Celestia Forums

Hi,

Another work in progress...

I was planning to do some new models of the basic package of Celestia, and the first one will be the MGS...
For now, this model is 70% done, I have use an existing model from Nasa (vrml) as a template, and complity rebuild it with more details, new panels and... the foil.
Now what about a version with a semi transparent foil? Here is a shot:



Is something like this (textured) can be acceptable for Celestia?

Wow! That looks impressing.
What is the estimated filesize of the model?

maxim

Impressive!!!
I am not worthy!

maxim wrote:Wow! That looks impressing.
What is the estimated filesize of the model?

maxim

Less than 1 meg...
But what about the foil? transparent or not? perhaps 3 versions, opaque, without, and semi transparent?

I suggest without the foil, if the details are interesting. Foils are hard to make as realistic parts. Actually, here's my perception of realistic models for Celestia :

All realistic models (as opposed to SF models) should be made as 3D schematics only, with a minimum of textures. Ideally, textures shouldn't be used on them, except for logos and solar panels.

In Celestia, there's no way we can render realistic models as the real things, because the real things tend to have a lots of complicated details on them which cannot be shown within Celestia. Celestia doesn't allow realistic rendering, anyway, like real reflections, shadows, etc.

I think all realistic models should be used as 3D representations of the real things, like some kind of "3D schematics". There's no way we can have a "photographic" like representation of the real satellites, for example. So I believe all models should be made as some "simple" but convincing representations, with the principal features clearly in evidence. Also, a model with too much details may give the impression that it is very accurate, and this may be an illusion. The user must know, just by looking at model, that it's just a nice representation of the real thing. Keep things simple.

ElChristou wrote:But what about the foil? transparent or not? perhaps 3 versions, opaque, without, and semi transparent?

Looking good.

I assume you are wanting to show off some detailing of the "insides" of MGS ?

I would definately produce a "true-life" model with full covering of foil, and maybe a 50%+ transparent foil version(depending how it looks), to show the inside detailing, that can used as an alternative.

Well, I think I will propose the 3 versions, with, without and semi transparent foil...

Cham wrote:Also, a model with too much details may give the impression that it is very accurate, and this may be an illusion.

Right. I won't invent some missing details... I'm only modelling what I can find on the Nasa photos, with the greater care... For example, the vrml model they use for presenting the craft is quite incorrect... (see http://mars.jpl.nasa.gov/mgs/movpics/mgs_vrml/mgs_vrml.html) I had to rebuild it competly to have the most accurate possible model...

Bye

Hi,

As I'm not really sure (but the logic tell me it should be), someone can confirm if the antenna of the MGS is covered with foil as the MO one?

Radio antennas can't be wrapped in the standard insulating foil. It includes aluminized material which would block radio transmissions.

The only photographs I've found so far of the MGS with its antenna are of ground testing. Only transparent plastic was wrapped around everything.

http://www.msss.com/mars/global_surveyo ... _test.html

Selden,

Between the two 3D shots above, a thumb from a picture used on Nasa sites for the Mars Odyssey, show a wrapped antenna... As the 2 crafts are quite similar in concept I was thinking of the same foil for MGS... Now as seen in this thumb, the foil in front of the antenna is a silvered one... perhaps a special material??

Any conductive surface will reflect electromagnetic waves. As you can see, it reflects light It has to reflect infra-red, too, otherwise it wouldn't be much good as a thermal protective layer. If the coating is really thin and the radio wavelength is long enough, then some of the radio waves will get through. The question is if the attenuation is small enough not to matter.

[edit]

It seems that there was indeed a layer put over the MGS' antenna. See http://mpfwww.jpl.nasa.gov/mgs/movpics/ ... /actv.html
and especially
http://mpfwww.jpl.nasa.gov/mgs/movpics/atlo_pics/actv/7whblankets.JPEG

[/edit]

selden wrote:...It seems that there was indeed a layer put over the MGS' antenna. See http://mpfwww.jpl.nasa.gov/mgs/movpics/ ... /actv.html
and especially
http://mpfwww.jpl.nasa.gov/mgs/movpics/atlo_pics/actv/7whblankets.JPEG

Yes! here it is... Tx a lot, no doubt now...

Hi,

Now the model is 90% over... here without the central foil...



As the mgs is in mapping mode, someone is able to do a ssc with the model always pointing his instruments to the surface?

For a circular orbit, one sets the object's rotational period to be the same as its orbital period. Unfortunately, that doesn't work so well for elliptical orbit: the object will seem to "wobble" back and forth; but that's the way it is in real life, too.

Hi,

Last shots before release... completly foiled version...



Bye

ElChristou wrote:Last shots before release... completly foiled version...

Very nice.

How have you managed to get such good looking gold foil ?

A quibble: that's gold-looking foil, not foil made of gold. (It's an amber-colored coating over aluminized Kapton.)

TERRIER wrote:...How have you managed to get such good looking gold foil ?

Well, you won't believe this: I have done a small model with a pile of cd, small boxes etc, and wrapped the whole with aluminium foil from my kitchen... Some shots with my digital camera, some works in photoshop, desatuation of the layer (=grayscale foil) then a blending of a second layer in color mode with a brown/golden color...
The result is good but not perfect, because of the UV mapping... also the foil details should be much smaller but I won't do a 2mx2m model in my living !!

Bye

Ok everybody, here are the releases:

*****
EDIT

The downloads link have been mooved further in the thread...

*****

Before releasing on ML, I need confirmation that all is ok.
On the semi transparent one, some rendering bugs can exists, but apparently it's the Celestia engine fault, seems there is a problem with transparency...

Bye

Is your transparency being done with the texture's Alpha channel or as part of the material defined in the model?

I had a discussion with Chirs about transparency a while ago. It seems that when transparency is created using a texture's Alpha channel, that it's essentially impossible to do in a single pass and multiple passes would make Celestia unacceptably slow.

Transparency that's part of the material defined in the model seems to work well, however.