No stars beyond 16308.35 ly's from Sol

[jrobert]

I travelled to the edge of the celestial star universe in an attempt to find the farthest star from Sol (our home star) and found something interesting. The farthest stars are exactly 16308.35 ly's from Sol. I counted about 200 or so stars with this distance. Now that can't be right, can it? Or is this the limitation of the Hippocros catalog?

PS: I'm running a P4 1.6Ghz, 512MB, Geforce2 GTS w/32MB DDR VRAM.

[Rigel]

I think that this is not a bug in Celestia

When the stars database of Celestia is generated from the Hipparcos catalog, and when a parallax value is dubious in this catalog, then Celestia put this star at such a distance from the sun

Tell me if I'm wrong...

[jrobert]

I think I should rephrase my question... What I'd like to really know is: Why are there no stars beyond 16308.35 ly's from Sol?

[Rigel]

:o

I hope there will be an astronomer around there to give us a correct answer, but from what I know about astronomy, that's because stellar distances in Celestia are based on the parallax value of the Hipparcos Catalog, and this method - ie : measuring the parallax angle of a star" - can only be done on quite near stars. When astronomers want to know distances of more distant stars, they use other methods than parallax...

[NecroBones]

Actually, I think the more important question is why are three so many with the SAME distance. That shouldn't happen, unless the distances are being truncated, or rounded, or otherwise hard-set based on unavailable data. So is it a limitation of Celestia? Or the Hipparcos catalog? Or both?

[julesstoop]

Might this have to do with Hipparchos' maximum measuring resolution? In other words, when a parallax comes close to the smallest measureable value, it gets truncated to that value; if it's yet even smaller, it gets rounded of to zero. Hence, many stars end up at the specific distance associated with this smallest possible parallax, and no stars beyond.

By the way, I have noticed it as well in quite a funny way. If you fly up to about a distance of 100.000 ly +, and look down towards the sun, with the star brightness on maximum, you can actually see brighter stars sitting on a circle a circle at this 16000 lightyear distance, and (besides the galaxy clouds) darkness beyond.

[billybob884]

Maybe its because people dont know the exactdistances of the stars and they dont know of anyother stars beyond these. Just a guess


Mike M.

[ou8poop2]

well, i updated the star database to the newest version, and found several stars with this as the farthest distance:
put this into your favorites.cel file

"fathest" {
isFolder false
parentFolder ""
base [ 7089.238800749067 -1580.958021665661 24015.40832756709 ]
offset [ 1.275541319145646e-005 -4.664833141987401e-008 7.505682233954758e-006 ]
axis [ 0.103199 -0.0962932 0.989989 ]
angle 2.4815
time 2447763.83118801
selection "#0"
coordsys "ecliptical"
}

now, i realize that this is most likely eroneous data from tthe catalogs, but i thought there was a limit placed by chris when coding that prevented stars from showing farther than 16000 ly. So why can we see this star? if you move too far in any direction, it disappears, so are these the effects of those limits? It seems that as long as you face home to Sol, you can see it, so can you place objects farther out for now, and get the same results? And Chris, when will you ( or how can we now ) change those limits so as to see farther obects?

[granthutchison]

16308.35 ly is exactly 5000 parsecs, corresponding to a pretty tiny parallax of 0.2 milliarcseconds. I don't know what the parallax limit for Hipparcos was, but it seems to me that this sphere of stars represents a whole bunch of stars with marginal parallaxes that were all rounded to 0.2 milliarcseconds. I wouldn't be surprised if this represented the "edge" of the Hipparcos dataset, beyond which parallaxes were deemed unobtainable to any usable precision - a star with a parallax of 0.1 milliarcseconds would be twice as far away as one with 0.2 milliarcseconds, so rounding errors would swamp the data.

Grant

[Rigel]

You're right.

Those stars correspond generally to Hipparcos entries with negative parallaxes, that could not be corrected by the mean of their B-V color indexes, which is the procedure I used to generate the new stardatabases for all Tycho stars with unknown parallax values.

Therefore, they are put at that limit of 5000 pc (a few hundreds of stars on more than 2 millions). There is no mean to have a better estimation of their distances. They could be eliminated, but it seems better to keep them in the database.

Also, don't forget that we have only a good estimation of the distance of about 100000 stars today and no more, as a result of the Hipparcos mission. And as Alexis said in another post, the detectors of the Hipparcos satellite were not distance limited, but magnitude limited.

Pascal

[granthutchison]

And as Alexis said in another post, the detectors of the Hipparcos satellite were not distance limited, but magnitude limited.

This is interesting. So the "edge" to the data (apart, that is, from the artefactual 5000 pc sphere) is determined by the fact that there are vanishingly few stars bright enough to allow Hipparcos to detect their parallax at such extreme distances? So could the Hipparcos detectors have detected less than 0.2 milliarcseconds of parallax, if the star had been bright enough?

Grant

[Rigel]

It's a bit more complicated than that. If you are interested, you should visit the ESA website

http://astro.estec.esa.nl/Hipparcos/hipparcos.html

to learn more about the Hipparcos mission...

[granthutchison]

It's a bit more complicated than that. If you are interested, you should visit the ESA website
http://astro.estec.esa.nl/Hipparcos/hipparcos.html
to learn more about the Hipparcos mission...

I do find that site a bit user-beligerent - I've spent a while there in the past and still haven't quite got to grips with the observational limits of the Hipparcos survey. (As is evident from my question, no doubt! )

Grant

[alexis]

This is interesting. So the "edge" to the data (apart, that is, from the artefactual 5000 pc sphere) is determined by the fact that there are vanishingly few stars bright enough to allow Hipparcos to detect their parallax at such extreme distances?

Well, yes and no. The average parallax error of the Hipparcos catalogue is about 1 mas, slightly dependent on the apparent magnitude (brighter stars usually have somewhat more accurate parallaxes than fainter ones) but generally independent of the distance to the star.

Thus a measured parallax of (2 +- 1) mas would place the star anywhere between 300 pc and 1000 pc with 1-sigma confidence (68% likelihood), and a parallax of (0.2 +- 1) mas would correspond to a distance between 800 pc and infinity with the same confidence.

The Hipparcos sample is complete down to about V magnitude 8, though some stars in it are as faint as V = 12 mag.

To take this into perspective, the nearest known stellar neigbour of our Sun is Proxima Centauri, at a parallax of 772 mas (distance 1.3 pc) and an apparent V magnitude of 11.0. Thus even out to the very nearest stars the Hipparcos catalogue is not complete down to the luminosity of a Proxima Centauri class star!

At the other extreme we have the red super giants that at a distance of 2000 pc still could be as bright as V = 8 mag.

So could the Hipparcos detectors have detected less than 0.2 milliarcseconds of parallax, if the star had been bright enough?

Hipparcos cannot distinguish parallaxes smaller than 1 mas from zero, even though it could (and did) measure such small parallaxes. That's why one has to take care when using such measurements. The relative error of the parallaxes just inflates for small parallaxes, as you pointed out yourself.

It is therefore very sensible to clamp distances derived from Hipparcos parallaxes, as done in Celestia. One can be quite certain that all stars seen by Hipparcos must be closer than 5000 pc (except possibly some rare cases). And it really doesn't make much sense using parallaxes that yield infinite relative distance errors anyway. A practical absolute lower limit on sensible parallax measurements would be about 1 mas, corresponding to 1000 pc.

That's all for now, have a very Merry Christmas, all of you!

/Alexis

[granthutchison]

Alexis - thanks for taking the trouble to reply.
I am grateful; and better informed!

Grant

[chris]

There are actually two other limitations in Celestia related to stellar distances . . .

First of all, the size of the root node of the octree structure used for fast visibility processing is 15000 light years. That's an easy value to change, but the bigger problem of floating point precision comes into play . . .

Single precision floating point numbers have 23 bits in the mantissa. This is a precision of about 1 in 8,000,000. At a distance of 80,000 light years, one 'step' is 1/100 light year. Celestia does some special handling of stars closer than one light year to the camera, but outside of that range, it's all single precision arithmetic. At such great distances from the Sun, nearby stars would jump as you moved by them . . . There are several solutions to this that I'm considering. I could simply increase the range for special handling, but this has performance implications. The performance problems could be offset by increasing the special-handling range as the camera's distance from the Sun increased. This isn't very general, but it would work reasonably well with the existing database, where the star density increases toward the origin. The real long-term solution is to allow multiple star databases, each referred to a different center. This is analgous to the way that solar systems are handled now, with the coordinate system for planets 'rezeroed' to the location of central star.

--Chris

[Larre]

I am an not really an expert at the Celestia source code, but I did download it from the Source Forge repository.

This Octree structure that I've been looking at had a lot of nodes, but I did not see the root node, but that's because, I've just been skimming over the code.

I really need to take a closer look at the octree.cpp file. If I find any bugs in it, I will let everyone know on this forum.

I can see where Chris is coming from on this topic. The best thing to do is to create many octree structural files for the many star databases that would be rendered in Celestia.

These should have a center, in the center of every galaxy. The larger octree files would need to be expanded in larger galaxies, while the current size is perfect for the smaller galaxies and globular clusters.

Keeping in mind that larger galaxies, like M31, also known as Andromeda, are over 50,000 ly in diameter. These galaxies may need more than one star database per galaxy to be rendered.

The other option would be creating and octree file for every star, and that would take a very long time, since there are billions of stars out there in space, and would slow Celestia down significantly..

Hope this helps..

Peace, Larre..

[Chuft-Captain]

The real long-term solution is to allow multiple star databases, each referred to a different center. This is analgous to the way that solar systems are handled now, with the coordinate system for planets 'rezeroed' to the location of central star.

Code: Select all

"Sol" "MilkyWay"
{
    Class "Star"

    OrbitFrame {
        TwoVector {
            Center "MilkyWay"
            Primary {
                Axis "x"
                RelativePosition { Target "MilkyWay" }
            }
            Secondary {
                Axis "y"
                RelativeVelocity { Target "MilkyWay" }
            }
        }
    }

}


[Reiko]

The real long-term solution is to allow multiple star databases, each referred to a different center. This is analgous to the way that solar systems are handled now, with the coordinate system for planets 'rezeroed' to the location of central star.

Code: Select all

"Sol" "MilkyWay"
{
    Class "Star"

    OrbitFrame {
        TwoVector {
            Center "MilkyWay"
            Primary {
                Axis "x"
                RelativePosition { Target "MilkyWay" }
            }
            Secondary {
                Axis "y"
                RelativeVelocity { Target "MilkyWay" }
            }
        }
    }

}


Does that work??

[selden]

Reiko,

No, it doesn't work. Yet.

ChuftCaptain is teasing us!