I was looking for a space exploration game, and couldn't find one I liked, so I decided to make my own ;) Anyway, while my interest holds, I'll try to work on it. I'm having difficulty with forming my galaxy. (I'd like to form it randomly for each game, so the game has a replay value) Basically, I'd like to be more or less realistic, and form stars with spectral and luminosity classifications (ie: our star is G2V, etc.) Master of Orion 3 seems to use this (I don't know if it's just tacked on or not, but I suspect not.) I don't know where to find data on the relative frequencies of the different star types. I'm willing to fudge it if it's necessary to make the game a game, but I'd like to be as realistic as possible. I know the majority of stars fall along the "main sequence" curve on the H-R diagram (<- that's what 3 hours of net research yielded ;P) and I know that the other types, (ie: supergiants, whitedwarfs, etc.) are less numerous, and represent later stages of stellar evolution, but I want to know relative freuencies in numbers, so I could code it in. Ideally, I'd like them as functions of galactic age, so that the player could set an older galaxy, and have more white dwarfs, or whatever the case may be). Like 80% main sequence, 2% white dwarfs, you get the idea. I can't find it anywhere. I'd be willing to base the numbers on the data of stars close to Sol (our sun), which I know exists, even if I can't find it. Also, I need some rubrick for deciding planets around these stars. This is admittedly far more difficult, as real science has only discovered a handful of interstellar planets, so I'd be more than willing to fudge this part. But I suspect that there are theoretical models which would suggest, say, numerous planets around K to M type stars... etc. and perhaps fewer around O type, with O type planets being more gaseous, you get the idea. If anyone with an astronomy degree would like to help, I'd be sure greatful ;) Really though, if anyone knows where any kind of information like this can be found, I'm all ears ;) Certainly, Master of Orion 3 seems to have occomplished this reasonably well, so if anyone knows the rubrick they used for their galactic creation, that'd be good too. Thnx

Here are a couple of sites I'm using for my current project:

A galactic model - good for layout

Uni of Illinois has alot of good info

BBC has a great site

Atlas of the universe has a ton of info

This guy has put an AMAZING amount of work into solar system generation using RPG rules It's too bad he didn't finish.

Good luck!

[Edited by - Wavinator on July 23, 2004 2:30:38 PM]

I'm not sure how much I can add that you haven't researched already, or isn't covered by one of Wav's links. One thing that might be worth bringing to your attention is the different generations of star, referred to as Population I (youngest), Population II (old) and the hypothetical Population III (oldest, unknown whether they even exist)
link

You might also find this useful:

Quote:

1. The most luminous and apparently hottest stars are found within gaseous clouds containing much cosmic dust. These stars are confined in clumps to a thin plate that forms the equator of the Galaxy. Similar stars define the highly visible spiral arms seen in other galaxies. 2. Bright, cooler stars like Sirius are located near the equator of the Galaxy but are not confined to the galactic arms. 3. The disc of the Galaxy is populated with moderately hot stars (with 5000 to 8000 K surface temperatures); these stars resemble the Sun and populate the arms, the spaces between the arms, and make up part of the stars that occupy the central core of the Galaxy. These disc stars are the most numerous group of stars observed. 4. The disc of the Galaxy is enveloped in an ovoid shell of red giant stars whose spectra show fewer metals than stars of comparable type in the disc population. That these stars are mostly giant stars is usually explained by claiming that the smaller stars in the population are not likely seen because of distance from the Earth. It is possible that the latter are absent. Most of what is known about these stars is from the study of giant stars within star clusters and intrinsically varying giant stars, where the star's luminosity varies in some characteristic way over an interval of days to months. 5. The Galaxy itself is embedded in a halo of cooler stars. Most of what is known of the galactic halo is deduced from a study of a few nearby small stars and 120 globular star clusters which surround the core of the Galaxy. One of these globular clusters, Messier 13 in the constellation of Hercules, has been described as a "celestial chrysanthemum" (Baker, R. H., p451). The number of stars in this cluster cannot be counted; but estimates around 500 000 are made. Averaging this number of stars over the volume of the cluster (not precisely known) it would seem as if the stars are about two light-years apart, much closer than the stars near the Sun. Some small halo stars are observed passing through the disc stars in the Sun's vicinity. Barnard's star is an example.

link

I can't find any online references to the exact percentages of different star types online. It does however, depend on whereabouts in the galaxy you're looking, as the previous link says. I might have some more information in one of my textbooks, I'll see if I can dig anything up when I get home.

EDIT: You might also find this interesting.

[Edited by - Sandman on July 22, 2004 5:19:45 AM]

Hey, thanx for the information. Wav, the link you provided about the solar system generation is broken, so if you could repost that, that'd be great.

However, despite the fact that I now know more about stars and spectral types and luminosity than I ever wanted to know ( :P ) , I'm still unsure how to go about creating a mathematical method for creating stars.

I'm thinking of pretending that all spectral types are equally likely ( a small stretch, but perhaps not too bad), and then deciding their luminosity and type (ie: dwarf, giant, etc.) as best as I can figure from the H-R diagrams of the nearby stars.

The luminosity would determine from how far away you can detect the star. I'm toying with the idea of having large dust clouds (ie: numerous nebulae) to obstruct whole parts of the galaxy from the player, so that they'd have to navigate around them until they found what lies on the other side, at which point they could shortcut through the nebulae.

However, I'm having the same type of problem with nebulae's that I was with stars and planets. I'm starting to think that all the games that create galaxies probably fudge it a bit :P

I haven't even started on the planetary formation of the different solar systems. If anyone knows a link that describes what scientists think are likely to form, (both in number and type) around different kinds of stars, I'd appreciate it a great deal.

Lastly, just so you understand kind of what I'm after, I'm trying to create a game that'd be a sort of cross between Noctis and 7 Cities of Gold. (For those of you who don't know what these are, you can find links about them and download links at http://www.the-underdogs.org , search for them by title).

Noctis fundges its galaxy a great deal, yet it is still quite remarkable. Evidence that as long as you fudge well, it doesn't matter to what degree you do it to.

As just a fast side note, I was wondering what coordinate system would be best for a galaxy. I'm thinking that spherical coords are probably the best, but since the galaxy is roughly flat out where we are, maybe cylindrical are better? Or I could just stick with good ol' rectangular coords. I don't know, and I wouldn't want to go halfway through the development only to find out that I should have stuck with rectangular or cylindrical.

Thanx for your time.

Quote:

Original post by Numsgil Hey, thanx for the information. Wav, the link you provided about the solar system generation is broken, so if you could repost that, that'd be great.

D'oh! Done!

Quote:

However, despite the fact that I now know more about stars and spectral types and luminosity than I ever wanted to know ( :P ) , I'm still unsure how to go about creating a mathematical method for creating stars. I'm thinking of pretending that all spectral types are equally likely ( a small stretch, but perhaps not too bad), and then deciding their luminosity and type (ie: dwarf, giant, etc.) as best as I can figure from the H-R diagrams of the nearby stars.

In one of those links I listed there's a breakdown of the percentages if you're interested. Could you turn that into a simple random number generation scheme?

Quote:

However, I'm having the same type of problem with nebulae's that I was with stars and planets. I'm starting to think that all the games that create galaxies probably fudge it a bit :P

You could feed data directly by checking out the start information and converting right ascension and I forget what the other variable is. This would give you exact positioning, luminosity, etc. But it's going to be alot of work.

Quote:

As just a fast side note, I was wondering what coordinate system would be best for a galaxy. I'm thinking that spherical coords are probably the best, but since the galaxy is roughly flat out where we are, maybe cylindrical are better? Or I could just stick with good ol' rectangular coords. I don't know, and I wouldn't want to go halfway through the development only to find out that I should have stuck with rectangular or cylindrical.

I'm going with rectangular, simply because it's easiest and you won't get alot of weird crowding as you get closer to the center like you might if you used some sort of cylindrical scheme.

Wow, you weren't kidding when you said he put an AMAZING amount of work into the planetary model. AMAZING is an understatement. Using his model as a basis, I bet I can chop off two weeks of mental anguish and research. Thnx alot.

You mentioned that in one of the sites you listed, there's a breakdown of the percentages. Try as I might, I couldn't find it amidst all the wealth of information :P If you could list it specifically, I'd be really greatful.

Just so I can list a source for any future readers of this thread, I'll list this site:

http://www.codemoon.com/cm/archive.php?index=main&year=2001&month=2

which provides a rather well developed method for creating the spiral arm shape and distribution of the galaxy.

Wav, you've been a great help, thanx alot. ;)

I must agree, working on a space game (with a procedurally based universe) myself, i found these links very useful.

Another source of data (which i've been primarily concentrating on) is the Hipparcos sattelite database. It lists the 100 000 brightess/closest stars to our solar systems, but most of the stars are in a radius of < 1000 LY. I'm going to use it as the primary input data near our solar system and will procedurally generate the star systems that are more than 1000 LY, on a statistics model based on Hipparcos data.

Getting the database into a "reable" format for the amateur like me (and probably you) is already a nice little project in itself but once it's done, there is a lot you can do with each star data. It would then be pretty straightforward to analyze what is the percentage of stars in each class and replicate it.

Y.

Quote:

Original post by Numsgil Wow, you weren't kidding when you said he put an AMAZING amount of work into the planetary model. AMAZING is an understatement. Using his model as a basis, I bet I can chop off two weeks of mental anguish and research. Thnx alot.

No problem, I was pretty psyched to find it myself. I'm only sorry he stopped at the climatology stuff, I would have loved to have seen what he would have done with alien races and atmospheres.

Quote:

You mentioned that in one of the sites you listed, there's a breakdown of the percentages. Try as I might, I couldn't find it amidst all the wealth of information :P If you could list it specifically, I'd be really greatful.

Hah, actually that was no fair given the amount of info overload, but I was running late this morn. Here ya go!

You'll see that the vast majority of your stars, if you want to be kosher, are dim red M class stars, and human friendly G class stars are only 4%. You may want to fudge this or 80% of your galaxy will end up a bit dim and maybe dull.

Quote:

Wav, you've been a great help, thanx alot. ;)

I'm always whining about how few sci-fi games there are out there (FPS's aside), so I'm glad to help a fellow fan! [grin]

Wavinator, is there any chance we can get some info on your space project? The game should be very interesting (based on your threads). Perhaps you have a website...?

I've gathered nearly all the data that I need to create a sample galaxy. I just need one more thing:

in order to form the galaxy, I need to create a model that will determine the distance from the center of the galaxy (which I have) and the star's distance from the mean line that can be thought of as a plane extending from the center of the galaxy, cutting it into two halves. (That is, into symetrical top and bottom halves).

Ihave that the galaxy is 2000 light years thick where the sun is, and ~3000 light years thick at the bulge. But I'll need more data points to form a statistical correlation which follows the shape of the galaxy.

Once I have that, I can create a test program that creates the stars and their location, spectral class, and luminosity.