Where is VY Canis Majoris?

[ElChristou]

ElChristou, I quit the community some time ago, and said my goodbyes at that time. As I explained then, I got to a point at which I could make no particularly useful contribution that was of interest to me, and I had other things I could be doing. (I do feel a responsibility to maintain certain data files in the background, which is why I refer to myself as a "maintainer" rather than a "developer" these days.)

The above tweak to star depiction would have been a small development contribution I could have made (a brief guest appearance on the developer list, as it were) but I knew Fridger well enough to guess that such a development would not be welcome. Fridger confirmed that, and so I now drop back into my previous obscurity.
Hence: adieu.

Grant

Tx Grant for the little explanation.
As far as I understand, you and Fridger just have a different point of view, right? Despite I know that sometimes this can turn people mad (hello Cham! ) I'm sure it's not soo dramatic...
...so I will just say "au revoir"!

[ElChristou]

Celestia should say something like "estimated radius", or some kind of symbol ("?", or something else), just to warn the user about the unsafe value.

To really warn the users, it would also be nice to print all uncertainties in a different color...

[t00fri]

Celestia should say something like "estimated radius", or some kind of symbol ("?", or something else), just to warn the user about the unsafe value.

Agreed, but every physicist at least would know what the term

"black-body estimate"

means physically and where the limitations are.

In short, I vote against Chris' suggestion since:

-- such a jigger factor approach is intrinsically untransparent, getting Celestia exactly back into "soft" i.e. undocumentable terrain.

Where are the table entries from etc...

-- black-body estimates are certainly wrong in a number of cases, but they represent the simplest, yet transparent estimates, IF properly announced as such.

Bye Fridger

[granthutchison]

As far as I understand, you and Fridger just have a different point of view, right? Despite I know that sometimes this can turn people mad (hello Cham! ) I'm sure it's not soo dramatic...

No drama, no anger.
I just have no desire to do work which I know will not be well received, since Fridger is generally quite direct with his negative feedback. As a user of Celestia, my sympathies are with Cham on this; but given the choice between doing work which is welcome and work which is not, I trust you'll understand which choice I'll make.

As to "data purity", I use the phrase to indicate Fridger's well-known desire to use single, validated datasets which are processed without hand-editing; it seems a pretty accurate phrase, since the data are unmodified and unmixed. I also felt that it was quite unobjectionable, until Cham chose to make the link with "racial purity": a connection that was of course never intended and which I find offensive.

Grant

[hank]

It's certainly better than reporting grossly incorrect radii.

I certainly agree with that.

I always found it misleading that Celestia prints star radii out, based on a very primitive estimate, as if it was reading out a sophisticated data base!

Celestia should say something like "estimated radius", or some kind of symbol ("?", or something else), just to warn the user about the unsafe value.

To really warn the users, it would also be nice to print all uncertainties in a different color...

Isn't the estimated radius also used for sizing the model that's displayed for the star?

- Hank

[ElChristou]

...No drama, no anger. ...

Happy to read that!
So really, no more "Adieu" and only a few "Au revoir"...

[t00fri]

It's certainly better than reporting grossly incorrect radii.

I certainly agree with that.

- Hank

That's precisely not an attitude that is defendable from any scientific point of view. Wrong or inaccurate values can be both valuable and physically instructive.

An incorrect radius obtained from a black-body estimate teaches us a LOT, since the black-body picture is not just a formula. It is a transparent physical concept!

A radius value from a "semi-empirical jigger factor" approach tells us NOTHING, least whether the value is RELIABLE. Fine, if it happens to be more accurate than the black-body value, but what if it is ALSO WRONG?

What do you guys say in such cases?

Too bad ...I suppose... What else could you honestly say?

Bye Fridger

[hank]

I just have no desire to do work which I know will not be well received, since Fridger is generally quite direct with his negative feedback. As a user of Celestia, my sympathies are with Cham on this; but given the choice between doing work which is welcome and work which is not, I trust you'll understand which choice I'll make. :)

Grant,

Of course nobody wants you wasting your time and talent doing work for Celestia that wouldn't be used. But as far as I know, while his opinions naturally receive great respect, Fridger doesn't have a veto over what gets used in Celestia.

Also, perhaps it would be possible to include a configuration switch that would allow the user to choose between using an unmodified blackbody estimate and using a more flexible alternative that includes other factors.

- Hank

[ElChristou]

...Too bad ...I suppose... What else could you honestly say?

Beuhh.... ...Razvouhniac ...?



(please don't google that!)

[t00fri]

I just have no desire to do work which I know will not be well received, since Fridger is generally quite direct with his negative feedback. As a user of Celestia, my sympathies are with Cham on this; but given the choice between doing work which is welcome and work which is not, I trust you'll understand which choice I'll make.

Grant,

Of course nobody wants you wasting your time and talent doing work for Celestia that wouldn't be used. But as far as I know, while his opinions naturally receive great respect, Fridger doesn't have a veto over what gets used in Celestia.

Hank,

I never talked of a veto, but I put in my vote. And I justified my vote from the experienced perspective of a senior professional scientist.

Long before you joined, notably Chris and the Celestia devs (to a considerable extent also Grant, if I correctly remember) agreed on the general strategy to base Celestia's data exclusively on documented published scientific data.

That was the agreed basis on which I spent lots of my time over the years to help getting Celestia in this desirable stage of it's database. At that time we also agreed on LOK masks for incomplete body textures for example. At that time I also introduced human readable PERL scripts as a concise documentation in the sources of all deviations from published data (binaries, galaxies,...). Grant also subscribed to this attitude by quoting the publications he used for his mods of our HIP star database in revised.stc. He and others have also adopted meanwhile quite strict and useful citations of orbital paramter sources in solarsys.ssc and other data files.

And now you come along, suggesting that I don't have a veto and rather we should return to undocumented "jigger factors"...

Bye Fridger

[ajtribick]

Also brings up the question of what exactly do we mean by the radius of a supergiant? Quick estimate using a few representative values I poached off Wikipedia (not my favourite source of information by a long way, but I'm just doing a few arm-waving not-particularly-accurate calculations here) gives that the scale height kT/mg (where m is the proton mass, not the mass of the supergiant!) for a supergiant can be something of the order of 10^6-10^7 km. Add to that the dense stellar winds, pulsations and the like and the radius isn't looking too well defined.

[hank]

I never talked of a veto, but I put in my vote. And I justified my vote from the experienced perspective of a senior professional scientist.
...
And now you come along, suggesting that I don't have a veto and rather we should return to undocumented "jigger factors"...

Actually, I believe it was Grant who suggested using a table of adjustment factors to improve Celestia's stellar radius estimates, and Chris also endorsed the idea. Presumably the values used would have a reasonable scientific basis, which would be clearly documented. I think many users would find that approach preferable to the simple black-body calculation currently employed.

- Hank

[t00fri]

Presumably the values used would have a reasonable scientific basis, which would be clearly documented.

Unlike you, Grant correctly stated:

I doubt if there is a scientific approach that could generate accurate radii for red giants and supergiants without using prohibitively complex calculations.

So, unlike Grant, you feel that this is "presumably" no problem to do a scientific approach. OK, let me know what you know about this issue, once you write "presumably"...

I think many users would find that approach preferable to the simple black-body calculation currently employed.

- Hank

Many users like many things as we all know. Few have a solid physics training, while this matter IS about physics. Anyway, let me know your hard arguments, as to the possibility of scientific documentation and seriousness of such a "jigger factor" approach. Let's see on what basis YOU are placing your own arguments?

Yet let me try to be as constructive as possible without violating our joint longstanding agreement. The only acceptable alternative that I can see might be a double printout on the canvas, e.g. like so:

---------------------------------------------------------------

Code: Select all

Radius:   xxxxxx km     (black-body estimate)
          yyyyyy km     (semi-empirical)


---------------------------------------------------------------
where "semi-empirical" might be whatever seems feasible with a potential to improve the accuracy of the radius predictions. By clearly stating the origin of the two numbers, users can notably learn from their differences whether there is a certain amount of reliability, perhaps, and are notified about their (shaky and quite different theoretical physics basis) at the same time!

Bye Fridger

[selden]

And what about variable stars with changing radii?
According to Simbad, VY CMa is a pulsating variable. Celestia currently has no way to indicate its varying radius or luminosity.

Heck, it isn't even a single star.
See http://www.aavso.org/vstar/vsots/0399.shtml

I'm amazed that it was included in the HIC.

[t00fri]

And what about variable stars with changing radii?
According to Simbad, VY CMa is a pulsating variable. Celestia currently has no way to indicate its varying radius or luminosity.

Heck, it isn't even a single star.
See http://www.aavso.org/vstar/vsots/0399.shtml

I'm amazed that it was included in the HIC.

Selden,

don't worry, our "physics expert" Hank knows the answer

Presumably the values used would have a reasonable scientific basis, which would be clearly documented.

Bye Fridger

[hank]

Presumably the values used would have a reasonable scientific basis, which would be clearly documented.

Unlike you, Grant correctly stated:

I doubt if there is a scientific approach that could generate accurate radii for red giants and supergiants without using prohibitively complex calculations.

So, unlike Grant, you feel that this is "presumably" no problem to do a scientific approach.

My use of the term "reasonable" was intended to indicate a recognition that some compromises might be necessary to avoid the prohibitively complex calculations needed for a precisely accurate scientific determination, as noted by Grant.

The only acceptable alternative that I can see might be a double printout on the canvas, ...

The issue is not simply what radius value to use in the information display, but what radius value to use in rendering the object. For the information display, I don't really see a great value in including the black-body estimate in cases where it doesn't apply, but I wouldn't particularly object if you think it's important.

And what about variable stars with changing radii?
According to Simbad, VY CMa is a pulsating variable. Celestia currently has no way to indicate its varying radius or luminosity.

Variable stars in general is an area where Celestia might be improved.

don't worry, our "physics expert" Hank knows the answer ;-)

Fridger, your incessant personal attacks are of no value to the discussion.

- Hank

[Cham]

Code: Select all

Radius:   xxxxxx km     (black-body estimate)
          yyyyyy km     (semi-empirical)


I vote for this approach. Both informations could also be used as an indication of "uncertainty".

[t00fri]

Code: Select all

Radius:   xxxxxx km     (black-body estimate)
          yyyyyy km     (semi-empirical)


I vote for this approach. Both informations could also be used as an indication of "uncertainty".

Cham,

that's of course the correct attitude of a physicist .

I have meanwhile also communicated per email with Grant. He is also quite happy with this proposal.

Of course, when quoting 2 values for the radii, one still has to select ONE for actually rendering the star. But here I would have less worries about "data purity". It's mainly the numerical database that should be reasonably clean and reproducable. Since any acceptable scheme with semi-empirical correction factors could be expected to be more accurate on average than black-body radiation , the semi-empirical choice of radii for /rendering/ the stars seems OK to me.

Moreover, Grant, would favour in addition a general flagging scheme with ?, ??, ??? to quantify the uncertainties of data or calculated values. This is very fine from my point of view.

Bye Fridger

[Cham]

Ok,

and what about you, Chris ?

[chris]

Ok,

and what about you, Chris ?

I'm OK with printing two values for the radius. The one used for actually drawing the star should be the value computed with the jigger factor. I also advocate only showing a single value (labeled just 'Radius') when then radius of the star is specified in an .stc file, as it is for the Sun. There are 'knowledge flags' in the Star class that could be used for this.

--Chris