue is the AI
-
It's always amused me that Descent II, the project from which the FS code base was semi-constructed, featured really great object avoidance AI...and then in the team's next game, fighters can't even manage to fly around an Orion without plowing into it. :p
Actual Mars terrain, though? That's awesome. :yes:
-
I look forward to the Noctis strike mission.
-
Actually, I do have a question/task for anyone who may be willing...
AC3D does have a terrain generation tool which will take a greyscale map and develop a landscape, but in my opinion, it's not that great a tool. Perhaps it's been refined in the latest version of the program, but I don't have that at my disposal. Blender, however, is free. A lot of you here also are quite comfortable using it. As it happens, it also appears that Blender has a similar tool for developing 3D terrain from greyscale images. If anyone would like to volunteer to assist in research or generating models, we'd appreciate that a lot (you might just get yourself a badge, too). Venus in particular, as it's from an old mission, has data (and a lot of it) which is broken down into tiny pieces so as to accomodate the computers of the early 90's; getting useful results for that might thus take a while, and a research partner could be most useful in making a clear picture of how to use the information.
-
I'll help, but RL may cause me to be absent from time to time.
I hope it won't be a problem, but I'll most likely be gone for the duration of winter holidays.
-
That shouldn't be a problem at all Dragon. I'll get us started with putting our information in order soon enough - having an assistant in this project will be a tremendous help!
By the way, would you be willing to take up a little Blender for this project?
-
I've had problems with it, but if I'll be able to install it, I may try.
-
Actually, I do have a question/task for anyone who may be willing...
AC3D does have a terrain generation tool which will take a greyscale map and develop a landscape, but in my opinion, it's not that great a tool. Perhaps it's been refined in the latest version of the program, but I don't have that at my disposal. Blender, however, is free. A lot of you here also are quite comfortable using it. As it happens, it also appears that Blender has a similar tool for developing 3D terrain from greyscale images. If anyone would like to volunteer to assist in research or generating models, we'd appreciate that a lot (you might just get yourself a badge, too). Venus in particular, as it's from an old mission, has data (and a lot of it) which is broken down into tiny pieces so as to accomodate the computers of the early 90's; getting useful results for that might thus take a while, and a research partner could be most useful in making a clear picture of how to use the information.
Are you talking about a possible Displacement mapping techinique?
Basically all you do in displacement mapping is get an appropriately hi-poly quad-based sphere, and apply a greyscale bitmap to move the vert's up and down depending on their proximity to the bitmap.