First galaxy test

Just been playing around with a script to generate a galaxy, which so far looks like this:

It’s fully 3D, and just consists of layered meshes consisting of several thousand vertexes with appropriate media. I’m contemplating a completely unrealistic plan of recreating the feel of one of those Hubble deep space images ( and then doing a fly through of it.

Finally here’s a fly around of the thing:

EDIT: Here’s a youtube version, but it looks pretty ordinary since most of the fine detail is lost, and because there are some serious artifacts added in by youtube’s encoder

Home sweet home :^)

Cool stuff manbites dog!

You should go to the Galaxy Map website
and introduce yourself to the webmaster. He has been looking for a good Blender plug-in to do nebulae for his map. Perhaps the two of you can work something out.

Thanks for the link, Nyrath.

I followed it back here. I’m the creator of the Galaxy Map website ( and I am indeed looking for an artist to work with to create a realistic image of the Milky Way (that is, one based on hard data, rather than a schematic four arm galaxy).

Anyone interested?


Cheers mate.

For an hours work it’s ok, but it’s not quite what I was aiming to achieve. I really wanted to get some fine detail in there (like the dark regions in the image below), but I couldn’t think of a way to do it easily.

My understanding is little is known about large parts of the Milky Way, is that true? If so you’d really be aiming to create a plausible rather then realistic 3D model of our galaxy. For the regions where there is good data, it should be straightforward to generate a 3D model provided the data (location, size, color etc) for the stars you want to show has been cataloged.

Can you tell us a little more about your requirements, and what data exists?

Cheers for the link, it’s perfect to have a quick play with.

So what are you after exactly KevinJ?

manbitesdog are u from holland?

No, southern hemisphere.

As Nyrath implicitly pointed out with his link to my sources, quite a bit is known about the Milky Way, especially within about 10 thousand parsecs (about 30 thousand light years).

So it is surprising that there have been very few attempts to create an accurate image. The Milky Way is not quite the grand scheme spiral like the Whirlpool galaxy whose image appears above, and in particular, the Perseus arm is not very distinct along most of its length.

But that doesn’t mean that it would not be possible to create an image. Some astronomers have suggested that the Milky Way might look a bit like NGC 1232:

which has lots of spurs and indistinct sections.

Take a look at

to see some of the maps on my site.

What I’d like is an image that incorporates that kind of detail.

Sorry, I wasn’t very clear. What I was getting at earlier is do you want a new version of this image (ie a lot of artistic license taken), or do you only want the data you have visualised (like the area you shaded in this pic

I imported the data into blender to see what it looked like in 3D if you’re interested by the way EDIT: blend had an error.

ok, if you wan realism…do NOT use the blender star feature, it is quite ugly imo…ad there are no stars outside of a galaxy…

other than that, its lookin pretty nice.

r u sure there isn’t iliketosayblah?

Fair enough, blender stars are a bit ordinary.

I’ve imported the data from the excel spreadsheet, this time rendering the cl (white), co (red), hii (green) and st (yellow) types (cluster, molecular cloud, ionized hydrogen and stars I presume?) separately. Obviously this is just a test render.

And a fly through:

So (Edit, sorry I meant KevinJ!), is something like this what you had in mind? I could use the visualization for a competition I’m considering entering, so I am interested in working on it.

Glad to see you’ve made some progress!

As HII regions are red in reality, I think it would make more sense if you rendered those as red and CO clouds as green.

In terms of the kind of image - if you just use the data, you get something that doesn’t look much like a galaxy.

So what I’d like to see is something like the NASA image (which has the correct overall architecture for the Milky Way according to the current scientific understanding) with more details driven by the actual data.

Oooo! Nice work! :yes:
How did you import the data? Did you write a Python script?

Ok I understand now. Since I can’t see any discernible structure I’ll just use the same basic parameters as used in the Nasa image.

By the way, I’ve been unable to download the star_data.xls file on your site. It seems to download ok, but when I open it Excel tells me it is corrupted. Could you upload it somewhere else perhaps?

Cheers mate.

I opened the spreadsheet in Mathematica, and wrote a little program to process the data and then export it as a mesh in dxf format. Python would also work, but I’m more familiar with Mathematica.

That picture’s in my screensaver! Nice galaxy. It looks better than my galaxy but mine was made of halos.

Hi mate, I think I might have confused you. I didn’t make the pic you quoted, I was just saying that’s what I was aiming for! The galaxy pic in the first post was made with halos by the way.

So Kevin, I came across this schematic of the arms of the milky way I’m considering to use as my base, does it look ok?

Hi manbitesdog,

A better starting image (although it is rotated 180 degrees) is in the section “The Shape of the Milky Way - The Evidence” here:

This appears to have been copied directly from one of the maps in a series of important scientific papers on Milky Way structure from Hiroyuki Nakanishi and Yoshiaki Sofue. Eg:

Actually, even better to use the image in that paper. I suspect that the Wikipedia image is a dumbed down version of that.

A while back, I overlaid my data on top of the Nakanishi and Sofue molecular hydrogen map and found a reasonably close match.


The good thing about the Hiroyuki Nakanishi and Yoshiaki Sofue map is that it reveals how “flocculent” different parts of the Milky Way are. Flocculence is a measure of how much a galaxy departs from the grand design spiral arm scheme. For example, the Perseus arm seems quite flocculent, and the Sagittarius/Carina arm less so.

Here’s an example of a highly flocculent galaxy:

Keep in mind that the idea that the Milky Way has four distinct spiral arms is actually far more speculative than many people realise. It is not that easy to see these arms in the Hiroyuki Nakanishi and Yoshiaki Sofue map although they do their best to find them in the article I linked to above.

I have re-uploaded star_data.xls and you should be able to get it now from my Sources page. Thanks for pointing out the problem.