Stars and galaxies appear too bright in 1.5.0pre5

[fsgregs]

Sorry about posting here instead of the tracker, but I have several screenshots to show you.

It is great that stars again have halos and flares, but when I go to Earth orbit and look out, many of the stars now appear unnaturally too bright, due to residual halos/flares around them. For example:

Shot 1 - view of Saturn with four bright stars around it. Two of them are Betelgeuse and Rigel, and are hundreds of light years away. Yet when viewing them from Saturn, you can see a small halo around them. They appear unrealistically bright, don't you think?



Shot 2 - is of Rigel Kentaurus A from Earth orbit - It is much brighter than its true apparent magnitude dictates, because it has a noticeable halo/flare.



I am happy to see the flare/halo effect back, but maybe it should be toned down a bit.

The other problem I have is with the galaxies. Here is a screenshot of our sun in front of the Milky Way and Sagittarius.




Frankly, that is NOT what we would ever see if we looked in that direction. Sagittarius is WAY too bright, as is the LMC and SMC and Andromeda. I have auto-mag enabled and galaxy brightness at 0%. This problem has been present since 1.5.0 first came out.

If Celestia is to be hailed as an accurate planetarium program, should not the galaxies look realistic in brightness? What do you think?

Frank

[t00fri]

Frank,

both your overbright displays of stars and galaxies are correlated. I bet you have turned up the "Automag Limit at 45 degrees" ( ] key ) to some value above 7.0.

Note that the ']' key affects BOTH stars and galaxies. If you want to see stars to 9th or even 11th magnitude at FoV of 45 degrees, you will also be correspondingly sensitive to galaxies! They will correctly be brighter than corresponding to naked eye view.

Bye Fridger

[granthutchison]

I find the galaxies generally look good if I choose a magnitude cut-off of 6.0 or 6.5, as Fridger says.

But there does appear to be an exception, in Sagittarius dSph: it's quite intrusively bright, to an extent that seems unrealistic to me.

Fridger, does a galaxy's displayed brightness in Celestia depend on its integrated magnitude, or on its surface brightness? The Sag dSph covers a fairly large area of the sky, so a display based solely on integrated magnitude might account for its apparently anomalous brightness.

Grant

[t00fri]

I find the galaxies generally look good if I choose a magnitude cut-off of 6.0 or 6.5, as Fridger says.

But there does appear to be an exception, in Sagittarius dSph: it's quite intrusively bright, to an extent that seems unrealistic to me.

Fridger, does a galaxy's displayed brightness in Celestia depend on its integrated magnitude, or on its surface brightness? The Sag dSph covers a fairly large area of the sky, so a display based solely on integrated magnitude might account for its apparently anomalous brightness.

Grant

Indeed, Grant,

I am of course aware of the overbright Sag dSph that was only discovered a comparatively little while ago. With normal observational means it is really hard to make out, although it is heading to collide "soon" with our MilkyWay...

The /solid/ data about Sag dSph are correspondingly still scarce and we should surely turn down it's brightness somehow. But by how much relative to the others?

Since a long time I was thinking to implement the 'surface brightness' of galaxies as another independent piece of catalog data. This could well allow to render low surface brightness galaxies like M33 and also Sag dSph much more appropriately. Sag dSph appears of course so big since it is really close by (on galactic distance scales). It's luminosity is smeared out over a big area as you stated.

Bye Fridger

[chris]

But there does appear to be an exception, in Sagittarius dSph: it's quite intrusively bright, to an extent that seems unrealistic to me.

Fridger, does a galaxy's displayed brightness in Celestia depend on its integrated magnitude, or on its surface brightness? The Sag dSph covers a fairly large area of the sky, so a display based solely on integrated magnitude might account for its apparently anomalous brightness.

I was wondering about the anomalous brightness Sagittarius dwarf, too. I notice that even though it's apparent magnitude (seen from Earth) is 3.7, it looks much brighter than the LMC with an apparent magnitude of 0.3. The LMC is a bit larger, but even taking that into account, it's brightness per unit area should still be significantly higher than Sagittarius dSph. Perhaps the rendering of irregular galaxies needs to be adjusted so they appear a bit brighter--more sprites?

Also, is the intrinsic brightness of Sagittarius dSph adjusted to compensate for attenuation by gas and dust along the line of sight? If it is, that could account for the extra brightness when it's seen from Earth.

--Chris

[granthutchison]

It seems to me that M 31 is visually satisfying when the magnitude threshold is set to 6, so perhaps any new "surface brightness" scale could reasonably be weighted so as to maintain M 31's appearance.

However, after Chris's comments, it might turn out that irregulars would require independent weighting in order to maintain the appearance of the Magellanic Clouds.

Grant

[fsgregs]

It seems to me that M 31 is visually satisfying when the magnitude threshold is set to 6, so perhaps any new "surface brightness" scale could reasonably be weighted so as to maintain M 31's appearance.

The problem I see with setting the star cutoff magnitude around 6 or 7 so as to tone down galaxy brightness to realistic appearance is that the number of stars in the night sky drops to quite low numbers. I am not sure what we could visually see with the naked eye if we were above Earth's atmosphere and looking out into space, but a mag limit of 6 removes so many stars that the sky looks ... well ... sparce.

I don't know if a formal study has been ever done to catalog the magnitude cutoff that a person with 20/20 vision would see "from space" without fancy optics, because Astronauts and Cosmonauts look through tinted glass or plastic visors or windows, and cameras used in space almost always employ some level of magnification. I suspect that a magnitude of at least 8 to 8.5 or so would be visible above Earth's atmosphere (I have no proof .. just a hunch). If so, then the galaxies at a mag limit of 8 or 9 are too bright.

Is there a way to simply remove galaxies from being changed when the mag limit is changed? Perhaps a simple cfg file line entry could do it for those of us who want truly realistic galaxy appearance and a bit more stars in the sky. Since we have a separate galaxy brightness control [ the ( ) keys) ], anyone who wanted to boost galaxy brightness could simply use those keys to do so. Alternatively, set galaxy brightness to a reasonable level at a mag cutoff of about 8 or 8.5, and anyone who uses a mag cutoff of 7 could boost the galaxy brightness with the other keys.

Just a thought

Frank

[granthutchison]

The problem I see with setting the star cutoff magnitude around 6 or 7 so as to tone down galaxy brightness to realistic appearance is that the number of stars in the night sky drops to quite low numbers. I am not sure what we could visually see with the naked eye if we were above Earth's atmosphere and looking out into space, but a mag limit of 6 removes so many stars that the sky looks ... well ... sparce.

Well, the number of stars depicted is accurate for the magnitude threshold set. And magnitude 6 is better seeing than most people get most of the time, on Earth. It's the standard threshold for depictions of the night sky, being approximately what a dark-adapted eye would see, far from city lights. So any perceived sparseness is presumably a problem with your expectations, rather than the depiction in Celestia.

I don't know if a formal study has been ever done to catalog the magnitude cutoff that a person with 20/20 vision would see "from space" without fancy optics, because Astronauts and Cosmonauts look through tinted glass or plastic visors or windows, and cameras used in space almost always employ some level of magnification. I suspect that a magnitude of at least 8 to 8.5 or so would be visible above Earth's atmosphere (I have no proof .. just a hunch). If so, then the galaxies at a mag limit of 8 or 9 are too bright.

But if your astronauts are able to see down to magnitude 9, then they will see galaxies three or four magnitudes brighter than we're used to. And they'll see 1st magnitude stars as very bright indeed.

You can't really wind up the magnitude limit to an unusual value of your own choosing and then complain that things look too bright. It's like wearing sunglasses in a road tunnel and complaining that the lights aren't bright enough.

Grant

[t00fri]

Of course I agree completely with Grant's reasoning. It reemphasizes what I wrote earlier above.

It was precisely the basis for my coding of the galaxy brightness display in Celestia. After much effort to present a reasonably concise scheme of display for the /catalogued/ visual (apparent) brightness values of BOTH stars AND galaxies, we should not untangle this connection dictated by optics by a single user demand.

I usually set the automag apparent magnitude limit at 45 degrees to 7.0 (instead of 6.0), which may account to some extent for the lacking atmospheric extinction in Space.

With this setting, the MilkyWay does not seem unnaturally bright compared to such exceptionally good visual conditions.

Throughout the years, I have personally observed the MilkyWay and other galaxies, as well as my personal app magnitude limit for stars in very dark areas and /high/ elevation (Mt Fuji/Japan 3800m , lake Titicaca area/Peru 3800m , Mt. Bromo/East Java 2392m, Annapurna base camp/Nepal 4090m ...)

What still remains to be improved is to specially account for the brightness of a few galaxies that are distinguished by a particularly low surface brightness (LSB).

As I wrote earlier above, It is well known that dwarf galaxies are usually LSB galaxies, possibly since they were recently found to be DarkMatter dominated also in their centers, unlike HSB galaxies! So dwarfs don't have large over-densities of stars in their central regions which supports LSB ...

Here is a corresponding Wiki reference, confirming what I was writing earlier:
http://en.wikipedia.org/wiki/Low_surfac ... ess_galaxy
or also
http://www.obspm.fr/actual/nouvelle/may03/lsb.en.shtml

The surface brightness may be read out of standard galaxy catalogs. So I am still thinking about a /quantitative/ scheme with solid physics support that allows to account for the exceptional LSB, typically of the Dwarf galaxies in the local group that I have recently added.

Bye Fridger