Celestia Forums

Hello Celestia users !

You can download two new generated stars database for Celestia at the following link:

Version 2.0 A : 1 044 025 stars
Version 2.0 C : 2 102 323 stars

http://perso.wanadoo.fr/celestia.stars/index.html

I hope there will be no connexion problem with my webpage, and that Chris will mirror those databases "one of these days" (ah... Pink Floyd ...!)

The stars in the two new databases were extracted from the ASCC catalog of 2.5 millions stars (more information on this catalog on my web site).

They are more precise than the previous 1.x releases.

Version 2.0 C is a huge file (33.4 Mb compressed) but seen from Earth, the Milky Way looks beautiful in Celestia with it ...

Best regards,

Pascal

Wow..and I thought 2.0 A was pretty looking
I'm downloading 2.0B right now, and I can't wait to see how it looks now

Wow. Looking at the screenshot of 2.0 C brings to mind that famous line from Arthur C. Clarke's book, "2001: A Space Odyssey".

And yet it's still only a tiny fraction of the entire Galaxy...

- Hank

Pascal,

The new stars databases are great! They add a lot to the visual impact and provide more things for us to explore. Thanks, a lot!

For example, I had fun looking and staring and twirling it around, investigating the way the stars defined the galactic plane, how some clusters were organized, etc.

However, I think I see some hints of non-physical structure in both 2.0A and 2.0C. It's actually easier to see in A. Some of it seems to be "systematic errors" -- I think I see a hint of an arc of a few bright stars centered on the Sun's position.

There also seem to be several spokes radiating from the sun in the direction of the arc. I can imagine several possible causes -- systematics in the observing program (studying some areas more than others), errors in assignment of RA and Dec, or even dark nebulae blocking (other) parts of the view.

At any rate, below I've provides a viewpoint taken from favorites.cel that seems to make some of the spokes more easily visible in 2.0A using Celestia v125pre3. One spoke points up toward about 12:00 and another toward 10:00. The sun is roughly in the center of the screen. Rotating the "sun" back and forth a little (with the right mouse button) makes the arc a little easier to see as the relevant stars move back and forth together.

"spokes" {
isFolder false
parentFolder ""
base [ -9560.51608016755 8635.427028550232 -1641.979748980645 ]
offset [ -8.851289749145508e-006 3.421157671393282e-013 2.870250153004517e-013 ]
axis [ 0.871896 0.299327 -0.387558 ]
angle 1.88479
time 2452539.473444656
selection "#0"
coordsys "ecliptical"
}

Thanks again.

Selden,

It's not easy for me to see what you mean, because I am using V2.2 of the Mac OS X release. Thus, scripts don't work, there is no stars browser, and so on...

I don't think there could be some systematic error on the data of the ACSS catalog which is based on Hipparcos and Tycho catalogs, at leat in RA or DEC, because those data were correlated with ground based catalogs.

In which constellation did you see an arc of bright stars ? (as seen from the Sun)

Best regards

Pascal

Pascal,

Unfortunately I only have versions 1.2.4 and 1.2.5pre3. I tried downloading 1.2.1 just now and the transfer hung after barely getting started.

I did a "capture image" (creating a 128KB Jpeg) from a point of view that seems to show the "spokes". It's on the Web along with a description of how to find that pov, which is near HD 109214. The commands include only Select, HCG keys and the Goto menu window, so hopefully you can use them.

See http://www.lns.cornell.edu/~seb/celestia/index.html#spokes for details.

I hope this helps.

Seldenm

I've followed your procedure to see these "spokes". And what I saw doesn't seem to be anormal.

The extraction procedure has rejected stars with very bad color indexes or bad parallax values, and the Tycho-2 catalog which is included in the All Sky Compiled Catalog reveals some low star density due either to nearby galactic dust clouds, or to variations in the limiting magnitude caused by the scanning pattern of the satellite (as said in the documentation of the Tycho 2 catalog).

Nothing is perfect and as Hank said, these 2 millions stars are still only a tiny fraction of the entire Galaxy...

Pascal

when i try to d/l 2.0 C it gets to 1 second left then stops, but the window says it's going at 54kbps

Pascal,

Thanks for taking a look. I'm not at all surprised that thee'd be effects like that -- like sunlight shining past the edges of clouds, although in this case it's our satellite's vision seeing past their edges

Thanks again for the effort you've put into generating the databases. It's greatly appreciated.

Thank you Selden for your appreciation !

16 of 18, I think the problem comes from my web site. As I said before, it is a personal web page with no support.

Could someone here mirror both the A & C databases or does anybody know where I could find some free web space for them ?

16 of 18,

Don't forget that it is a very large file. C is 34MegaBytes, which will take about 5,000 seconds at 54Kilobits/second. That's almost an hour and a half. (kb is ambiguous; if you meant bytes, it'll still take about 10 minutes or so.)

You might want to investigate getting one of the 3rd-party download assistants. Apparently some of them are able to restart hung file transferrs. I don't use one myself, so I can't suggest a good one.

I hope this helps a little.

So how does celestia run with all stars visible?

On my system (500MHz P3, 128MB Ti4200) Celestia1.2.5pre3 takes about 10 seconds before opening the initial screen while loading v2.0A, about 20 seconds for v2.0C.

With maximum star visibility, the initial view of the earth (filling most of the screen) and a full-screen display at 1600x1200, frame rates are rather low, about 1-4 fps while moving the earth slightly. Turning down star visibility increases the framerate substantially: 30 fps is easy to maintain with a significant number of stars visible. (It'd be nice if the "star visibility" controls included a disply of "limiting magnitude".)

i'm on bb, what do ya think i mean seldon

Rigel wrote:Thank you Selden for your appreciation !

16 of 18, I think the problem comes from my web site. As I said before, it is a personal web page with no support.

Could someone here mirror both the A & C databases or does anybody know where I could find some free web space for them ?

Well, I think Rigel's server is potentially quite powerful and this is a problem on the /client/ side!

Yesterday, I downloaded both 2.0A and 2.0C with amazing 1500KB/sec
in less than half a minute!

Bye Fridger

Well, gee, Fridger, you do live practically next door to Pascal, after all

Links across the pond tend to limit the transfer rate to somewhat less.
Also, some of the routes that packets travel among the various networks in the U.S are, shall we say, less than optimal. For example, I live only three blocks from work, but packets from my home system to there have to travel almost 200 miles through a dozen routers to another city and back.

I suspect the popularity of the new stars DB has something to do with the problem, overloading some of the slower links between here and there.

My impression is that the links to Shatters.net are somewhat overloaded at the moment, too. I suspect someone uploaded some rather large textures recently, which are extremely popular

selden wrote:Well, gee, Fridger, you do live practically next door to Pascal, after all :)

Links across the pond tend to limit the transfer rate to somewhat less.
Also, some of the routes that packets travel among the various networks in the U.S are, shall we say, less than optimal. For example, I live only three blocks from work, but packets from my home system to there have to travel almost 200 miles through a dozen routers to another city and back.

I suspect the popularity of the new stars DB has something to do with the problem, overloading some of the slower links between here and there.

My impression is that the links to Shatters.net are somewhat overloaded at the moment, too. I suspect someone uploaded some rather large textures recently, which are extremely popular ;)

;-)

Yes, but still: 1.5MB/sec is quite something, and it may still be 1500Km to Rigel's server from Hamburg...Of course, we do entertain pretty fast connections to "certain" places on the other side of the pond, like Cornell;-)

Bye Fridger

PS: Uploading those mentioned files to shatters.net went with only 35KB/sec via ssh with the fastest Blowfish encoding...

only 35kb/s? im lucky if i get 5

I've mirrored Pascal's star database files on shatters.net:

http://www.shatters.net/celestia/files/starsdb2.0A.zip
http://www.shatters.net/celestia/files/starsdb2.0C.zip

These are wonderful additions to Celestia . . . I have an idea for a further improvement. I notice that there are a number of star clusters visible from Earth, but when you approach the clusters, it becomes apparent that the stars are spread out over a great distance radially from the Sun. This is expected because of uncertainties in the distances of stars--it's apparent to a smaller degree for clusters like the Pleiades in the default Celestia star database. Is there a catalog of star cluster members (ideally using Tycho catalog numbers) that would allow you to constrain the distances of cluster members so that they fall in a roughly spherical region of space. This would be both a realistic and visually spectacular enhancement. Imagine drifting through young star clusters where 100+ bright stars are packed into a region only a 10 light years across!

--Chris

The Astronomical Data Center http://adc.gsfc.nasa.gov/ at NASA Goddard has a very large collection of online star catalogs (large as in *all* of them, I think). If you specify the name of the cluster plus an appropriate modifier, like "parallax" or "distance" or "tycho", its search engine at http://adc.gsfc.nasa.gov/adc/sciencedata.html will list all of the associated catalogs.

For example, specifying Hyades and Tycho returned "J/A+A/367/111 A Hipparcos study of the Hyades cluster (de Bruijne+, 2001) " (and others). Apparently they managed to reduce the Hipparcos errors for that cluster (6%) by a factor of 3. See http://adc.gsfc.nasa.gov/adc-cgi/cat.pl?/journal_tables/A+A/367/111/