Page 1 of 1

Galaxies and Star Systems

Posted: 2004-12-22 01:32pm
by Slartibartfast
I'm trying to get as many sources for a random galaxy generator I'm making (just for testing purposes, I already have a PHP one based on GURPS). So I'd be grateful if anyone knows of any books or websites that describe the probability of a binary star, neutrons star, etc, or that give me the proportions of these in the Milky Way galaxy, etc...

Also RPG books or algorithms that deal with such random generation are cool. The only one I have is GURPS Space, and it's a bit, well, weird. I give you the following example:

"As many as one in three star systems may be a multiple star. Roll 2d to determine how many stars there are in the system, adding 2 to the roll if the area is in a cluster or galactic core:
2-9 - Single star.
10 - Double star.
11 - Triple star.
12+ - Four or more stars..."

Makes sense right? No, actually, my spider-sense went "WTF?!". Note that by adding 2, you decrease the chance for a single star and increase the chance for a quadruple+ system, while all the others (double, triple) stay the same!

Anyway, you know what I want. Thanks.

Posted: 2004-12-22 02:01pm
by NecronLord
I don't think our knowledge of the universe is advanced enough to give reliable figures.

Posted: 2004-12-22 02:04pm
by Tasoth
Use a random number between 1 and 100, giving the chance for a single star a large chunk. What little I know of astronomy tells me that the chances for multistar systems having planets dropping with the more stars you have.

Posted: 2004-12-22 03:00pm
by Slartibartfast
NecronLord wrote:I don't think our knowledge of the universe is advanced enough to give reliable figures.
No, but we can sorta model and predict based on the sample we already have.

We also don't really know many planetary systems (with planets) outside the Solar, but that hasn't stopped astrophysicists to make up hypotheses about orbit distribution and such ;)

Posted: 2004-12-22 03:02pm
by Slartibartfast
Tasoth wrote:Use a random number between 1 and 100, giving the chance for a single star a large chunk. What little I know of astronomy tells me that the chances for multistar systems having planets dropping with the more stars you have.
Yes I think that's quite the case. Apparently these chances also increase the higher the density of the cluster. Encyclopedia Britannica says that probably about half the stars in the Milky Way are single, the other half is a mix of binaries and bigger.

I'm thinking, I'm just gonna make a picture of a galaxy with a couple color channels defining density and width, and randomly fill each "pixel" with stars according to some formula :P

Posted: 2004-12-22 04:36pm
by Tasoth
you could do something like so:
1-50 single
51-70 double
71-80 triple

and the rest would be various things, such as quadruple, Pulsar, a star and a blackhole, blackhole. Also, you could maybe make it so that a triple star system would have a smaller star farther out or some such. No clue if thats feasible or not.

Posted: 2004-12-22 05:56pm
by Slartibartfast
Tasoth wrote:you could do something like so:
1-50 single
51-70 double
71-80 triple

and the rest would be various things, such as quadruple, Pulsar, a star and a blackhole, blackhole. Also, you could maybe make it so that a triple star system would have a smaller star farther out or some such. No clue if thats feasible or not.
That's the general idea. Not sure exactly how I'm going to handle binary systems, because they also can be close binaries or not. Planets orbit either star, or they might orbit the center of the two stars.

Here's my idea: this is supposed to be some kind of "on-demand" random generator, which means no actual data is stored except some basic seed, and in this case, the galaxy bitmap.

I'm gonna make a grayscale bitmap with two channels. The main channel will define density, which will shift the probability table. The secondary channel is just thickness, so it doesn't alter the chance at all, merely "stretches" the "height" coordinate.

Each pixel will represent a sector, and a sector is a grid, and a star (or none) is generated for each gridpoint. So in effect, i guess it'd be something like this:

1-70: nothing
71-80: single star
80... and so on

Probably gridpoints will be one parsec apart.

Posted: 2004-12-22 06:05pm
by Deathstalker
This site has has a star map for generating small sectors Space map

If you need planet building charts, I can find my D20 Traveller for some ideas.