Celestia Forums

I'm releasing a model of Curiosity on Mars.

This is the best conversion I could make from the original Blender model found on the NASA web site :
http://nasa3d.arc.nasa.gov/detail/curiosity-dirty

Two previews of the model in Celestia :



Here's the model for you to enjoy :
(4.1 MB zip file)

Looks pretty good! Well done!

I added the shorter name "MSL" to the ssc, I do this with most of my spacecraft Does the .DS_Store files serve any purpuse? I just removed them...
How about uploadng it to the Motherlode? I'm feeling like this is the better option releasing an addon unless something is left to discuss.

The .DS_Store files are a quirk of MacOS. They're ignored by Celestia.

Well, I can use one word only: astonishing!
It looks like the real things, so thanks a lot, Cham, you made a wonderful job.
Now finally I can show to my students this wonder on the red planet.



Just one thing is missing, anyway, but it's not Cham's fault ... the model shadows.
I remember that in the last posts of the former forum, Chris wrote about the cmod models shadows, and I found the thread here:
viewtopic.php?f=25&t=15775&start=20&sd=d
He says

The shadows in cmodview will be introduced into Celestia once the code is perfected.

Any idea on what kind of code perfecting is needed?
Anyone still in contact with Chris can ask him about?
Just try to imagine the ISS labs, structures and panels shadowing each other, or Curiosity shadows on Mars surface.
Bye
Goofy

You can add vague shadows under objects by making a surface texture that fits under the model, with a darkened patch on it. Not ideal, but doable.

The problem with this idea is that the shadow would be fixed, and inconsistent with the position of the Sun. It would feel weird.

Hi. In the search of a better "environment" for the wonderful Cham's Curiosity model, I found here
http://nasa3d.arc.nasa.gov/detail/gale-crater
an intersting model of the Gale crater.
But (there is always a "but"!) it's in .stl file, that's used for 3D models printers.
I know absolutely nothing of 3d modelling, but anyway I tried to do something, so I found here
http://www.nchsoftware.com/3dconverter/
a simple and very little freeware, the NCH Spin 3D File Converter Software.
In a very short time I obtained a 3ds model, that installed on Mars surface results this way:



But when I go to the Curiosity model, results are not very nice.
Visible triangles, no texture, no normal.... ugly at least!



Moreover the model should be rotated 180° horizontally to match the real Marrs surface. as you can see in this comparison image.



And to fill the cup, it should have to be "thinned" at the edges, in order to smooth that horrible step from the planet's surface!

But my ability on modelling stops here, so I'm asking if here is any willing guy that could solve these problems, and give us what Curiosity deserves, a nice Gale Crater landscape.
Will you?
Thank you so much.
Goofy

This is a job for John VV !

You told it, not I!
Anyway, if John is tuned.... I absolutely agree.
Please!
Goofy

PS: BTW, NASA's 3X heights enhancement in this model is a bit too much, I would prefer a 2X, no more.
My opinion, obviously.

The STL format doesn't support textures, then the conversion in 3DS cannot display them. You could make VT normalmap of the area from xyz elevation data written like ASCII STL by importing the file in Blender or through a GIS program.

Added after 1 hour 7 minutes:
The Curiosity model is beautiful. It can be optimized with the CMODview by merging the meshes and uniquifying the vertices.

As it is: Meshes: 66, Materials: 9, Vertices 152265, Triangles: 50755 -> size on disk: 6.97 mb
optimization: Meshes: 1, Materials: 9, Vertices 65415, Triangles: 50755 -> size on disk: 3.32 mb

I tried to follow Fenerit's suggestions with CMODview, but the only thing I obtained was a dull black terrain.
This gives me the final confirm that between me and 3d modelling there is an invalicable canyon.
Given this, I tried something else.
Searching Curiosity "selfies" I found this image, comparing the rover on 2012 and on 2015:



As you can see the dirt accumulation increased a lot (probably due to sand storms; do you remember Spirit's solar cells problems on 2008?)
So I decided to try to make a "dirtier" rover, and to compare it with Cham's issued here.
The images are "Cham's", and "dirtiest" , i.e. mine.
I tried to take "consistent" images, as possible.
Start with left (LH) images:





and the right (RH) ones:





Follows in the next post.
Bye
Goofy

Let me go on:
here the upper ones:





The side view of the wheels:



and the oblique view.



This is a view of the "real" wheels on Mars, but obviously their appearance changes as soil changes.



I think that this "dirtyer" change makes Curiosity a bit more "a live object" than a "plastic" one, well, this is obviously my opinion, may be I'm wrong.
What do you think about?
The end in the next and last post.
Bye
Goofy

Moreover I found here

http://gizmodo.com/curiositys-wheels-are-falling-apart-and-how-we-can-sol-1626826935

this comparison image taken at sol 708 (now we are at sol 1434, so surely things are worst now than here!), showing the six wheels all-around, with the many damages occurred up to then.



This is the Front-Right wheel damage



I think that at least the front wheels should have such damages, do you agree, Cham?
Anyway, just suggestions, nothing else.
Thanks God, I finished!
Beg your pardon for these very long posts.
Bye
Goofy

There is no question that I'll add any damage to the wheels. The model depict the rover a few months after it landed on Mars, lets say.

About Fenerit's suggestion to compress/optimize the model file, my experience in the past showed me that we may get some problems with the models in Celestia. So this is why I did not used the "optimize" options.

Fenerit, look at your model in Celestia using the wireframe mode (control-W). You may get a bad surprise.

OK Cham, I can agree with you about damages.
But regarding dirtyness?
Opinions?
Bye

Goofy

I like the textures the way they are. My only desire would be to have higher resolution, so the cables are better defined.

But else, that model is perfect.

Sorry, but I do not see weirdness...

normal


wireframe

Hmm, then this is weirder that it isn't weird in this case.

FYI, the optimization works fine also on linestrips got from georeferenced shapefiles, where their polygons shares lots of common vertices. In this case the vertices decimation is close to 50%. On geological SHPs I've gained half a million of lesser vertices throghout.

Added after 17 minutes:
My fault: half-a-million only on Canada shape: thus say more than a million throughout.

Its kind of sad to think that it's feet are hurt and it may stop being able to be mobile before it goes to sleep for the last time. [sad panda face here]