Celestia Forums

Hello everybody.

I am compiling an add-on file for all asteroids more than 70 km in radius. So far I've got orbital data for all of them with radius greater than 100 km, and I'm progressing swiftly. I plan to include all information available, and particularly, color (since light curves are known for most of them).

However, having no expertise with 3D modeling, I'm using the same mesh for all of them (roughsphere.cms). It's a good approximation to most of the shapes (since they haven't been photographed in close range), but it's somewhat boring to see the same asteroid in different orbits, time and again.

I think a practical way to overcome this problem would be to add a new configuration parameter for asteroids, allowing to scale their mesh in three axis. The idea is somewhat similar to the Oblateness parameter used in gas giants, but not only for Z, but also for X and Y. The format would be something like:
Scaling [ xScaling yScaling zScaling ]
with x, y and z scaling parameters between 0 and 1 (1 meaning no scaling at all).

This way we could use the very same mesh to represent different asteroids whose general shape is known and similar to a distorted sphere (and there are plenty of them, believe me!) This could also solve in part the "lunar boredom" problem, as all shapes of tiny moons are drawn from the same 3D mesh.

Tell me what you think, or if this would be too difficult to implement. Thank you for your attention.

Bruckner

I'll see what I can do . . . adding scaling terms would be straightforward, but I may be able to come up with a way to generate asteroid models with shapes that vary in more interesting ways than just scaling.

--Chris

That would be wonderful. I think scaling would be a start, but if you think you're able to implement other modifications, that would only add to the variety.

I was thinking of randomizing the surface appearance. That, however, has a problem... It's true that only general shape is known for a lot of asteroids and small moons, but a random appearance would keep objects appearing in different ways for different runs of the program, and that wouldn't be correct, IMHO.

However, a wider choice of textures would also help, and I may be able to come up with some. I've already got a color-adjusted texture for reddish asteroids like Juno, Iris, Eunomia... A credible representation can be achieved just by scaling a basic shape before applying the texture, I believe.

Keep up the good work...

Bruckner

Bruckner, I think randomizing the mesh is the best idea. Individual asteroids don't have to look different each time you regenerate the mesh as long as you reset the randomizer with the same number everytime. You can "invent" the seed value each time you need one by, for example, adding up the integer values of each letter in its name, or by adding all of its orbital parameters together.

Each asteroid will then look unique but consistent!

I'm ashamed I didn't think of that possibility myself... it's just brilliant!

Best regards

Bruckner

I guess no one likes my dozen asteroids zip:

http://www.vicware.com/space/DozenAsteroids.zip

I have to use these right now, because any cms reference crashes
Celestia on my cpu (1.2.4).

I'll have a report this weekend on 1.2.4

Vic

Finally I managed to post my compilation of orbital data for 50+ asteroids (the bigger ones, except for Ceres, Palas and Vesta). It's here: http://bruckner.homelinux.net/addons.html.

I haven't had time to look at your asteroids, but it seems that I could substitute the generic asteroid shape I used (oops, a cms file ) by yours. Since the general shape is known for many of the asteroids I listed, I think there is some work to do... Would you accept bundling my orbital data with your asteroid shapes in the next revision of my asteroids.ssc file? I would package them together and offer them from my site.

By the way... Is it too costly for you (time-wise) to come up with a specific shape for an asteroid? I always wanted to have a 3ds mesh for Hektor, a strange double asteroid (think two tangential rough spheres and you'll have an accurate image) which is the largest member of the Trojan group. This web page: http://www.cen.uiuc.edu/~sulfridg/asteroid/asteroid.html shows a collection of known asteroidal shapes, including Hektor. Maybe it could be useful as a reference...

Best regards.

Bruckner

Absolutely please use them.

They actually need better textures. I just quickly ran them thru
3dsmax - it's variations on a themes.

I wish I had more time - wish we all could make a living
just working on Celstia - wonderful project.

Vic

bruckner wrote:
The idea is somewhat similar to the Oblateness parameter used in gas giants, but not only for Z, but also for X and Y. The format would be something like:
Scaling [ xScaling yScaling zScaling ]
with x, y and z scaling parameters between 0 and 1 (1 meaning no scaling at all).

Technically we could just have two such scaling parameters without any loss of generality, since the overall size is already independently scalable. But I suppose having all three would be a convenience.

A few more parameters could suffice to make "peanut" or "dumbbell" shapes, since so many asteroids whose shapes are known seem to be like this: WaistWidth and WaistPosition, perhaps.

On the subject of actual shapes of some smaller bodies, you might want to take a look at Wm. Robert Johnston's page of shape models. The URL is:

http://www.johnstonsarchive.net/spacear ... odels.html

He includes some POVRay images that are quite impressive, based mostly on shape models and shaded-relief maps from Phillip J. Stooke. These sources can be found at:

http://publish.uwo.ca/~pjstooke/plancart.htm

and

http://www.ssc.uwo.ca/geography/spacemap/contents.htm

The Celestia credits indicate that we're already using some shape models based on Stooke's work, but I don't think we're using his maps.

- Hank