Celestia Forums

I'm making a fictional solar system based around a (fictional as well) star named Magellan's Sun, in the LMC. I looked on the Motherlode and I have not found any addons with stars for the LMC. Can somebody post a link or something? Also, inb4 Globular Cluster Generator.

Try this.

Thanks.

OK, I tried it and it absolutely will not work. I've tried everything, from replacing every "dead" star with a neutron star (as AFAIK there's no "dead" spectral class in Celestia) to removing all the #'s that somehow appeared in front of every line but the first. Apparently the problem is that there are no empty lines between the star definitions, and neither Notepad nor WordPad allow me to replace every } with } followed by a blank line.
EDIT: Never mind, it works. But the stars still refuse to show up from Magellan's Sun. I'll try moving it.
EDIT2: Moving the star worked. But now I have to wear nuke-proof sunglasses to see them. It's like every star in the galaxy is exploding at once. Looking closer at the STC file I see that the stars' absolute magnitude has E+04 stuck on the end. How do I get these stars' brightness to realistic levels? Removing the E+04 made half of the stars invisible. and turned the rest into dim candles.
EDIT3: I don't care what the stars look like now. After I added a nebula the light from Magellan's Sun turned red.
EDIT4: Turns out that red light wasn't from Magellan's Sun at all, and wasn't caused by the nebula. It was from some of the other stars in the cluster, which I had replaced with black holes. By now I'm seriously considering using a nebula using a starfield texture for stars.

W0RLDBUILDER wrote:OK, I tried it and it absolutely will not work. I've tried everything, from replacing every "dead" star with a neutron star (as AFAIK there's no "dead" spectral class in Celestia)

You shouldn't have to "do" anything. Sorry, I should have included some instructions:

Put the file cluster.stc in your "extras" directory,
start Celestia, and
GoTo clu000001 (for example).

The stars are named clu000001 through clu010000. The cluster is centered at Celestia's position for the LMC , which is at
RA 5.3928 ( = 80.892 degrees)
Dec -69.7561
Distance 1.631e+05

The cluster generator that I used ( MCluster) "discards" about 25% of its intermediate results by flagging them with an invisibly dim luminosity value. I left them in but commented out and with the SpectralType set to "dead". You won't see them when using the original file above.
( MCluster doesn't generate spectral types, although it does include luminosities. I had to write a translation program which guesses at the SpectralType from the physical characteristics that MCluster generates. MCluster normally is used to create input files for gravitational studies.)

to removing all the #'s that somehow appeared in front of every line but the first.

They're in front of about 25% of the stars.

Apparently the problem is that there are no empty lines between the star definitions, and neither Notepad nor WordPad allow me to replace every } with } followed by a blank line.

The file has Unix-style line terminators (bare line-feeds). Wordpad displays it properly. Saving it from Wordpad produces a file which Notepad displays properly. Celestia doesn't care. It recognizes all three styles of line termination: Windows <cr><lf>, Unix (Linux) <lf> and MacOS <cr>.

EDIT: Never mind, it works. But the stars still refuse to show up from Magellan's Sun. I'll try moving it.

EDIT2: Moving the star worked. But now I have to wear nuke-proof sunglasses to see them. It's like every star in the galaxy is exploding at once.

I suspect you are using Star Style Scaled Discs. Fuzzy Points looks better in this situation: when looking from nearby at many stars which are very close to one another.

Looking closer at the STC file I see that the stars' absolute magnitude has E+04 stuck on the end. How do I get these stars' brightness to realistic levels? Removing the E+04 made half of the stars invisible. and turned the rest into dim candles.

Larger positive AbsMag values produce dimmer stars, not brighter ones. (An AbsMag of 9999.9 effectively turns a star off.)

The stars look bright because they're close together. The viewpoint isn't very far from them when the cluster fills the screen. A couple of screengrabs on my system are below, using Star Style Fuzzy Points. There are about 7500 stars within about 15LY, with a typical distance between adjacent stars of about 1LY. They're even closer together near the center of the cluster. That makes it easy for interstellar travel, unlike the relatively low star density where we are.

EDIT3: I don't care what the stars look like now. After I added a nebula the light from Magellan's Sun turned red.
EDIT4: Turns out that red light wasn't from Magellan's Sun at all, and wasn't caused by the nebula. It was from some of the other stars in the cluster, which I had replaced with black holes.

If you use the original file above, you shouldn't see that effect. Below are a couple of screengrabs taken on my system.

By now I'm seriously considering using a nebula using a starfield texture for stars.

If you want to create pictures with a dramatic backdrop, that's certainly an effective way to produce them. Some of the pictures of the Tarantula Nebula are amazing.

I think I've managed to duplicate the "screen filled with red" problem, or at least something similar to it.

As best I can tell, it's a bug in Celestia. It looks to me like it sometimes draws stars that actually are immediately behind the point of view. I'm guessing it's related to the stars being so far away from the center of the coordinate system (the solar system barycenter) so that the clipping planes aren't being calculated accurately. Drawing stars more than 16K LY from the Sun is a relatively new feature. I doubt it's been used very often with stars so tightly packed.

With the cluster stc provided above, I see the picture shown below when at the viewpoint produced by this URL:

cel://Follow/clu000001/2011-05-29T19:35 ... rc=0&ver=3

The screenshot you posted is nothing like what my problem is. Mine's not so much a "red sky" effect, as it is a "red light" effect. The planet Hurricane in my star's solar system located at exactly the same point as clu000001 (but with clu00001 deleted so there's no conflict) looks like its sun has no influence on it at all. I rotate around it to see only the planet's dark side and it's not lit by the sun at all. It instead looks like it's being lit by an M-star in a completely different direction. I'm fine with the red light on Firestorm (a small volcanic planet with a dense atmosphere) but on Hurricane, a blue gas giant? I'll try using your original cluster file (again with clu000001 removed as my system's sun is at the same place as clu000001) and see if it works. Maybe I was doing something wrong. I might also try adding a nebula, but seeing as I know next to nothing about the Tarantula Nebula it'll be a fictional one.
EDIT: Still having problems with the light on Hurricane and Firestorm. The sun has absolutely no effect. Celestia renders the planets as if the sun didn't exist and the planets orbited other stars. I think I'll just use a fake starfield to avoid these problems.

W0RLDBUILDER wrote:EDIT: Still having problems with the light on Hurricane and Firestorm. The sun has absolutely no effect. Celestia renders the planets as if the sun didn't exist and the planets orbited other stars. I think I'll just use a fake starfield to avoid these problems.

It sounds to me like those other stars outshine the sun. Check the apparent magnitudes of the sun and those other stars and see how they compare. If those distant stars far outshine the sun, Celestia will render the effect from the sun in a muted way.

I think Hungry is right:
1. Celestia supports multiple light sources.
2. Celestia does not implement 1/r^2 falloff in illumination
3. Several cluster stars are well within the 1LY limit that Celestia seems to use for calculating planetary system effects.

My candidate for the star providing the brightest illumination is clu004609. It's at a distance of about 0.9 LY and has a luminosity of about 78x our Sun.

In other words, you might consider replacing a star that's more distant from the center of the cluster, where the stars are farther apart.

Alternatively, I've generated a larger cluster. See below.

CLU2 is displaced slightly from the first one, although it overlaps just a little. See the picture below for a comparison, with them both in the field of view at the same time. The first cluster is the tight, bright one. Of course, you don't have to have both clusters installed. The star CLU2_000001 seems to be maybe halfway in, and its nearest star (CLU2_000984) is about 1.17 LY away.

Thanks, that second cluster really helped! I replaced clu2_00001 with my star and put the nebula back. I've attached two screenshots of Hurricane and Firestorm. Both planets are lit properly now.

Not going to lie... seeing those clouds (shadows?) in front of the first planet's ring system is kinda awkward.

It's not like I can fix that. AFAIK it's a bug in Celestia.

It's been a while since I last worked on this addon. I have all the textures done and I'm working on coding the rest of the system (everything other than Firestorm and Hurricane). I've removed Phoenix (because it's next to impossible to make the planet fall apart near the sun) and Bluestorm (because it's a near-clone of Hurricane and two objects with storm in their names are enough). I've named all three large moons of Hurricane (New Earth, Thundercloud, and Sandstorm). What I would like is a way to generate 10-40 asteroids in orbit around Hurricane (because I'm too lazy to code them all myself )(do you( think I (overuse) par(en(t)h)es)es? i think i do). The planet's mass is about 100x that of Earth.

A spreadsheet-based asteroid generator is available on the MotherLode's Utilities page:
http://www.celestiamotherlode.net/catal ... ities.html

I already knew about that. Is there any way to make the asteroids orbit Hurricane at somewhat realistic speeds?