Ecliptic and Galactic coordinate grids

[selden]

I've created a spherical CMOD model of a grid with lines separated by 10 degrees. With appropriate DSC definitions, it provides a background of Ecliptic, Equatorial and/or Galactic coordinates.

The model is rather large (~350KB) because I wound up using 360 points around each of the 17 latitude circles so they'd look like smooth curves. I suspect the model could be made quite a bit smaller, but none of the methods I've thought of so far would make more than about a 30% difference.

http://www.lns.cornell.edu/~seb/celestia/coordinate-spheres.zip (90KB, 6Mar04)

Here's a picture of the Galactic coordinate sphere in use:

(as usual, this links to a much larger image)
The picture is rather cluttered with other Addons that I'll be announcing shortly, but the 10 degree Galactic coordinate system can be seen in the background.

See http://www.lns.cornell.edu/~seb/celestia/graticules.html for more information.

[Guest]

Cool. Thanks Selden!

One slight problem, I'm looking at Earth with the default new grid spheres (ecliptic?) in the background, and the grid lines are all glitchy - bits of them keep flickering in and out of existence. I tried looking at the Galactic grid and that seems more stable (it's not redrawing the grid lines all the time even when I'm not moving), but bits are still missing.

[Evil Dr Ganymede]

Cool. Thanks Selden!

One slight problem, I'm looking at Earth with the default new grid spheres (ecliptic?) in the background, and the grid lines are all glitchy - bits of them keep flickering in and out of existence. I tried looking at the Galactic grid and that seems more stable (it's not redrawing the grid lines all the time even when I'm not moving), but bits are still missing.

Dagnabbit. That was me. Darn these stupid forum bugs!!

[selden]

Evil Dr.,

Of course, I don't see that problem at all on my system. I turned on all the rendering options, went into sync orbit around the Earth, ran time at 10,000x and the Earth just sailed calmly past all the rock-stable lines.

The only differences between the three grids is the orientation of the model and the locations of the labels. They all use exactly the same model.

You might try disabling the label catalog to see if that makes any difference. I would expect not, except for an improvement in Celestia's fps rate, just because of all the objects it defines.

I have to believe that there's something buggy in the OpenGL code your card is using.

At least one person with an ATI card has reported somewhat similar problems with the model of the 2dF redshift survey.

Which version of Celestia are you running?
What are your system specs these days?

My system:
256MB 500MHz P3, WinXP Pro SP1
128MB GF4 Ti4200, Nvidia drivers v53.03
Celestia v1.3.2pre5 (and pre4)

[Bob Hegwood]

Selden,

Sorry to bother you, but I'm afraid that I don't have any idea how to view the Coordinate Spheres model. Can you tell me what I need to do to in order to view them?

I tried to rename the mesh from the current single blank " " to something I could reference, and then select it and go to it, but that didn't show me anything and then the control system via mouse went completely screwy.

Could you please elaborate on how to view this model? Us "Brain-Dead" types need a bit more help please.

Thanks, Bob

[selden]

Bob,

It doesn't work that way

The grids are intended just to be background. You don't GoTo them.

If you restore the zip file into your extras (or addons) directory and then run Celestia, the Equatorial grid should be displayed behind whatever you're looking at. In that respect, it's much like Celestia's builtin "Celestial Grid".

Uh oh, it may be one of those "Of course..." things that I didn't think to mention.

The Grids are defined as Nebulas.
In order to see the grids, you have to have "show Galaxies" enabled.
In order to see their labels, you have to have "label Galaxies" enabled.
(Nebulas = Galaxies so far as current versions of Celestia are concerned.)

I'll update the Web page and the Readme to include this.

Does this make them visible for you?

[Bob Hegwood]

Selden,

I'm afraid that I can't even use this feature. I did as you said to do, and my whole display immediately went screwy.

The eastern half of the Earth suddenly lit up to an Albedo of 2.0 or more, I got all kinds of strange half-arc lines on my screen, and the display basically just went Blooey!

Probably my limited graphics cababilities again. Sigh! Well, that just tears it. I'm simply going to have to get an NVidia graphics card installed.

Thanks though... It *did* display coordinate numbers on the screen, but the lighting was so screwed up that I simply had to remove the whole addon before Celestia acted as it should again.

Thanks, Bob

[selden]

Bob,

I'm sorry: I think you've encountered a bug in Celestia's CMOD code.

I just noticed something similar myself, but didn't want to report it until I'd had a chance to try to duplicate it on my system at work.

When Celestia uses "Basic" or "Multitexture" render paths, and a CMOD line drawing is in the background, the coloration of foreground objects gets screwed up. On my home system, the dark side of the Earth turns shades of bright orange.

If Celestia is using the "OpenGL vertex program" rendering path, it draws things fine.

*Arrgh*

Of course, now I can't duplicate it.
I think this is related to the problems Rassilon is having with Emissive objects, too.

[ogg]

As an old sky3d user I love that Celestia can now do this stuff. Displaying the stuff we know... mapping the universe... brain candy instead of eye candy (though I have nothing against eye candy). Unfortunately I have the same viewing problems as the people on this thread.

a) pinked-out night-side planets

b) the lines and dots don't show up at all most of the time, but 'twinkle' when rotated.

I have a bad, generic 16meg laptop graphics card, a Trident 'cyberblade' (you'd think with a name that pretentious it might actually do something) with only a handful of OpenGL extensions showing up in Celestia's OpenGL info (have been kicking myself over that purchase ever since). I can handle not being able to display big graphics, dds's and nicely shaded spacecraft models, but if it turns out I can't even see plotted points or gridlines... well... I'm gonna cry. (sad but true)

[MB]

I got similar problems as Evil Dr, The grids are flickering but the labels are OK.

MB

My system:
laptop P4 2.6 GHz
512MB , WinXP
64MB GF4 488GO Nvidia drivers v 49??
Celestia last cvs (today)

[selden]

MB,

Please read the faq: Q1/A1.

The new CMOD code requires more recent Nvidia drivers than you have. I had problems with planets being transparent with v43.03 on my Ti4200 and seeing triangular patches on planets with 47.xx (don't recall the last digits) on an FX5200. Upgrading to v53.03 fixed them both.

Ogg,

Be patient. I'm sure the Basic/Multitexture bugs will be stomped soon, certainly by the time v1.3.2 final is available.

[MB]

I just download the last 53.03 driver: Same problems. Actually, even before I had no problem to see objects such as amalthea.cmod.
MB

My system:
laptop P4 2.6 GHz
512MB , WinXP
64MB GF4 488GO Nvidia drivers v 53.03
Celestia last cvs (today)

[MB]

I just download the last 53.03 driver: Same problems. Actually, even before I had no problem to see objects such as amalthea.cmod.
MB

My system:
laptop P4 2.6 GHz
512MB , WinXP
64MB GF4 488GO Nvidia drivers v 53.03
Celestia last cvs (today)

[Evil Dr Ganymede]

I have to believe that there's something buggy in the OpenGL code your card is using.

I've got a GEForce4 Ti 4200 (can't recall how many MB it has - it's 64 or 128MB).

Either way, I installed the latest NVidia drivers and it works fine now - no flickery problems!

That galactic grid is an absolute godsend though (especially when checking star locations for my near stars project), selden - thanks!!