Celestia Forums

A sleek spacecraft first seen in the movie "Der Schweigende Stern" (1960) (also known as "The First Spacecraft on Venus" in the U.S.)

This Addon is a long way from being complete, but I suspect it'll be a long time before I finish it. In the meantime, I thought perhaps someone might like to add this prototype spacecraft to their fleet

Celestia v1.5.0pre1 or later is required.



http://www.lepp.cornell.edu/~seb/celest ... trator.zip ( 413 KB)

No problem of depth sorting?

Of course, problem with depth sorting: the flames are partially transparent.

The viewpoint was carefully chosen.

selden wrote:Of course, problem with depth sorting: the flames are partially transparent.

The viewpoint was carefully chosen.

Mmmf, I new it...
Well patience...

Even so... sweet. Thanks, Selden! I'll have to look up information on the movie, and see if it woul be a good rent (or even purchase).

...John...

From a story by Stanislaw Lem, eh? I expect it could have been a good film. However it seems to have been mangled somewhat in the creative process, if this review is anything to go by.

http://www.dvdtalk.com/dvdsavant/s138first.html

Sweet.

Is there an orbiter version of it (made by anyone, of course)

I dunno if an Orbiter version is available. i probably could generate a model file: an Anim8or plugin that exports Orbiter's model format supposedly is available now. I don't know enough to be able to generate a scenario, though.

selden wrote:Of course, problem with depth sorting: the flames are partially transparent.

The viewpoint was carefully chosen.

There is a fairly simple solution to the depth sorting problem for translucent things like these flames. Luminous objects such as glowing flames can be rendered with additive blending, which is sort independent. I can make a simple modification to model rendering which will assume additive blending for mesh components with an emissive color.

--Chris

chris wrote:

selden wrote:Of course, problem with depth sorting: the flames are partially transparent.

The viewpoint was carefully chosen.

There is a fairly simple solution to the depth sorting problem for translucent things like these flames. Luminous objects such as glowing flames can be rendered with additive blending, which is sort independent. I can make a simple modification to model rendering which will assume additive blending for mesh components with an emissive color.

--Chris

If this could solve the flame problem for the shuttle reentry (the model I send you a time ago), it would be great!

In this case, the problem is that the flames and the spacecraft are two separate objects, and the spacecraft is surrounded by an invisible cube.

To explain why this is a problem, you have to understand that Celestia only does very coarse sorting of translucent objects. Generally, translucent objects are rendered in order from back to front after all opaque objects have drawn. This doesn't solves opaque/translucent ordering issues, but translucent objects may not be ordered properly with respect to each other if they overlap.

Obviously, the flames are translucent. But, the invisible box causes the spaceship to be marked as translucent as well, and thus the usual translucent/opaque sorting fails. If the flames were part of the spacecraft model, they would be sorted correctly. A workaround that I attempted was to ignore completely transparent objects (opacity = 0) when determining opacity. Unfortunately, the boxes in the Cosmostrator model are not quite completely transparent. If I nudge the complete transparency threshold up to 0.01, the sorting works fine.

So, in summary: I have a simple fix that will make the Cosmostrator look right, but the opacity of the bounding boxes should really be zero, not ~0.007 as it is now.

Ideally, the boxes shouldn't be necessary at all, but I understand that they're required right now to accurately position the flames relative to the spacecraft, because Celestia normalizes mesh sizes at load time. In upcoming post, I'll propose a scheme in which the rescaling could be avoided, making mesh positioning much easier.

--Chris

chris wrote:In this case, the problem is that the flames and the spacecraft are two separate objects, and the spacecraft is surrounded by an invisible cube...

Nope there wasn't any transparent cube... I had to change the size and MeshCenter of the flame to match the correct position of the mesh using the same orbital datas than the suttle...

ElChristou wrote:

chris wrote:In this case, the problem is that the flames and the spacecraft are two separate objects, and the spacecraft is surrounded by an invisible cube...

Nope there wasn't any transparent cube... I had to change the size and MeshCenter of the flame to match the correct position of the mesh using the same orbital datas than the suttle...

Sorry, I was talking about Selden's Cosmostrator. I'll try your mesh tonight.

--Chris

chris wrote:

ElChristou wrote:

chris wrote:In this case, the problem is that the flames and the spacecraft are two separate objects, and the spacecraft is surrounded by an invisible cube...

Nope there wasn't any transparent cube... I had to change the size and MeshCenter of the flame to match the correct position of the mesh using the same orbital datas than the suttle...

Sorry, I was talking about Selden's Cosmostrator. I'll try your mesh tonight.

--Chris

Oups, I thought the "in this case" was refering to the last post (mine)... sorry...

The not-entirely-transparent cube is a fudge.

1. I like to make models in pieces that are assembled for viewing by an SSC file. This is especially useful for objects that have pieces that are lost during flight, like multistage rockets.

It simplifies the SSC catalog if the pieces all are the same size with the same center: all can be specified to have the same radius and follow the identical orbit.

To ensure the separate models all have the same size, I usually place tiny transparent cubes at equal distances on the +/- xyz axes beyond the bounds of any of the pieces. This was an experiment to see if I could do the equivalent using a single cube instead of 6. It causes various problems while modeling, though, so I probably won't use that method again.

2. Anim8or has an annoying tendency to optimize things away when values are set identically to 0, so I usually specify 0.001 instead of 0. I accidentally specified 0.01 in this case.

selden wrote:A sleek spacecraft first seen in the movie "Der Schweigende Stern" (1960) (also known as "The First Spacecraft on Venus" in the U.S.)

This adaptation of the Stanislaw Lem novel is also known with these titles: L'?©toile du silence, Milcz?…ca gwiazda ,Silent Star , Planet of the Dead or Spaceship Venus Does Not Reply. But the English (USA) version is grandly recut from the original (and rewrited:An International mission replace the Soviet one...) and 14mn are missing. BTW, it was mocked in the humorous collection Mystery Science Theater 3000 (ep. #45, 1990 IIRC)!
I like these old sci-fi movies... This one is not so ridiculous, a good sumary of the book...

Jeam,
just a short message because I have no really time fort Celestia for the moment: I spare my free time with a listing... of complete Science-fiction published stories turned on screen! Sorry...