Page 1 of 1

Boomerang

Posted: 15.06.2006, 16:18
by eburacum45
I thought I'd try my hand at a 3d nebula; here is the results so far.
The Boomerang Nebula is supposedly the coldest place in the universe yet found outside of the laboratory;

here is an APOD of the object
http://antwrp.gsfc.nasa.gov/apod/ap030220.html
the cooling is due to expanding gases, apparently.

My model so far- about 5 megs- by golly these 3d models get big quick
Image

Now if you go round this object it starts looking like other celestial objects in the sky-
here is a planetary nebula
Image

here it looks like the Helix nebula
Image

this kind of looks like Eta Carina
Image

Another view
Image

Please ignore the radius shown onscreen, which is just the size of the smallest, innermost component. The model will be about 2 light years long - when it is finished.

Posted: 15.06.2006, 16:24
by ElChristou
Looks pretty good so far!

Posted: 15.06.2006, 17:33
by Johaen
Very pretty. I really like these 3D nebulas. They look much more real in a 3D enviroment then the flat ones.

Posted: 15.06.2006, 21:19
by jll
Wow 8O , your multi-layered model seem to be very impressive from each side !

I'm very excited to know how you managed to build such a model !

When will it be downloadable ?

I can't wait !!

JLL

Posted: 16.06.2006, 01:13
by buggs_moran
Vast lack of sharp edges, very nice. How many layers are there?

Posted: 16.06.2006, 07:01
by Paolo
Impressive! 8O

Kind regards

Posted: 16.06.2006, 07:11
by eburacum45
Oh dear, I have just counted- there are about forty layers in the model.

But only one texture!

My strategy was to make a small cloudy patch on a .png texture, then apply it to lots of shapes which I pieced together to make a wasp-waisted structure. So it has lots of layers, but they are all the same texture.
If I were to make a different type of nebula a different strategy would be required, using several textures.

Posted: 16.06.2006, 16:10
by Paolo
Can you please post some screenshots of the inside of the nebula with the observer watching at some planetary remnants?

Kind regards

Posted: 16.06.2006, 18:17
by buggs_moran
So stacking the meshes resulted in the brightness and rise in opacity? Or was there some the texture pretty homogenous?

Posted: 16.06.2006, 20:31
by eburacum45
Not sure why you want to see some planetary remnants inside the nebula, but here they are,
Image

actually making this image made my computer crash, so it is obviously pushing the boundaries of my set-up.

To Buggs;
the texture is fairly transparent, but stacking it up, flipping and rotating it the nebula appears quite thick in places.

Posted: 17.06.2006, 02:51
by DonAVP
eburacum45 wrote:Oh dear, I have just counted- there are about forty layers in the model.

But only one texture!

My strategy was to make a small cloudy patch on a .png texture, then apply it to lots of shapes which I pieced together to make a wasp-waisted structure. So it has lots of layers, but they are all the same texture.
If I were to make a different type of nebula a different strategy would be required, using several textures.


Regarding the 40 layers and the size of the file. You might try deleting half the layers that would reduce the file size. Check out the results you might be supprised that there is not that much difference. Just a thought. You might even get by with fewer that 20 just have to play with it.

Don

Posted: 25.06.2006, 19:23
by eburacum45
Well I tried making the model with fewer layers, but it lacked the fullness of the forty-layer model- so I tried another trick- I made several .dsc declarations for the same model, each one a slightly diferent size and oriented at a slightly different angle. This made the resulting model thich and full-seeming.
Here is the result.
Image

Posted: 26.06.2006, 01:44
by buggs_moran
I have had some success with this method as well eburacum. I have also toyed with changing the oblateness of models. When using one texture and one model with different sizes and oblateness it cuts down on load times. Do you have any flicker problems?

Posted: 26.06.2006, 17:45
by eburacum45
I don't notice a flicker with this model. I have (however) noticed a flicker on some of my spaceships.
One trick I like to use is to make the port and starboard lights and other similar lights with a semitransparent halo- this semitransparent halo flickers quite noticably. It almost looks like deliberate animation...

Posted: 11.07.2006, 20:30
by eburacum45
As we were talking about flicker with semitransparent meshes, here is a particularly strange example. I have been updating my model of the Bishop ring habitat concept; I decided to add a moving transparent cloud layer.
You can see an early version of the clouds moving here;
(movie)
http://www.orionsarm.com/movies/Movie_0001.wmv
and here is a more recent version of the model with better textures.
http://img66.imageshack.us/img66/9339/manaus1dj.jpg

but once you get inside the ring the clouds disappear altogether;
http://img66.imageshack.us/img66/7838/manaus38fl.jpg

It ia quite annoying, although it does allow you to look at the ring floor texture a bit closer. I have tried flipping normals and all the usual tricks.

Posted: 12.07.2006, 14:28
by Chuft-Captain
eburacum45 wrote:...but once you get inside the ring the clouds disappear altogether;
http://img66.imageshack.us/img66/7838/manaus38fl.jpg

It ia quite annoying, although it does allow you to look at the ring floor texture a bit closer. I have tried flipping normals and all the usual tricks.


eburacum45,

I think I experience the same problem when I put clouds inside my O'Neill Habitat. From certain angles the clouds just disappear. For example in the case of the video "Golden Gate in Habitat" at http://traitorsclaw.sitesled.com/videos.html you can see the rotating clouds perfectly OK from the viewing position high on the endcap, but from some positions they will just disappear.

My theory is that this may have something to do with the well known depth-sorting problems or disappearing models problems documented in other threads, as it seems to exhibit some of the same symptoms.

Just speculating. I know that's probably not much help, but that's my theory for what it's worth.

Maybe, just maybe, these problems willl disappear with Chris's current work on enhancements to cloud rendering.