Celestia Forums

I'm developping a family of 15 generic planetary nebulae models, all made of sprites (3D volumetric models).

I intend to use the models in Selden's old addon, which shows 73 PN in the Milky Way. I'll release the addon later this weekend, when I'm satisfied with the models. The full addon takes about 8MB on disk (3.5MB zipped). The frame rate is very good on my now "old" Mac mini.

Here are three previews. The first picture below is showing all 15 models. The second picture is a close-up of 6 of them, while the third picture is a close-up of one member in the family (click the pictures to get a larger view).





These models are used as DSC objects. I'm inspecting them to be sure that they all stay beautifull from all angles and from the inside too (sprites may be capricious, at times...).

As you can see from these pictures, 10 models have an ashes shell and an ionized core (the ashes shell is the same for all these generic models). Five models are just an ionized shell.

I changed the color of 5 ashes shells, for more variations in the family and color consistency with their core.
Here are two examples :

Two new models :


I'm not sure I'll include them in the pack...

--- edit----

Thanks John.

The pack now includes 17 nebulae. I'll release it soon, after I do more tests.
It's a pitty that there isn't any lobes-like models (the results aren't very satisfactory yet).

The frame rate is good enough for my taste.

Here's a download link to the addon :

http://fsgregs.startlogic.com/Public_Fi ... ebulae.zip (3.9MB zip file)

Take note that with Celestia 1.6.1, you'll get lots of annoying remaining dots at a large distance from the nebulae, like this :


This very long standing bug is finally fixed in Celestia 1.7.0 (not published yet, though). It was about time !
(During the next few weeks, there should be some great news on the forum... shhh ! It's a surprise !)

I can't stop creating planetary nebulae !

There are so much variations possible, I don't know anymore which models to use.
Here are three models I made today, at breakfast :




I tried to do some irregular Orion-like nebulae, but it's very difficult to get satisfying results.

My 73 planetary nebulae addon is now using 23 nice spherical models (I'll probably have to update the addon above).
I have to use a CELX script to visit each one of them.

From now on, I'll work on axi-symetric nebulae. Here's an example :


Same model, with reduced luminosity :



I'm very tempted to play with the Hydrogen atom orbitals, to get more complicated structures. The problem with mathematics, is that there are so much possibilities, it may take years of experimentations with sprites...

I like very much the colors and luminosity on that one, even if it's not realistic (but who knows...)

Very nice work indeed. Reminds me a bit of cell division.

Could you make sculpture like this and to see how they looks?



(without the extruding base, this is a heightfield)

Fenerit wrote:Could you make sculpture like this and to see how they looks?

You're touching the problem I want to solve : filaments !

I want to add random filaments structures to the nebulae (like in the Crab nebula, for example). I simply don't know how to do it yet.


Basically, the method I'm currently using in Mathematica can't do filaments. The statistical distributions I'm using are ploting random points, and each point is generated independently from all the previous points.

I need something which could generate random points with some correlations (filaments, percolation, Poisson statistics, etc...).

If I could solve this mathematical problem with Mathematica (there is certainly a way), I would then be able to reproduce the Crab nebula (that would be awesome !), and much more interesting models.

EDIT : Note to self : a textbook about random fibers with Mathematica :
http://books.google.ca/books?id=dl9aGrM ... ca&f=false

I used one of the previous experiments to do this one. It's pretty in Celestia and I may keep it.

I think that I'll include these two models in the final pack. So currently, there are 25 "generic" models for the old 73 planetary nebulae from Selden.

A HUGE revolution in my way of creating nebulae models for Celestia has occured today !

With the help of some folks on a Mathematica forum, I can now generate a new family of models, much closer to the Crab nebula, supernova remnents, etc. I still need to adapt the Mathematica code to export the data to Celestia, and this means a lot of work to do during the following weeks. Here's a preview, direct from Mathematica :



In place of the small balls, there will be a Celestia sprite for each ball, of the same color (as defined in Mathematica). I still need to define the sprite size and transparency...

Guys, try this in Celestia !

It's just a first and very crude model from my new Mathematica code, but it's really promising !

In Celestia, just type "Crab ball" and go ! Move around, especially at the exterior.

Little Crab, I'm gonna get ya !

Here's another test model. The Mathematica code is getting better, but there's an issue with the number of sprites which are too dense in some filaments intersections. This will be very hard to fix and may even kill my project


Just replace the cmod file in the small addon given above.

Preview :

I think that's rather good!

abramson wrote:I think that's rather good!

Well, it could be improved **a lot**. I need to give some thickness to the filaments, with smaller sprites in them. Currently, I'm not sure how to tweak this. When done, the resolution would be much better, with finer details.

The Mathematica code isn't functionnal for the sprite size yet. This is giving me some troubles.

At least, the colors are very easy to change.

EDIT : The sprite size code is now functionnal.

Now, this is beginning to look like a real supernova remnent !





This model is just a bit too big to be published here (zip of 300KB).
I now have to refine the filaments, with more and smaller sprites, and to select a better color palette.