I generated a few stereo images of the etched terrain from Viking to Victoria craters. I'm not 100 percent confident in my renditions, simply because the landscape looks pretty bizarre: undulating terrain with parallel east-west trending ridges that even affect the ejecta around Victoria and Erebus.

Here are my source images. Thanks to Alan for pointing them out to me:

Left: http://www.msss.com/moc_gallery/r10_r15/im...5/R1500822.html
Right: http://moc.rlproject.com/end_vic_1m.jpg

Anaglyph:
733 Kb: http://www.marsgeo.com/Photos/Opportunity/Etched_LRa_s.jpg
2.0 Mb: http://www.marsgeo.com/Photos/Opportunity/Etched_LRa.jpg

Parallel method:
267 Kb: http://www.marsgeo.com/Photos/Opportunity/Etched_LR_s.jpg
1.6 Mb: http://www.marsgeo.com/Photos/Opportunity/Etched_LR.jpg

Crosseyed method:
265 Kb: http://www.marsgeo.com/Photos/Opportunity/Etched_LRx_s.jpg
1.6 Mb: http://www.marsgeo.com/Photos/Opportunity/Etched_LRx.jpg

Wow! If it is accurate, it looks like we have 3 ridges to climb shortly after leaving V & V - just to reach Albert. Then between Albert and Victoria - even more.

I wonder if Victoria Craters invisbility so far might be due to the crater itself sitting in a larger depression? The white band of outcrop surrounding the smooth region outside of Victoria might define the depression? Just a thought.

Thanks sooo much, Aldo! You did just what I was suggesting here earlier. (I was first a bit confused with the source for the right-eye image.) The parallel east-west ridges that were in my attachment image of Vic flank are there again.

It's admittedly pretty odd looking. I think, could it be so that it's indeed really 3-D in some level, i.e. concerning the relative heights between individual details - but that there is additionally some flaw which renders all that undulating? Looking at it, even Viking & Voyager seem to be sitting in a trough between two ridges. That doesn't quite make sense - we should see the ridges in MER images also!

I do hope we get somebody commenting who knows something about the technical side of MOC images.

The thing that I find fishy is that the ridges are almost perfectly straight, parallel, and equidistant peak-to-peak. And Erebus steems to be straddling one.

That could easily be an artifact from vertical scaling of one of the eye images, and it only becomes apparent when viewed in 3d against the other eye which is not scaled. Sort of like an inference pattern if you will. There is no way that it's a real feature.

Yes, the ridge pattern is an artifact, it's clear after a thorough look. I could also see (quite vaguely) in places a north-south lineation of narrow, alternating lower/higher parallel patches, another artifact that must be? These patterns make it virtually impossible to decipher if any real features show in 3d about the etched terrain. Alas.

Still, if you look at the crater rim in the image, the floor clearly drops below the flank - and that is the only case where we can see any significant changes of height in the east-west direction. I assume the drop should be big enough to be able to counter the artificial effects. This makes me believe it's basically a stereo pair after all, albeit the flaws destroy the real 3d-effect badly.

Is there anyway to remedy this with a scale change?

scott

Regarding the topographic relief artifacts in the stereopair of the etched terrain: I've seen extended abstracts from either Lunar and Planetary Science Meetings, Mars Workshops <mostly linked to from the LPI Lunar and Planetary Institute> or International Mars Conference a couple years ago on this.

The MOC Camera on Mars Global Surveyor was designed to take vertical monoscopic black and white images with the highest resolution practical. It's geometric precision is perfectly adequate for that purpose but is far from ideal for stereo work, especially in very flat terrain. Also, the spacecraft is *NOT* a stable platform, wobbling slightly in attitude during the "push-broom" sweep of the field-of-view of the camera's line-scan CCD across the scene. These imperfections put various "corregations" in stereo pairs and there isn't enough precise info on attitude in the telemetry to precisely correct for it. The experts have had a hard tiem trying to get artifact-free stereopairs from the MOC images, despite access to camera calibration files and the spacecraft engineering telemetry.

Basically, we're trying to do something the camera wasn't designed to do.

While it is true that the MGS is a "wobbly'" platform and can often result in wavy stereo images like this, it's certainly not true all the time. The MOC is in fact perfectly capable of superb stereo pairs for anaglyphs or for free viewing. The experts should only have trouble trying to use these images for digital terrain models. A bigger problem for stereo image creation arises from the fact that there are months or years between images of the same terrain. Any differences in the images such as mismatched sun angles (different season, different time of day), surface patterns that are present in one image but not the other (such as dust devil tracks or new slope streaks) or differing amounts of dust in the atmosphere can all play havoc with stereo pairs. Image quality also varies widely.

Even with all of this there are still literally hundreds of great stereo pairs available. Here is an example of some Terra Meridiani craters not too far from Opportunity:



Here's an Aram Chaos anaglyph:



Lots more here:MarsUnearthed.com

You'll need your red/CYAN glasses.

Very impressive relief on that bottom image; lots of ice and big dunes. Looks interesting, but it'd kill any present rover designs.

Looks like the craters on the top image are in a good place to spawn regular dust devils.