MRO 3D views, Unbelievable... Options

[jekbradbury]

I did the spinny polar pan a while back. Does that count?

[CosmicRocker]

Wow! Yes! Absolutely! Fantastico! ...

I'm all for more of those, but aren't they simply mini-MMB views?

[djellison]

If we can have the FLA from which they were made, there's enough 'knowledge' out there to bash out a load of these, I'm sure. There are only 20 or so 360 pancam pans to be done anyway. I'll happily put together a Flash-pan page to host them all.

[mhoward]

I'm all for more of those, but aren't they simply mini-MMB views?

That one is sort of equivalent to an MMB view using James' color images, which I'm in favor of.

[jamescanvin]

As am I

I'd love to see more of these endless panoramas as well - and as I think I said at the time, if I had the code, I'd happily produce them and host them alongside the normal jpgs on my site.

I have a feeling a couple of new ones are not far away

[ngunn]

There are only 20 or so 360 pancam pans to be done anyway.

And then there's the Huygens descent ones, maybe some lunar surface ones from Apollo, at least two of East Anglia from the edge of space , all the mercator maps of every body in the Solar System - in fact everything which has been snipped to make a rectangle when it should be continuous.

I'm thrilled that this looks like it may happen, even if it's happening in the wrong thread!

Here's the link to the original for anyone who missed it:
http://www.unmannedspaceflight.com/index.php?showtopic=4735

[fredk]

I've suggested before having a virtual cube (or sphere?) in the corner of the image to provide a scale key in all three dimensions. If you see the cube as a cube and not as a skyscraper it means you're the correct distance from the screen.

I like that cube idea.

One thing to keep in mind is that most of these images are intended to extract science. Most of the targets don't have extreme relief, so in order to help extract as much information from the stereo images as possible it helps to exagerate the relief. Recall that the pan/navcam spacing on the rovers is about 30 cm, much more than human eye separation, so at a fixed distance rover pancams will exagerate the relief over what we'd see.

Of course as people have pointed out, for any stereo view there is a distance from the image at which the angular separation of your eyes matches that of the cameras that recorded the view. At this distance a cube will look like a cube etc, but everything in the image will be shrunk by a factor of the ratio of the camera baseline to your eye baseline. So the result is like looking at a miniature 3D scale model of the scene.

The point is that that distance from the screen at which a cube looks like a cube etc is in most viewing arrangements considerably closer than we would normally view an image. So at normal-ish viewing distances, the relief is heavily exagerated. As I said, the main goal is to extract information, not to provide realistic relief.

But I agree a cube 3D scale would at least give us the option of getting the distance roughly right to get a realistic sense of the relief. I'm all for that.

One analogy is with false colour. Tweaking the colour is often done to extract more info than you'd easily get from true-ish colour. But at least true-ish colour is often provided to give us a sense of what the scene might look like to our own eyes.

[ngunn]

. . and nicely back on topic. I assume there is enough geometrical information attached to the pairs used to make the anaglyphs for the image experts here to construct the cubes, should anybody feel it's worth their while. I wonder if that would inevitably be a laborious process, different for each anaglyph, or whether it could be 'automated' in some way so as to work for them all?

[djellison]

You would have to render a pair of them uniquely for the angle between the two observations for each one. And then it would be a judgement call on how to co-register the two renders to be similar to the coregistration of the HiRISE image. It really wouldn't be worth it.

[ngunn]

Two variables: the angle between the views and the distance to the virtual image. It sounds impractical to do 'post hoc' as you describe, but maybe not so for the Hirise team themselves. Maps have linear scales so it seems to me that anaglyphs should have vertical scales too, as a matter of routine. Impractical or not it would be entirely in keeping with the UMSF mission if someone here were to pioneer the idea. I only wish I had the know-how to do it myself.

[charborob]

I wasn't sure this warranted a new thread, and I didn't find it mentioned elsewhere on this forum, so I bring to your attention some flyover movies on the HiRise website.
http://hirise.lpl.arizona.edu/media/

[djellison]

http://www.unmannedspaceflight.com/index.php?showtopic=5073 is similar.