Bob,

Sounds like a very clever technique. It would improve the spatial resolution of images, but doesn't have any bearing on the question of color.

Similar techniques have been used in an adhoc way to synthesize high resolution images from existing images. I think it was Ames that took advantage of subpixel pointing variations in Pathfinder images to create images of double the resolution. When the camera took two images of a scene whose center points were different by about half the angular resolution of a pixel, it was possible to meld the images by calculating the differences between partially-overlapping pixels to produce an image with twice the (virtual) resolution.

Color's a whole different animal. Historically there have been cameras with pretty good filters (color resolution and accuracy) landed on Mars, but the orbital cameras have had pretty poor color capabilities. Green has gotten shortchanged--both Viking and MGS provided red and blue and we've had to synthesize green as the average of the two.

All I could find out about the Mars Express camera is that it has "red, green, and blue" filters. I don't know anything about its accuracy, but I'd observe that the released images have an oddly uniform color that makes me wonder about its calibration.

And then there's Odyssey/THEMIS. I haven't tried to work with the raw data, though that's been on my long to-do list for a long time. What I hear is that the images are hard to work with, consisting of offset bands of subimages making up full images; and that getting images from each of the filter channels registered as the camera moves between images is a tough task. I can't comment intelligibly on the THEMIS controversy.

Looking at what I just wrote about color, I guess I don't actually have much to say about orbital color images. As for the controversy about MER color images, this is a good time to update what I've learned about it in the last few months of researching the issue and writing color image software. The PanCam is a fine instrument for producing good color; the problem is calibration (as always).

I have a copy of Justin Maki's 1999 paper on the "true color of Mars". It seems very solid, but it's based on results from only the Pathfinder site and some correlation with Viking Lander sites. It's a good place to start, and I can well imagine Maki spending the next couple of years integrating MER results into an updated paper. There were no "blue"berries at the other sites, and he has a lot of new colors to add to his model. But I don't think that he'll be changing the baseline average color of the Mars landscape illuminated by the Martian atmosphere; I think his 1999 calibration is pretty solid.

What I found writing the software is that it is possible to get into the same spectral ballpark as NASA even without precise calibration information for each image. The real difference between the images I produce and the images NASA releases are in how we treat enhancement variables like brightness, contrast, and saturation. NASA's playing it safe, producing Maki-esque images that probably do accurately represent what an unaided human eyeball would see standing on Mars (or to be precise, embedded in the head of a human standing on Mars--I had this weird mental image of an eyeball with legs...). They've taken so much heat over the years with charges of "playing" with the colors that I can understand why they'd want to release images backed up by a credible scientific paper.

Unfortunately, the safe images discard a huge amount of visual information that would beyond a doubt improve the viewer's comprehension of the scene. It's up to people outside NASA to go out on a limb and enhance the images (within the bounds of starting with well-balanced color) to bring out useful information. I've said it before: It's all fine and dandy to theorize what the unaided human eye would see on Mars, but in reality humans who actually go to Mars would bring along flashlights and other artificial illumination so they can see what they're doing, and would probably be equipped with vision enhancement gear descended from the night-vision goggles soldiers use. So it is legitimate to enhance images to a more useful standard, what I sometimes call "true color(augmented)".

I guess what I'm saying is that from my perspective there's actually not much controversy about those images. NASA's producing good images, and people like me and Keith Laney who make an effort to balance the channels are producing another kind of good image. The least-good images floating around the Net are the ones which are simple uncalibrated RGB combination. They misrepresent the landscape in a way that's sure to cause a lot of misunderstanding and confusion as time goes on. In the early days I contributed more than my share of those, and I have some work to do to make up for it.