Just a quick posting so this is in the right forum:
The maps of Mars shipped with Celestia have 0 degrees at the right edge.
solarsys.ssc follows the IAU standard and assumes the use of maps with 0 degrees in the center.
As a result, Mars is drawn with the wrong side toward the Sun.
See http://63.224.48.65/forum/viewtopic.php?p=8720#8720 for details.
Added later: Of course after complaining, it turns out that the real problem is indeed with the definition of the rotaiton of Mars in solarsys.ssc.
See the thread http://63.224.48.65/forum/viewtopic.php?t=1331
sigh.
Mars Maps off by 180 degrees
-
Topic authorselden - Developer
- Posts: 10192
- Joined: 04.09.2002
- With us: 22 years 2 months
- Location: NY, USA
Mars Maps off by 180 degrees
Selden
-
granthutchison
- Developer
- Posts: 1863
- Joined: 21.11.2002
- With us: 22 years
Orientation and registration of texture maps
No, you are right, too!
The various manipulations I've been carrying out orientate the planet so that the seam of the applied texture map is in the prime meridian position - it so happens that the Mars maps are laid out that way, so everything looks fine.
But if you try to use Goto Object to get to a specified longitude on Mars, you quickly find out that the Mars map is 180 degrees out of register - Celestia, as you say, expects the prime meridian to be at the centre and not at the edge.
Now that I'm spending time getting the planets and satellites correctly orientated, this is becoming an annoyance - while you can easily get to the right longitude on Mimas, you'll find yourself in the wrong place on Io or Callisto or Mars.
Ideally, all Celestia maps should be edited to bring the zero meridian to the centre: this means a cut-and-paste job on a relatively small number of maps - the Gallileans, Mars, a couple of others. I'll be doing this for my own use anyway, so I'll happily provide the altered maps to anyone who wants them, together with revised information for the RotationOffset.
I've just finished generating a surface map for Venus, but the current north-pole down orientation of that planet in Celestia creates another problem for the Goto Object function. The Magellan data call the pole north of the ecliptic the north pole - so the Magellan map has to be flipped over to display in the correct orientation on Celestia's Venus. But then Goto Object will take you to the map's southern hemisphere when you enter a northerly latitude. (Simplest solution is to use the standard Magellan map, and flip Venus into a north-up, but retrograde-rotating, orientation.)
Finally, there's a Goto Object problem relating to the 3ds shape models - because of their inside-out definition (which flips applied maps into mirror images), all the latitudes and longitudes in the Goto Object function are reversed, too. I can't see a solution to that one without someone knowledgeable in 3D Studio fixing the shape definitions for these satellites.
Grant
The various manipulations I've been carrying out orientate the planet so that the seam of the applied texture map is in the prime meridian position - it so happens that the Mars maps are laid out that way, so everything looks fine.
But if you try to use Goto Object to get to a specified longitude on Mars, you quickly find out that the Mars map is 180 degrees out of register - Celestia, as you say, expects the prime meridian to be at the centre and not at the edge.
Now that I'm spending time getting the planets and satellites correctly orientated, this is becoming an annoyance - while you can easily get to the right longitude on Mimas, you'll find yourself in the wrong place on Io or Callisto or Mars.
Ideally, all Celestia maps should be edited to bring the zero meridian to the centre: this means a cut-and-paste job on a relatively small number of maps - the Gallileans, Mars, a couple of others. I'll be doing this for my own use anyway, so I'll happily provide the altered maps to anyone who wants them, together with revised information for the RotationOffset.
I've just finished generating a surface map for Venus, but the current north-pole down orientation of that planet in Celestia creates another problem for the Goto Object function. The Magellan data call the pole north of the ecliptic the north pole - so the Magellan map has to be flipped over to display in the correct orientation on Celestia's Venus. But then Goto Object will take you to the map's southern hemisphere when you enter a northerly latitude. (Simplest solution is to use the standard Magellan map, and flip Venus into a north-up, but retrograde-rotating, orientation.)
Finally, there's a Goto Object problem relating to the 3ds shape models - because of their inside-out definition (which flips applied maps into mirror images), all the latitudes and longitudes in the Goto Object function are reversed, too. I can't see a solution to that one without someone knowledgeable in 3D Studio fixing the shape definitions for these satellites.
Grant
-
chris
- Site Admin
- Posts: 4211
- Joined: 28.01.2002
- With us: 22 years 9 months
- Location: Seattle, Washington, USA
Orientation and registration of texture maps
granthutchison wrote:Ideally, all Celestia maps should be edited to bring the zero meridian to the centre: this means a cut-and-paste job on a relatively small number of maps - the Gallileans, Mars, a couple of others. I'll be doing this for my own use anyway, so I'll happily provide the altered maps to anyone who wants them, together with revised information for the RotationOffset.
I agree completely that all of Celestia's planet maps should have the zero meridian at the center. But . . . it should be possible to compensate for a nonstandard map--without screwing up Goto Object--by using the Orientation field. For example, this line:
Code: Select all
Orientation [ 90 0 1 0 ]
should have the same effect as specifying a RotationOffset of 90 without messing up Goto Object. Still, just modifying the maps is a better idea . . .
--Chris