A planetarium sphere within Celestia

[selden]

This addon requires Celestia v1.3.1pre3 or later.
It does not work in Celestia v1.3.0 or earlier versions.

Starting with v1.3.1pre3, Celestia supports large Deep Sky objects. You now can use a sphere within Celestia as a "planetarium". By this I mean that a sphere can be used to show panoramic maps of the entire sky surrounding the solar system. Various all-sky surveys are starting to make their results available as low resoution mosaics that can be displayed this way.

Here's a simple example, showing a map of the infrared sources that are in the galactic plane.



Oh, sorry, I guess that view is a bit narrow-angled .
Here it is from a distance.
(As usual, these thumbnails link to much larger images.)

It's a sphere with a 10,000 LY radius centered on the solar system with a PNG image of the MSX IR mosaic as its surface texture. Details can be found at http://www.lns.cornell.edu/~seb/celestia/billboard.html#7.2

Addendum:
At 20:30 U.S. Eastern Daylight time on June 16th, I updated the model of the sphere so that a surface texture map with 0 degrees of longitude in the center will properly align with the galactic center. The original model was off by 90 degrees. I also corrected the accompanying MSX map to have 0 degrees of longitude in the center.
As a result, the appearance of the addon is unchanged, so you don't need to download the new version unless you plan to use this particular 3DS model for your own galactic maps.

[Borg Collective]

Exactly what we needed for our Kharak System from Homeworld...

[AstroBoy]

I had never thought about this possibility but now you pointed it to me, I have got some interesting idea.

For example a "Cosmic Microwave Background Map". I'm gonna look for it...

[JackHiggins]

With the new alternate surfaces feature, we could cycle through all wavelengths, from radio all the way up to gamma rays!

[Borg Collective]

9 months ago we posted a request for similar thing. Yesterday, it turned into reality.

[selden]

The annoying thing is that most of the all-sky maps posted on the Web are Aitoff projections. Surface maps for a sphere need to be Plate Carree projections (aka simple cylindrical). Maybe one of our friends at NASA can use a little gentle persuasion.....

[selden]

oops.

I just discovered that this add-on requires Celestia v1.3.1pre3 or later.

It does *not* work in Celestia v1.3.0 final.

[JackHiggins]

The annoying thing is that most of the all-sky maps posted on the Web are Aitoff projections

By Aitoff you mean this:

right? (taken from the 2MASS site http://www.ipac.caltech.edu/2mass/)

Are there any (free) conversion plugins for photoshop etc to make something like that into a cylindrical projection?

It couldn't be that hard, could it...?

[MalcolmP]

Nicely done Selden, I like it !

The annoying thing is that most of the all-sky maps posted on the Web are Aitoff projections. Surface maps for a sphere need to be Plate Carree projections (aka simple cylindrical). Maybe one of our friends at NASA can use a little gentle persuasion..... ;)

Or perhaps we can persuade Chris to incorporate an Aitoff projection option ?
The software "PROJ" described on this page http://stommel.tamu.edu/~baum/linuxlist ... ode37.html
seems to suggest the capability to convert between many systems including Aitoff, but when I went to the home site it is all in Unix-speak which I dont understand so I cant check it out.

Axel Mellinger( http://home.arcor-online.de/axel.mellinger/ ) has photographed a very nice all-sky mosaic and presents it in both Aitoff and Cylindrical side by side, so he should know how to do it ! He says that he used "GIMP for flat-fielding and Aitoff projection" but I cannot see Aitoff in my Windows-Gimp. Maybe we should send him a request !

[selden]

Jack,

Yup, that's an Aitoff projection. Some of the Web pages with maps like that explicitly say that's the projection they're using.

A related problem is that those maps are quite low resolution: a 1Kx512 image is only about 1/3 of a degree per pixel. Except for the line at the equator, transforming directly from one image format to the other can only make it worse, since the original pixels have to be blurred into adjacent pixels in the output and then merged back together again by the graphics hardware. I think it'd be best if the stewards of the data could be persuaded to create the appropriate projections directly from the original data.

Malcom,

My (possibly mistaken) understanding is that cylindrical maps are used directly by the graphics hardware when they're projected onto objects that are defined as spheres. I'll admit I'm not sure how the projection is handled for 3DS models like this one.

It would be great if Axel could be persuaded to provide a copy of his Milky Way mosaic for use with Celestia, but I suspect that won't happen. I can imagine lots of legal entanglements with his various publishers which could prevent it. He spent several years travelling around the world to photograph all the pieces, so it's worth quite a bit of money, too.

There's a low resolution version of the mosaic on his Web site which can be used by individuals, but it's copyrighted. That means that only Axel may reformat it and publish a version for Celestia on the Web for others to use. He'd have to give explicit permission for someone else to do so. That copy of the map also has some constellation lines which some people might find objectionable.

[MalcolmP]

Malcom,
My (possibly mistaken) understanding is that cylindrical maps are used directly by the graphics hardware

Ah, I see, yes that would sort of tie Chris's hands a bit ! Thanks for the clarification.

There's a low resolution version of the mosaic on his Web site which can be used by individuals, (snip) That copy of the map also has some constellation lines which some people might find objectionable.

Sorry, my mistake, I thought that the constellation lines were just on a frontpage demo of the cylindrical map, I was not concentrating sufficiently. My excuse is that I was reading and concentrating on understanding his description of how the various distortions were handled on this page http://home.arcor-online.de/axel.mellin ... n_web.html which was all very interesting. The ref to GIMP and Aitoff was made in another forum/page but he doesn't give any further info about what plugin or perhaps it was his own,,, but with all the legal stuff you mention it looks as though I introduced a red-herring into the discussion ! sorry !

So, it looks like back to waiting for a maths genius to rustle us up a converter prog !

[selden]

Malcom,

I'm not sure if it's a "red herring". I know I'd really like to have a high resolution copy of his mosaic to use with Celestia, but I'd like it to be one I could share.

As I mentioned in my response to Jack, I'm afraid that converting the public Aitoff maps wouldn't give us images that look all that great, although itt's certainly a place to start.

Note that the copy of Axel's map on his Web site is close enough to being a simple cylindrical projection that it works quite well when projected on a sphere. It just needs to be rescaled and have some padding added to the top and bottom.

And if it's used with the sphere I created, it has to be rotated 90 degrees. I still haven't found an accurate workaround for that. Manual adjustments aren't precise enough. *grump*

[JackHiggins]

Yup, that's an Aitoff projection. Some of the Web pages with maps like that explicitly say that's the projection they're using.

I was in a hurry looking at the page- I didn't see that on the caption page it said "rendered in an Aitoff projection"...

A related problem is that those maps are quite low resolution: a 1Kx512 image is only about 1/3 of a degree per pixel.

They actually have a link to a 4700x2400 image at http://www.ipac.caltech.edu/2mass/galle ... y_full.jpg but it's 4.6MB, so if someone wants to download it & check it out be my guest...

I think it'd be best if the stewards of the data could be persuaded to create the appropriate projections directly from the original data.

Have to agree there!

And as I mentioned in my response to Jack, I'm afraid that converting the public maps wouldn't give us images that look all that great, although itt's certainly a place to start.

For a start, someone could make a texture of just the galactic plane, like you did for your addon above- then the distortion wouldn't be too bad, even if it didn't cover the whole sky...

[selden]

At 20:30 U.S. Eastern Daylight time on June 16th, I updated the model of the sphere so that a surface texture map with 0 degrees of longitude in the center will properly align with the galactic center. The original model was off by 90 degrees. I also corrected the accompanying MSX map to have 0 degrees of longitude in the center.

As a result, the appearance of the addon is unchanged, so you don't need to download the new version unless you plan to use this particular 3DS model for your own galactic maps.

[Guest]

If you still have a use for it, here is a routine to convert Aitoff latitude/longitude to Galactic latitude/longitude. I imagine you can then transform Galactic l,b via the projection of your choice. The function assumes that all angles are in radians.

Code: Select all

void aitoff2gal(double alat, double along, double *glat, double *glong)
{
  double x, y, z, horiz;

   x = -0.5 * along;
   y = -0.5 * alat;
   z = sqrt(4. - x*x - 4*y*y);
   
   *glat = asin(y * z);
   horiz = 0.5 * x * z / cos(*glat);
   *glong = 2. * asin(horiz);
   if(horiz<0) *glong += PI;
}


[/code]

[selden]

Thanks!