Celestia Forums

The idea of using a really large data base of galaxies etc. in
Celestia is very exciting /and/ challenging, indeed.

I think Paul's work was a major pioneering step in this
direction.

Yet, it would be most desirable, to have the (familiar) galaxies
look more or less realistic and at /precise/ locations at least relative to
earth. Then amateur astronomers could use Celestia as well;-).
In Paul's extended galaxies.dat, most galaxies are effectively
oriented according to the default (axis,angle) values,

axis=(1,0,0); angle=0.0 deg,

which is mostly /not/ realistic.

The Tully catalogue Paul uses, still employs 1950 instead of 2000 coordinates.

In the ambitious NGC/IC project, http://www.ngcic.com,
there are meanwhile improved/corrected catalogues for all NGC/IC objects. In
Xephem we have used the one by Wolfgang Steinicke, where every object
matches the DSS/RealSky /photographic/ data by less than O(1 arc sec)!
Implementing this accuracy into Celestia costs virtually no extra effort...

Also, the galaxy input parameters in galaxies.dat are quite far from
the input that is given in standard catalogues.
I strongly vote for changing the input in galaxies.dat to a ~standard
one that is /close to actual measurements/ and calculate Celestia
related quantities in galaxy.cpp /after/ reading in the file.

This definitely opens more perspectives for the future.

Normally in catalogues, the appearance of galaxies is characterized by

mag (for some ref. light wave length)
smallest apparent size d
largest apparent size D
position angle (many PA's are actually incorrect in popular
catalogues; they must be mirrored at 90deg!),

and empirically, one may use in most cases the formula for the
/inclination i from face-on/:

i = 3deg + acos(sqrt(((d/D)^2 - 0.2^2)/(1 - 0.2^2)))


in galaxies.dat, however,

radius
distance
axis (a unit vector)
angle
----------------------------------------------------------------------
Chris:
======

It would save me and others quite some time, if you could quote from your
"notebook" the relation between those quantities as used in Celestia!
-----------------------------------------------------------------------

With O(1000) galaxies and more to come;-) it is now definitely time
for a small Perl programme (that I can produce quite easily) that
extracts the respective parameters from standard catalogues and
converts them into the form required in galaxies.dat!

This way, TYPOS are eliminated and experimentation is much easier.
So everyone who is able to run Perl, can try out various changes in a
reproducable manner...

Bye Fridger

Fridger, I agree with everything you just said . . . It's time to change Celestia's galaxy database to use parameters that more closely match the ones used in astronomical catalogs . . . What's the largest galaxy database out there with distances included? Is it large enough to warrant switching to a binary format for the galaxy database?

--Chris

chris wrote:Fridger, I agree with everything you just said . . . It's time to change Celestia's galaxy database to use parameters that more closely match the ones used in astronomical catalogs . . . What's the largest galaxy database out there with distances included? Is it large enough to warrant switching to a binary format for the galaxy database?

--Chris

Well, the distance estimate of galaxies is an independent and often quite uncertain issue, using e.g. the period-luminosity relation of Cepheids together with a particular value of the Hubble constant etc. Most catalogues therefore refrain from quoting the distance. Tully's catalogue does provide a distance estimate, but the very accurate one's usually do not. Joining distance info from one catalogue with high precision position/shape info from other catalogues via a Perl script is of course feasible and straight forward.

It seems preferable though to use accurate data on position/shape of galaxies and invent some more or less rough model for their distance in case it is not available directly.

The Tully catalogue Paul uses, still employs 1950 instead of 2000 coordinates.

Well, actually I use the 2000 coords calculated by VizieR - I don't use the 'raw' catalogue so much as a VizieR query of all objects in the catalogue, which calculates 2000 coords for me.
Once I've imported it into a spreadsheet I can then write formulas to create radius, distance, etc. At some point I would also like to implement an Excel script to export the data automatically; unfortunately I still haven't quite figured this out.
The main problems I have with Tully's catalogue is that it doesn't have complete orientation data, and also some of the distance values seem questionable to me. Keep in mind that I've actually examined images of every galaxy in galaxies.dat, and some definitely look like they should be much further away (sometimes, nearer). It's almost certainly just the shortcomings of using recession velocity as a distance indicator - after all, galaxies have their own motion.

In the ambitious NGC/IC project, http://www.ngcic.com,
there are meanwhile improved/corrected catalogues for all NGC/IC objects.

That's well and good, but nearly half of the galaxies in Tully's catalog aren't NGC objects, a few of which I have added to galaxies.dat. I was planning to add most of them at some stage. I think it's important to derive all the positional information from one catalogue, because it tends to be internally consistent w.r.t. the overall structure of the galaxy groups.

I agree with this as well...I also would implore a way to add in individual nebulae/galaxies in ssc or stc or maybe a different file alltogether in the extras directory...With the ability to use custom models or the default software rendered ones...That way we all can contribute to getting the task of filling the cosmos with realistic...relatively speaking...looking galaxies