Get Celestia v1.3.2 from SourceForge. They have many mirrors for best
performance:
http://sourceforge.net/project/showfiles.php?group_id=21302
[back to Contents]
Celestia makes use of the most advanced features of OpenGL that your computer's graphics driver claims to support. Many older OpenGL implementations have serious bugs. Here are some options for improving Celestia's display, with the most likely ones first:
Under Windows, open the "Display Properties" window. Select the "Settings/Advanced/Troubleshoot" tab. (not the "Troubleshoot..." button). Move the "Hardware acceleration" slider all the way to the left. Click on the "OK" buttons to change the settings in use. This will cause Windows to use Microsoft's Generic OpenGL v1.1 library, which is limited, but seems to have relatively few bugs. It does everything in software, works on 2D displays, and is quite slow.
Your graphics chipset and its drivers aren't drawing bumpmaps and normalmaps properly: the OpenGL routine "GL_ARB_vertex_program" is defective. (This is often seen with the newer Intel graphics chips.) Assuming you've already installed the most recent drivers,
celestia.cfg
. Remove the # that's in front of the line
# IgnoreGLExtensions [ "GL_ARB_vertex_program" ]
Addendum provided Tech Sgt. Chen with regard to performance issues:
Shut down all background programs on your system before running Celestia
(i.e., antivirus software, multimedia software such as REAL Player,
Musicmatch, etc.)
Graphics programs are notorious for consuming system resources and even
the best of graphics cards are better off without competing for those
resources.
[back to Contents]
Report the exact circumstances and details of your hardware and software in the "Celestia Bugs" forum.
For example:
Hopefully you know the System information. If you're running Windows, System details usually are available in the Control Panel's System Properties menu.
Some of the Graphics information can be found in Celestia's Help menu.
If you're running Windows, more details can be found in the Control
Panel's Display Properties menu.
[back to Contents]
FX 5nnn
or GF 6nnn
chipsets
with 128MB of memory or more will show all of Celestia's eye candy.
For example, an inexpensive FX 5200
will show all of the eye candy drawn
generated
by Celestia v1.3.2 and v1.4.0, although not as quickly as more expensive
cards.
All other cards have limitations or problems when used with the current versions of Celestia.
Background:
Starting with Celestia v1.4.0, you will be able to see multiple shadows
cast by multiple light sources if your card supports
"GL_ARB_shading_language_100"
and "GL_ARB_fragment_shader". These routines are part of the
OpenGL V2.0 standard, but are included
in the v66.nn and later ForceWare graphics drivers for
Nvidia FX 5nnn
and GF 6nnn
cards and
in the v4.nn and later Catalyst graphics drivers for
ATI's Radeon 9500
cards and for
their higher numbered Radeon cards.
Starting with Celestia v1.3.2, improved eclipse and ring
shadows are visible if your card has "GL_ARB_fragment_program" and
uses floating point to do the graphics calculations. Unfortunately,
Celestia v1.3.2 supports these features only on Nvidia
FX 5nnn
and GF 6nnn
cards and not on any others.
In order for Celestia to be able to show bumpmap shadows and specular reflections, your card's OpenGL library must support OpenGL v1.4. In particular, it needs to include "GL_ARB_vertex_program".
As of December, 2004, only Nvidia and ATI have invested the effort in providing OpenGL v1.4 and later. Most of the other graphics chip vendors have not yet upgraded their OpenGL libraries. However, Matrox Parhelia cards seem to have many of the necessary v1.4 features in their v1.3 libraries. The OpenGL library provided with Intel's newest embedded graphics chipsets claims to include "GL_ARB_vertex_program", but the routine does not yet work properly. See Q/A #1.1 above.
Note: The authors of Celestia only have cards with Nvidia graphics
chips. Other types of cards sometimes have problems. See
Q2 above.
[back to Contents]
Celestia v1.3.2 includes only stars that had their distances meassured by the Hipparcos satellite. Hipparcos was not used to measure the distances to many dim, variable or close double stars. Someone may have created an Add-on that includes your stars, though. Or consider creating the necessary STC file yourself and contributing it.
For example, see Grant Hutchison's Addon of stars within 25LY at
http://www.lepp.cornell.edu/~seb/celestia/hutchison/nearstars.html
This catalog and others are being added to Celestia v1.4.0.
[back to Contents]
The authors of Celestia spend their time improving the software. Documentation is provided by users. Some people have written about specific aspects of Celestia. An introductory User's Guide by Frank Gregorio is available in English and has been translated into several other languages. Please consider translating it into a language you know.
Links to various versions of the documentataion are on the Web.
Start with
http://www.shatters.net/celestia/documentation.html
and
http://www.celestiamotherlode.net/catalog/documentation.html
[back to Contents]
Their diameters are exactly right. As seen from the earth, they both are about a half-degree across. Celestia's window is about 45 degrees across, so the Sun and Moon are about 1% of that. They are drawn only 10 pixels wide if your screen is 1024x768.
Remember that your computer screen is only about 10-20 degrees wide in your own field of view. Celestia's 45 degree field provides a "wide angle" view of the sky. This makes objects look smaller than you might expect.
The apparent large size of the Sun and Moon as we see them in the sky is a psychological illusion. There are several different explanations for this. If you take a picture of the moon with a camera lens that has the same field of view as Celestia, you may be surprised at the small size of its image.
At least one book has been written about this effect:
The Mystery of the Moon Illusion:
Exploring Size Perception, by Helen Ross and Cornelis Plug
A review of their book is at http://www.astronomyedinburgh.org/publications/journals/46/moonillusion
A simple experiment was suggested by "HarrieS", a guest on the Celestia forum:
Here is something you might try: a finger at arm's length is about two degrees wide for most people. That means that you can fit four moons side by side. Go outside and check it. Now have a direct look at your finger indoors. Can you still believe that four moons will fit on it?
And Dalle of the Forum wrote
But if your head is positioned e.g. 70 cm away from your e.g. 17" monitor, which measures 32 cm across (at least mine does), then the effective field of view looking at the Celestia sky "through" your monitor screen is 2*arctan((32/2)/70)) ~ 26 °. Hence, if you decrease the Celestia field-of-view to 26° you may get a more accurate appearence of what you would see looking out the window.
Sometimes a bright area is recorded where the file requestor menu or any other window overlaps Celestia's main window.
This bug seems to be fixed in Celestia v1.4.0. Prereleases of v1.4.0 are available for Windows, but there are many new bugs in them.
Here are some suggestions for minimizing this problem:
Windows includes a "print-screen" command. Under XP, you can use the keyboard combination Ctrl Print-Screen to write the screen image to the clipboard. You can then use any Paint program and Paste the image into it.
The real things are dim, grey blobs. Your eyes are not sensitive to color at the very low light levels emitted by distant galaxies.
( Catalogs of additional dim, grey blobs for use with Celestia v1.3.0 and later can be found at http://www.lepp.cornell..edu/~seb/catalogs.html.)
Many of the colorful pictures you're used to seeing are enhanced by long exposures on sensitive color film. Others are imaginative "false color" combinations of narrow-band CCD images designed to make visible the specific features of interest to the investigators. The colors of those pictures aren't realistic at all.
You can add a colorful object to Celestia by creating a 3DS model with
appropriate images as surface textures. Define it as a Nebula in a DSC
file. Use the Search command in the Celestia "Development" and "Add-On"
forums to find examples.
[back to Contents]
This is a bug in Celestia similar to Selden's *FAQ #8* above. To make a good clean screenshot image, press [Ctrl + D] before you capture the image. This cancels Multiview.
(provided by ElPelado)
[back to Contents]
You have to press Ctrl+k. Markers can be turned on and off in the "Render/View Options..." menu of Celestia v1.3.1 pre3 and later.
(provided by ElPelado)
[back to Contents]
On the second page of the thread mentioned above, at http://www.shatters.net/forum/viewtopic.php?p=9424, Chris wrote about some of the tradeoffs in the current version of Celestia.
March 25, 1752, is when Great Britian and its colonies converted from the Julian calendar to the Gregorian calendar. This changed the day on which the new year started and also lost 11 days from that year.
The Windows version of Celestia uses Microsoft's date/time input routines. Microsoft didn't want to bother with the complexities of converting between Gregorian and Julian dates. This problem does not exist in the Linux version of Celestia. Chris hopes to replace the Windows "Set Time" routine with a better one in a future release of Celestia.
In the meantime, you can use the "reverse time" keyboard command [J] and
the time speedup [L] and slowdown [K] commands to get to the time you want.
[back to Contents]
To draw the orbits, Celestia only calculates 100 or so precise locations and then draws straight lines between them. If the orbiting body doesn't happen to be close to one of those 100 points, then it won't be very close to the line, either. The position of the orbiting body is calculated very accurately. The lines aren't.
Celestia only calculates a few points around the orbit in order to minimize the amount of computation needed between frames. The more calculations are required, the slower the frame rate is.
Starting with Celestia v1.3.2, you can specify the number of segments in
an orbit track. Use your favorite text editor to modify the file
celestia.cfg
. Change the line
OrbitPathSamplePoints 100Increase 100 to 500 or larger.
Briefly:
A slightly out-of-date summary of .CEL commands is at http://www.shatters.net/celestia/docs/scripting.html. Also see http://www.lepp.cornell..edu/~seb/celestia/celestia_notes.html#4.0 for a list of more of the .CEL scripting commands, somewhat cryptically organized.
An example .CEL script is at http://www.shatters.net/~t00fri/images/start-pandemo.cel
A few example Lua scripts for Celestia are available on SourceForge at http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/celestia/celestia/scripts/
When invoked from a command line, Celestia can be passed the name of a .CEL or .CELX script to run at startup:
./celestia --url name-of-script.cel
When the command line includes the qualifier "--once
", the command line
will be passed to the running copy of Celestia instead of starting a new
copy of the program. (note: use two hyphens)
./celestia --once --url name-of-script.celx
(Starting with Celestia v1.3.1, scripts may be in any folder. Previously they
had to be in Celestia's "root" folder.)
[back to Contents]
The human eye can adapt to a very wide range of illuminations, so it wouldn't be as dark out there as you think - about the level of moderate indoor lighting, in which you can see very well. At the other end of the scale, your computer monitor is physically incapable of generating the brightness of illumination that pertains on the inner planets. Fortunately it doesn't need to, since all that would happen would be that your pupils would constrict to reduce the incoming light to a more comfortable level.
So in summary -
a) There's no way Celestia can display "realistic"
brightnesses on your computer screen; but
b) such "realism" is
unnecessary because your eyes merely adapt to compensate.
(Provided by Grant Hutchison)
[back to Contents]
The developers maintain a list of all new functionality and bug fixes in the Celestia ChangeLog, which is located at SourceForge: http://cvs.sourceforge.net/cgi-bin/viewcvs.cgi/celestia/celestia/ChangeLog. Look at the top-most entry in the list, to the right of where it says "Revision #.##", click the "(view)" link to display the most recent ChangeLog.
(Provided by Don G.)
[back to Contents]
(Thanks to DaveMc for these tips!)
Here are three things you can try...
1. Check the OpenGL anti-aliasing setting of your graphics card. If it
is on, try turning it off, or setting it to another option. This seems
unrelated to the anti-aliasing setting inside of Celestia.
2. Set your graphics card OpenGL options to "default".
3. Get the most recent drivers for your graphics card.
(Provided by Don G.)
[back to Contents]
If you're running Celestia v1.3.0 or later
0) Select the body from which to observe; the Earth maybe icon_smile.gif
1) use the GoTo menu to specify your Longitude and Latitude and GoTo there
2) type a y = Sync Orbit
Under Windows:
3) type a Ctrl-G = GotoSurface
4) type a Ctrl-F = change Arrow keys to AltAzimuth mode
Under Linux:
3) type an Alt-S = GotoSurface
4) type an Alt-F = change Arrow keys to AltAzimuth mode
Unfortunately, there is not (yet) an alt-az coordinate system display,
but typing a ; will turn on the equatorial coordinate system display.
[back to Contents]
Set Celestia's simulation time to be when the spacecraft was in orbit. Celestia does not display spacecraft if they are not in orbit at the time of the simulation.
For example, Mir was launched on February 20, 1986, and reentered the Earth's atmosphere on March 23, 2001, at 05:55 GMT. Celestia will show Mir only if you set the time to be somewhere between those two dates.
This is controlled by Beginning and Ending directives in the definition of Mir in solarsys.ssc. If you remove those statements, Celestia will always draw Mir in orbit.
Addendum provided Tech Sgt. Chen:
Mir, along with certain other Space Craft models, were modeled
within an historical time frame and can only be viewed from their
starting date to their ending date. You can override ending dates
by opening the solarsys.SSC or other specific craft related .SSC
files and placing a pound sign (#) in front of the ending date string.
Then save the new setting. This way you can always view your
installed Space Crafts. Consequently, removing the pound sign
will return the craft to it's natural time frame.
[back to Contents]
You have a misaligned map.
Celestia requires that all maps have 0 degrees of longitude in the center, with 180 degrees of longitude at the edges. All of the surface texture image maps of all of the moons and planets that come with Celestia have this alignment.
In contrast, many maps of Mars were created with 0 degrees of longitude at the left and right edges, and with 180 degrees of longitude in the center. This includes the maps on WWW.Shatters.Net in Fridger's "Texture Foundry".
Using a map with 0 at the edge would cause the symptoms you describe.
You need to cut the map in half and exchange the halves or you need to
find a map that's properly aligned.
[back to Contents]
You need to use a smaller ring image, one that is no wider than your graphic card's OpenGL texture buffer.
To see how large your OpenGL texture buffer is, use Celestia's menu Help "OpenGL Info". Near the beginning it has a line that starts with "Max texture size:"
Older versions of Celestia scaled down large ring texture images so they'd fit into the smaller texture buffer of your OpenGL graphics driver. Celestia v1.3.1 and later no longer do that.
Some older cards, like 3dfx Voodoos, only have a 256 byte buffer. Modern cards have a 2K or 4K buffer. Microsoft's software version of OpenGL only has a 1K buffer.
The orbit of the International Space Station changes continuously in ways that are almost impossible to predict due to things like atmospheric drag, light pressure, cargo ship docking, etc. If you want an accurate orbit, you'll have to update it on a daily basis.
ISS TLEs (Two Line Elements) are posted to the AMSAT SAREX mailing list regularly by "Dave Larsen PhD". See http://www.amsat.org/amsat/archive/sarex/31day/threads.html
Here are the ISS TLE orbital parameters for August 16th, 2004:
ISS 1 25544U 98067A 04229.23839543 .00019757 00000-0 15906-3 0 4532 2 25544 51.6323 19.1941 0005251 117.9988 304.8582 15.70921896327755
Grant Hutchison has provided a spreadsheet to convert TLEs into Celestia SSCs at http://www.lepp.cornell.edu/~seb/celestia/hutchison/spreadsheets.html#2
Also, don't forget that Celestia models the shape of the Earth using a
spheroid. The actual shape of our planet is much more complicated. As a
result, a view from the Earth's surface in Celestia is not accurate
enough to show the correct path across the sky of satellites in low
Earth orbit like the ISS. In other words, you can't use Celestia to find
out where to look in the sky to see the ISS.
[back to Contents]
You can have more than one copy of Celestia on your system at the same time.
Rename the directory (folder) where you have Celestia now, maybe to Celestia131.
Verify that things still work: Double-click on the icon for Celestia131\Celestia.exe and look around in your universe.
Install the new copy of Celestia and tell it to use the directory Celestia. It'll create the folder again and install itself there.
Move your Addons at leisure.
Cel:// URLs run the copy of Celestia that is in the folder named Celestia. If you want to change back to using your old Celestia that way, just rename the directories again: rename Celestia to be Celestia132 and rename Celestia131 to be Celestia.
Also, please take a look at
http://www.lepp.cornell.edu/~seb/addon-intro.html for a description of
how to organize your Addons so they can be moved easily.
[back to Contents]
Many sources for planetary maps (such as the USGS) use a mapping convention called "ecliptic north" - the north pole of any planet or asteroid is defined as being whichever rotational pole points north of the plane of the ecliptic (the plane of the Earth's orbit around the Sun). Each planet or asteroid can then be classified as a "direct" rotator, if it rotates in the same direction as the Earth when observed from above its north pole (ie counterclockwise); or as a "retrograde" rotator, if it rotates in the opposite direction.
But Celestia uses an equally valid mapping convention called "rotational north" - north is defined as being the pole around which the planet appears to rotate counterclockwise, irrespective of that pole's orientation relative to the ecliptic.
For direct rotators, this difference is irrelevant - "north" turns out to be the same direction in both mapping conventions. But for retrograde rotators, Celestia's north pole corresponds to ecliptic south - so you will find many maps of retrograde rotators like Venus, Ida and the moons of Uranus that appear to be upside down relative to those in Celestia. If you want to convert such maps for use in Celestia, you'll have to turn them through 180 degrees.
(Provided by Grant Hutchison)
If you don't tell me that something's missing, unclear or wrong, I can't improve it.