Page 1 of 2
16000 LY limit
Posted: 29.08.2005, 02:24
by buggs_moran
Without getting too technical, and for my sake, too insulted by this question, could someone explain the 16000 LY limit on stars??? Is it due to the Hipparcos catalog? I understand that the idea is to create a limit of knowledge program, is that it? My reasons are purely selfish. My active black hole for galactic cores is getting close to where I'd like to release it, and I JUST saw/played with fridgers new galaxy models (THANK YOU Fridger). BUT, I can't put it anywhere it belongs except as a mishmash dsc object, which would mean a bit of redesign and no animation...
Thanks
Posted: 29.08.2005, 07:27
by Guckytos
Hi buggs,
the limitation to within 16k LY lies within the OpenGL, not Celestia itself, as far as i remember the explanation.
I am trying to recreate a SciFi universe in the Milky Way and can't place stars where they should be.
So, unluckily you will have to find another place for the black hole.
Regards,
Guckytos
Re: 16000 LY limit
Posted: 29.08.2005, 08:10
by t00fri
buggs_moran wrote:Without getting too technical, and for my sake, too insulted by this question, could someone explain the 16000 LY limit on stars??? Is it due to the Hipparcos catalog? I understand that the idea is to create a limit of knowledge program, is that it? My reasons are purely selfish. My active black hole for galactic cores is getting close to where I'd like to release it, and I JUST saw/played with fridgers new galaxy models (THANK YOU Fridger). BUT, I can't put it anywhere it belongs except as a mishmash dsc object, which would mean a bit of redesign and no animation...
Thanks
Celestia being a 3d program, NEEDS information on the DISTANCE of the stars it displays. Distances of close-by stars (< 15-20 ly) were usually measured by means of the standard /parallax/ method. The HIPPARCOS satellite was able to push the upper distance limit via a most sophisticated improved exploitation of parallaxes by orders of magnitude. But anyhow each (distance) measurement has its practical limits
The limit of astrometric resolution of the HIPPARCOS parallaxes are around 1 mas = 0.001 arc seconds and mag <= 12.5 . It's about a hundred fold improvement compared to the earlier limit. Hence accurately known distances are still below 1000 ly!
It would contradict Celestia's design philosophy to just place additional "luminous dots" at random distances
Bye Fridger
Posted: 29.08.2005, 10:24
by selden
Please take a look at the thread at
http://216.231.48.101/forum/viewtopic.php?t=822
where Grant Hutchison describes the Hipparcos limitations in detail.
On the second page of that thread, Chris also discusses the software design reasons why Celestia currently cannot reliably draw stars past about 16KLY: the depth of the octree and the floating-point precision of the stars' locations.
Posted: 29.08.2005, 11:35
by t00fri
selden wrote:Please take a look at the thread at
http://216.231.48.101/forum/viewtopic.php?t=822where Grant Hutchison describes the Hipparcos limitations in detail.
On the second page of that thread, Chris also discusses the software design reasons why Celestia currently cannot reliably draw stars past about 16KLY: the depth of the octree and the floating-point precision of the stars' locations.
I couldn't find any "detail" in that thread beyond what I have quoted above:
HIP resolution=1 mas and mag<12.5. The corresponding 68%CL (1 sigma) limit on the distance follows trivially from there and the known size of the earth orbit.
I have spent considerable time with that data and know the physics limitations of their used method pretty well.
What else?
In any case, the 16k ly =5000pc software limit seems good enough, given that we
know quite little quantitatively about star distances beyond 1000pc.
Bye Fridger
Posted: 29.08.2005, 12:13
by selden
Fridger,
One of the things I'd like to be able to do is generate globular clusters in appropriate locations, using photographs and monte-carlo techniques with appropriate stellar population profiles from existing catalogs.
Unfortunately, there are only a few globular clusters within 16KLY.
Posted: 29.08.2005, 12:51
by t00fri
selden wrote:Fridger,
One of the things I'd like to be able to do is generate globular clusters in appropriate locations, using photographs and monte-carlo techniques with appropriate stellar population profiles from existing catalogs.
Unfortunately, there are only a few globular clusters within 16KLY.
Selden,
that sounds like an interesting project, but is it really affected by the 16k ly limit for /stars/. Clearly galaxies and clusters can and will be much further away than that. Just make a coordinate transformation and MC generate your stars in a frame with origin at the cluster center. That's what we do also with the galaxies: we generate a distribution of points and brightness values (from photographs) with coordinates counted from the galaxy center.
Bye Fridger
Posted: 29.08.2005, 13:39
by selden
How hard would it be to extend that code to be able to specify offsets for the locations of individual stars?
Globular cluster blobs already are being created in Addons by declaring them to be E0 galaxies. Generating a globe of individual stars is the next obvious step. (Rassilon wrote a globular cluster generator already, but it doesn't provide a way to adjust "scientific" parameters -- like matching to measured HR diagrams or radial density distributions.) I've played a little with creating CMOD models of points with different brightnesses and colors, but they're just another stopgap.
Re: 16000 LY limit
Posted: 29.08.2005, 15:08
by danielj
Regarding this matter,I think the limit for the ATI video cards are smaller.They cannot draw star until a distance far smaller than that.What are exactly the differences between ATI and Nvidia cards?
Thanks[/quote]
Posted: 29.08.2005, 20:01
by buggs_moran
t00fri wrote:
In any case, the 16k ly =5000pc software limit seems good enough, given that we know quite little quantitatively about star distances beyond 1000pc.
Bye Fridger
I understand that we wouldn't want stars strewn about willy nilly in the galaxy or the universe for that matter. I believe my problem is due to the fact that I cannot create a barycenter outside the 16KLY limit. I don't even need stars per se. "Technically" my addon would invlove stars by default as would Selden's globular clusters, but they would not be "part" of the star catalog and wouldn't fall under any deviation rules (the glass is more than half full at 68%
).
The STC and SSC files are the only ones that implement movement (as far as I understand the file system.) My addons are built with CMODs and PNGs, so the only way to give them rotation and orbital characteristics is through a barycenter. Unless I am missing something, maybe we could implement artificial barycenters as needed in the universe ? (i.e. at the center of active galaxies...
granted, we would probably have to have a STC type file that can be read by the program)
Posted: 29.08.2005, 20:10
by buggs_moran
First, sorry about the double post...
selden wrote:Please take a look at the thread at
http://216.231.48.101/forum/viewtopic.php?t=822Chris also discusses the software design reasons why Celestia currently cannot reliably draw stars past about 16KLY: the depth of the octree and the floating-point precision of the stars' locations.
Chris from 2002 post wrote: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.
YES! That's what we need. Guess I'll have to wait till Celestia 1.9pre7
Posted: 29.08.2005, 21:52
by selden
I believe my problem is due to the fact that I cannot create a barycenter outside the 16KLY limit.
Strange. That works for me. When Barycenters were first defined, I was able to define one positioned at the galactic center and have a star orbiting around it. I defined the orbit''s SemiMajorAxis to be large enough that the star passed by the Sun within the 16KLY limit and I watched it travel around the galaxy. (I just tried it and it works with 1.4.0pre6)
Remember, the operational problem isn't that large distances don't work at all, but that objects defined that way aren't drawn reliably.
Posted: 29.08.2005, 22:08
by buggs_moran
selden wrote:I believe my problem is due to the fact that I cannot create a barycenter outside the 16KLY limit.
Strange. That works for me. When Barycenters were first defined, I was able to define one positioned at the galactic center and have a star orbiting around it. I defined the orbit''s SemiMajorAxis to be large enough that the star passed by the Sun within the 16KLY limit and I watched it travel around the galaxy. (I just tried it and it works with 1.4.0pre6)
Remember, the operational problem isn't that large distances don't work at all, but that objects defined that way aren't drawn reliably.
Gotcha!, will have to try that tonite. Thank guys. I get it now, I just can't have "stars" as they look within the 16KLY...
Posted: 29.08.2005, 22:27
by selden
This STC file works for me. Remember that RA in STC catalogs is measured in degrees (not Hours) and SemiMajor Axis is in AU (not LY). I specified a short period just to make it easy to see the star move.
Code: Select all
Barycenter "GC" {
RA 266.25 # 17.75
Dec -28.93
Distance 28000
}
Star "Big_Orbit" {
SpectralType "O5"
AppMag -10
OrbitBarycenter "GC"
EllipticalOrbit {
Period 0.001
SemiMajorAxis 1e9
}
}
Posted: 30.08.2005, 08:18
by Dollan
So, in theory, I could have stars in the Andromeda Galaxy, for instance, as long as they are placed in a barycenter system?
...john...
Posted: 30.08.2005, 09:55
by selden
Using a Barycenter is irrelevant.
You can place stars anywhere.
I was just playing with stellar orbits to see if one could portray the motion of stars within the galaxy. You can, but...
Celestia still can't draw stars reliably beyond 16KLY from the Sun. They're visible from some viewpoints and not visible from others.
Posted: 30.08.2005, 10:01
by Dollan
Ah,. I see.
Well, still reason enough to hold off on some plans for a while.... Hopefully someday stars can be added beyond the limit, and still show reliably regardless. It would be fascinating to see some of the star sdetected in the LMC, or even interglaactic space.
Okay, going to bed now. Night hift is done and I apparently have lost the ability to type properly.....
...John...
Posted: 31.08.2005, 09:35
by Michael Kilderry
I've got another question.
We may not be able to reliably draw up stars beyond 16KLY in Celestia, but what about planets?
Posted: 31.08.2005, 11:22
by selden
That's unreliable beyond 16KLY, too.
Posted: 31.08.2005, 15:25
by buggs_moran
selden wrote:That's unreliable beyond 16KLY, too.
Yup, totally unreliable. I tried your star orbiting the center of the galaxy and it worked.
When I tried to add my CMODs to it, kaput.
Oh well, maybe I'll stay in our galaxy for now and keep working on Mira...