Enjoy!

It's a command-line utility, so if you're not familiar with that, just drop .img file (with XYL label) into .exe

Not really useful, just clouds of points, but still interesting to see. Turns out they have LOTS of points in these files...
I only tested it on sol 419 for now, so let me know if it crashes on some other data.

And to think - I was about to go to bed.



I tried
http://anserver1.eprsl.wustl.edu/navigator...FF40AJP1745L0M1
http://anserver1.eprsl.wustl.edu/navigator...SF1129P1939L0M1
http://anserver1.eprsl.wustl.edu/navigator...FF0215P1320L0M1

All generated VRML files between 10 and 25 MB, 3ds Max is notorious at being crap at importing VRML, so I used a standalone VRML viewer to have a look at the clouds....

1 is Spirit look at Humphrey, 2 is Navcam wedge that includes the heatshield impact for Oppy, and 3 is Oppy RHAZ egress (that's the lander behind it)

Doug

Here's a Navcam wedge from Spirit Sol 12, you can even see individual cells on the array in the foreground

Doug

Almost forgot - if you have corresponding .png file in the same folder (with the label FFL), the converter will try to "colorize" the points.

Yes. yes it does indeed. Legendary altogether.

Oh the possibilities!

Pretty nifty. I hadn't been aware that they'd posted this sort of information.

The info is out there - we just need someone to make it into a 3d friendly format, and then the visualisations will come out like a fire hose

Doug

This thing claims support for PFB:

http://www.photo-freeware.net/3d-viewer.php

But I think to work properly it needs SGI Performer installed anyway...

Those NASA folks have an unnatural affinity for all things unix. Sad. Frustrating. Out of touch. I could go on... but we don't want to get into flame wars

Well, it's not really a platform issue, because there is a Windows version of OpenGL Performer too.
The problem is that PFB seems to be proprietary SGI format, so their Performer needs to be installed. And professional SGI software usually extremely expensive...

ViSTa, on the other hand, is NASA format, so there is some hope here.

How easy is it to calculate the the difference in location of the cameras at two consecutive sites?

The site location is taken as the centre of the rover id'ed by the purple dots here;
http://anserver1.eprsl.wustl.edu/roversite...FFA818P0783L0M1

Can the difference between these site locations be extracted from the published data?
i.e elevation difference and horizontal difference?

I don't know, but if so wouldn't it be trivial then to perform a transformation on the points created in Maxst's wrl files to combine multiple site locations?

Fair enough the files would be ginormagantuan, but still, just curious.

Was having a little trouble viewing the big files.

Just did this quick utility that downsamples the points in the wrl file.
Probably won't work if the points have been colorized with the png, didn't try it though.
Can't try now have to go the Ireland Scotland rugby match... Yayyy!

Have to use the command line though, sorry

java -jar wrlSampler.jar 1R128816115XYL0215P1320L0M1.wrl 3

This takes the wrl file and outputs every third point

java -jar wrlSampler.jar 1R128816115XYL0215P1320L0M1.wrl 1

Would just output the original file.
It's an exexutable java jar file so I just found out I can't upload it.
If anybody wants it just PM me and I can email it to you.

Edit: We won the match and it can sample the coloured wrl file if needs be....

OK - I am DESPERATE to do a lot of interesting work with this - but the 'point cloud' is just too hard to get into 3ds max as a mesh - it just dies. I am sure we are very very close to having something magnificent here - but we just need to go one tiny step further to get something that a lot of people will be able to use and do amazing work with - I just don't know what that step requires

Doug

Don't worry, I just found official ViSTa specs.

This is going to be fun!

Keep us informed

Doug

This sounds like really good news, good luck with it MaxSt!

So far so good... Sample output from the parser:

ViSTa v.0.9
Number of textures: 1
Number of vertices: 650850
Number of LODs: 7
LOD 0: Number of triangles: 2
LOD 1: Number of triangles: 638
LOD 2: Number of triangles: 2940
LOD 3: Number of triangles: 11386
LOD 4: Number of triangles: 43382
LOD 5: Number of triangles: 166874
LOD 6: Number of triangles: 650848

I guess the easiest way to handle all this would be to write each level of details into separate VRML file.

Your member title has been suitably adjusted

Doug

hehe... nah, xyz/xyl is nothing really.
-------------

Meanwhile I'm getting a little frustated... I can read .vst files just fine now, but can't figure out the best format to write the result.

I don't really like VRML, it's too limited. It doesn't support triangle strips, which is a main primitive in vst files.
I decided to switch to X3D instead, but the viewers I tried (Venues and Flux) either crash or refuse to display triangle strips.

Flux says:
Unknown node type 'IndexedTriangleStripSet' at line 10; skipping

Maybe I should find a better viewer, or output to some other format... I heard there are some good X3D viewers for Linux, but I only have Windows. Ugh, so frustrating. If anyone have any ideas, please share them...

dxf, 3ds, obj, stl - all good for me

Feel free to attach a file or two so people can try them out.


Doug