Milky Way map for the hardcore sci-fi strategists (Full Version)

All Forums >> [New Releases from Matrix Games] >> Distant Worlds 1 Series


Miletkir -> Milky Way map for the hardcore sci-fi strategists (1/13/2021 10:45:49 PM)

My thrill is to play 4x space games in a Milky Way setting, preferably accurate. This makes for a strong story background, future of humanity and what-nots. In reality, I've been mostly happy with 4x titles as long as they gave me galaxies with 4 spiral arms: a bit of renaming keeps me satisfied after that. It was nice to actually start with a Sol system in Stellaris, for instance, but the vast superiority of DWU in other areas lets you get over the fact that your home system is random with a human faction. Still, I'm secretly hoping we can have an option to start with Sol in DW2 - we might come back to that later.

With the new kind of scale that DW2 is going to offer, generating star systems in the thousands, I'm tempted to knock myself out. So I went back to Winchell's star databases. I played around with his HabHYG database and his Stereostar a bit - the Python program and the star database are aging, but still powerful.

I first generated a series of rather extensive starmaps up to 1000 parsecs around Sol (beyond that, there are tricky display and scale issues to overcome). Because of RAM and cluttering limitations, I had to generate these maps by parsecs increments: increments of 10 parsecs from 0 to 120, then 120-140, 140-160, 160-190, 190-250, 250-350 and 350-1000. Stars whose spectral class was uncertain were not displayed. The center is always Sol, and the X-Y axis is the galactic plane. Higher Y values point to the galactic core. And, even when culling stars inside a given radius, the area is not empty, because there are always distant stars higher or lower on the Z axis, making up the "thickness" of the Orion arm.

Of course, this amount of data is overwhelming, and has limited usability on the side of the player. I'm wondering if the coordinate data could be used to help recreate Milky Way-like galaxies in DW2 for a specific scenario, although a few thousand parsecs worth of data does not weigh much against a 100,000 light years wide galaxy. Data from Gaia should be more useful in that regard, and I've started playing with the Aladin software to see if we can extract something.

Anyway, from either the player or the developer point of view, it is more interesting to focus on named systems, which refer to constellations or astronomical history. Bodies from star catalogues (e.g. HIP) are important in showing the star pattern, the diversity of spectral classes, and a lot of them are even potentially habitable, but in the context of a galaxy that's procedurally generated anyway, it's just noise.

So one way to have a "cleaner" approach is to remove all stars with catalogue numbers. Now we have recognizable names associated with unique stars or constellations. Here are two maps, the first with all named stars in the database, the two others with named stars within a 1000 parsec radius (one was rescaled, the other not):

But there are still a lot of constellation stars differentiated with just numbers, and, as you can see, the maps can still easily become unreadable. However, even if the central area around Sol is a dense cluster, the map is still useful to find the name of systems a bit further away from the immediate vicinity of Sol, which is relevant to the game, since the nearest stars will be a minimum of hundreds of parsecs away, even with a 10,000 star galaxy.

Since we are looking for the big picture here and are not making an exact simulation, we can further remove some less interesting stars. So I took out all of the constellations stars starting with numbers, to leave only the single names and constellation stars with Greek designations. At a much smaller scale, this is the view that a game like Descent: FreeSpace chose to adopt. I called those "main named stars" which is of course entirely subjective. There is a map with all of those sub-selected stars, with a center still quite cluttered, a map with main named stars within a 50 parsecs radius around Sol, and one with these stars from a 50 parsecs distance and beyond. If a map looks like just a little splurt of white in the middle of the image, it's because there are very distant stars spreading the axes out!

We can push that even further and keep only the "landmarks", all those celebrity stars with proper, unique names, like Altair, Deneb, Aldebaran or Betelgeuse. There's a map for under the 100 parsecs mark, and another that is complete.

You can see that most of these famed star systems are quite close to Sol, Deneb being a notable exception (but still "only" about 1000 parsecs from Earth). At galaxy scale, this source of names might be quickly depleted if one respects scale. But again, this will depend on the generative capabilities of DW2. Even if it's only useful early in game, it's my kind of blast to check where I'm expanding and what are the real systems to most likely correspond to the one I see ingame.

In the context of DW2, perhaps what could be more interesting to some would be the surrounding systems that would actually be habitable. The beauty of Winchell's HabHYG is that it contains these data. As I mentioned above, collecting all the catalogue stars might not add much for our scope. So here's a map of all named potentially habitable star systems I found:

You can see that the "celebrity" star systems are not necessarily the most promising. It's mostly G5 and F5 classes.

Now, I'll be the first to mumble that those maps are unappealing visually. This was mostly a phase of data exploration, and I'll see if I can improve things visual-wise. The catch is that increasing font size often leads to more overlap of names. A more pleasing visual would also require to sacrifice relative distances. Winchell also used other Python software to generate a more game-friendly display:

But I haven't been able to reproduce it yet. My next goal will be to try and put together something more easy to navigate and use.

As I said, the Milky Way extends much beyond the data of the HabHYG, and, for an accurate experience with a human civilization, additional documentation such as the galactic density map gathered by Gaia will be useful. It's not the cup of tea of everyone to bother with such details, but I think the intricacy and depth of Distant Worlds is the right playground for those who do. If DW isn't a good game for that, no game is.

Osito -> RE: Milky Way map for the hardcore sci-fi strategists (1/14/2021 10:55:30 AM)

I have found that the way that works berst for me is to have a non-linear scale for stellar distance. In general, I'm most interested in seeing star names that I've heard of, rather than catalogue numbers. By having a non-linear scale, you can include dozens, or even hundreds, of well known stars in the nearer parts of the galaxy, with 'unknown' stars further away. I've found that was a reasonable mix, at least for me, and I think it worked well in the maps I did for DW1, and particularly the New Horizons mod for Stellaris.

However, there is definitely a place for the more rigorous type of mapping that you're suggesting, and I look forward to seeing what you come up with for DW2. I'll certainly be applying my own approach to DW2, so hopefully we'll be spoilt for choice.

This all depends on how they implement the modding for DW2. With DW1, a star map mod counted as a 'theme' in itself, and it wasn't possible to mix it with other mods, unless you built it that way from the ground up. I'm hoping DW2 will be a lot more flexible in that regard.

Miletkir -> RE: Milky Way map for the hardcore sci-fi strategists (1/14/2021 2:54:59 PM)

Your maps and scenarios are a beauty. Obviously there's a lot of fine handcrafting, and humour, but you didn't manually place all these systems? How did you do it? And which database did you use?

To be honest, I wasn't going to strictly hold onto the idea of linear distances. Even if DW2 can generate a colossal amount of stars, there's always going to be a need for compromise between reality and fiction (largely tilting towards fiction). In Descent Freespace, even if all the systems depicted were rather close to Earth, the issue was mitigated by the fact that intersystem travel had to go through jump points or gates. Thus the actual distance between systems can be anything, and this also circumvents other issues of physical travel across the galaxy, such as dark matter density. If DW2 stays with the high-speed travel concept, my mind wants to keep at least a pseudo-linear representation.

An important point for me would be first to get the spectral classes about right, as well as some notion of relative distance to Sol, and perhaps even cluster patterns. I don't see having a large part of the galaxy beyond our local area made up of fictional names being a problem, since the discovery of an unexplored galaxy is a big part of the fun.

The way I see it, we're both surfing on the same superstring, and I say we're better off combining forces to make a semi-realistic Milky Way map. A reference map to make manual modifications ingame was my first modest goal, but if you're interested in this, modding seems all the less out of reach. Let's first see what DW2 holds in store!

Page: [1]

Valid CSS!

Forum Software © ASPPlayground.NET Advanced Edition 2.4.5 ANSI