95 Herculis is wrong

[_MackTuesday_]

Celestia says it's G5V but it's more likely something like F5III. SIMBAD gives A5III but its color index is about 0.4. Wouldn't that indicate an F class star?

[Guest]

95 Herculis is probably more like F5II.

Another one that appears to be wrong: HIP 87902. Its absolute magnitude is 1.00 but Celestia says it's K0V. It's probably K0III.

[selden]

(I made a significant change in the wording below after my initial, hurried, response. Sorry.)

Remember that Celestia's star database is based directly on the Hipparcos catalog. Any errors or omissions in the Hipparcos catalog will show up in Celestia. That's not to say that the entries shouldn't be improved, though!

95 Her is a wide double star, so you have to be careful which of the two members you are referring to.

According to my Simbad search:
95 Her B = HD 164668 is G5III

HD 164668 = HIP 88267 and is the one Hipparcos (and thus Celestia) claims is G5. I seem to recall that if Hipparcos does not specify the luminosity class (which it doesn't for this star), Celestia assumes it's V.

95 Her A = HD 164669 is A5IIIn
and is not in Celestia.

The other star you mention also has no luminosity class specified in the Hipparcos database.

[Guest]

HIP 4786: Abs mag -0.78, Class G5V

[selden]

HIP 4786: Abs mag -0.78, Class G5V

I'm sorry: why do you mention this?

These Hipparcos (and Celestia) values agree with Simbad, except that Simbad doesn't mention any luminosity class (remember: default=V)

[granthutchison]

Another one that appears to be wrong: HIP 87902. Its absolute magnitude is 1.00 but Celestia says it's K0V. It's probably K0III.

HIP 4786: Abs mag -0.78, Class G5V

Or the Hipparcos distances are wrong, causing an error in the absolute magnitude calculation ... both of these are double stars, which Hipparcos notoriously has trouble with. I don't think we should "correct" stars.dat without being sure of the correct answer.
Have you considered signing in, so we at least know whether you're one person or two?

Grant

[selden]

sorry, I overlooked the sign. I could believe a glitch in data entry, since Simbad's B & V values are
7.587 and 6.489

[Guest]

HIP 14753: Abs Mag -3.72, Class G5V

[selden]

Guest,

If you're going to point out mistakes, please include the correct values and your sources for them.

It might be reasonable, for example, for Celestia to be enhanced so that Replace works for STC catalogs just as it does for SSC catalogs. Then an Addon could be used to update the stars database appropriately. However, I suspect that's not going to happen without appropriate citations.

[Guest]

HIP 7389: Abs Mag -6.12, Type K1V

[_MackTuesday_]

I'm the one pointing out these errors.

What I've been doing is listing stars whose absolute magnitudes don't match their luminosity class. I thought there were just a few but now I see that there are a ton of them. I probably should have just read the manual, I'm sure it says somewhere that a "V" gets stuck on the end when the HIP catalog doesn't give a luminosity class.

So I'll stop now. I hope I didn't annoy anyone. Just trying to help.

[granthutchison]

Celestia needs to at least make an assumption about the luminosity class and spectral subclass in order to come up with a temperature, and from that (plus apparent magnitude and distance), a radius for each star.
That said, my own feeling would be that such assumptions shouldn't be passed on to the user ... if Hipparcos says "G", then Celestia can internally assume "G5V" and come up with a best-guess temperature and radius (which won't be horribly adrift), but it should only print the original "G" to screen. This would only require the addition of a couple of new codes indicating empty fields for spectral subclass and luminosity subclass. However such a change, if Chris is happy with idea, would need to wait for the next complete rebuild of stars.dat.

Grant

[Cormoran]

Selden,

The idea of being able to replace stars by STC Catalogue entries is a feature I'd definitely like to see.

Pascal's huge stellar databases are wonderful, but since my interest tends to lie in the near spaces, I can't use them due to a single solitary error (36 Ophiuchi to be precise).

A STC replace feature would open up the database for me to use again, which would be very nice

Keep up the good work, guys

Cormoran

[chaos syndrome logged out]

Selden,

The idea of being able to replace stars by STC Catalogue entries is a feature I'd definitely like to see.

Pascal's huge stellar databases are wonderful, but since my interest tends to lie in the near spaces, I can't use them due to a single solitary error (36 Ophiuchi to be precise).

A STC replace feature would open up the database for me to use again, which would be very nice

Cormoran

I noticed a few dodgy parallaxes in the extended star database so I whipped up a quick VB.Net program which performs a merge operation on the assumption that the Celestia file is more accurate (so any file that appears in both versions of stars.dat is taken from the Celestia file), thus Eta Cassiopeiae isn't over 1,000 light years away any more and I have Theta Orionis back.

[chris]

Celestia needs to at least make an assumption about the luminosity class and spectral subclass in order to come up with a temperature, and from that (plus apparent magnitude and distance), a radius for each star.
That said, my own feeling would be that such assumptions shouldn't be passed on to the user ... if Hipparcos says "G", then Celestia can internally assume "G5V" and come up with a best-guess temperature and radius (which won't be horribly adrift), but it should only print the original "G" to screen. This would only require the addition of a couple of new codes indicating empty fields for spectral subclass and luminosity subclass. However such a change, if Chris is happy with idea, would need to wait for the next complete rebuild of stars.dat.

The format of stars.dat has to change for the next version of Celestia, so it's a good time to extend the information encoded in the spectral class to include an unknown subclass.

--Chris