Anaglyph and my3D stereo pairs of Homestake, sol 2765. This one took a lot of monkeying around but was worth it.

Sol 2765 Homestake micro image anaglyph and my3D stereo pair. This one works really well in the my3D.

I overlaid Bill's colors on the base MI image and came up with these rough composites. The L257 has a desaturated section on the left where the IDD arm joint was intruding into the color image.

My take on Homestake in 3D...



...and a colour view, too...



...and again, great work Dan

wow! That looks

My take on the L456 set:

and even better... almost looks like a calibrated image Ugordan, what is it you do?

First I try to get the filter combination to look like what I *think* the calibrated representation would look like, which for the L456 set looks more like a dark brownish-red color for the dust and soil. Then I run that through some code that linearly interpolates between the 3 wavelengths and uses the CIE XYZ formula to generate sRGB outputs. This converts the scene to the familiar rusty orange color for dust and soil - here's a comparison on an early mission sundial snapshot. Final step is then conversion to default sRGB colorspace for web output.

The simple linear interpolation is less sophisticated than a polynomial fit if you have more than 3 filters, but I found it perfectly adequate if the starting filter set is already pretty close to "red", "green" and "blue" colors as is the case with L456. It does a lousy job with L257 stuff on the other hand because the surface spectra is not that flat.

Lots of outstanding images of a remarkable feature! Given how it's standing up, wonder if Oppy could break off a piece by driving over it, to get a fresh surface?

John

The 'true colour' thing is really difficult. For a start everyone's looking at different monitors which represent the colours differently. Just walk past a television shop to verify this. Of course that goes also for the persons creating the images. But there is another factor that I think is even more significant, and that is the adjustment the human eye makes to the ambient colour scheme. This acts to compensate for an overall colour bias in a scene, moderating the perceived colours.

Gordan already knows how much I respect his judgment in these matters so I hope he won't mind me saying that actually I find Stu's colours (as viewed on my monitor) more 'believable' in this case.

I look at these images on an ~sRGB monitor and one that displays ~aRGB and still Gordan's colors (and overall contrast, etc...) look more like the ones on the Pancam site..

I like it! In the past I've advocated deliberately using the wheels to knock things over and look underneath but I never got a seconder for the proposal. Let's see what happens this time.

Yeah. The images on the pancam site don't look 'right ' to me either. I think that's because when I view them I'm not surrounded by 4 pi steradians of Martian landscape. Rather I'm surrounded by a room full of Earthly ambient light. To someone with eyes adjusted to this the Martian scene may indeed look strongly orange-brown. But if you actually went to Mars your eyes would compensate and the perceived colours would shift towards the neutral. I think that's what Stu is trying to represent. (If not no doubt he will correct me.)

somehow (based on some readings; the reproduction of colour by Hunt), I was thinking the colors look oversaturated in photographs just because of the fact photographs are confined to 'a box' surrounded by other light, contrast, etc.. but that in reality, the scene, if observed in real time 'in situ', is just as vivid, only it doesn't strike as being so vivid because it's 'all around you' and adaptation to pure white has alot to do with it. I hope Gordan or others could help me out/correct me on this if I'm wrong.

I don't want to derail this thread any more by getting into another discussion on color which won't really change anyone's mind so this will be my last comment. I never claimed that particular image is really "true" color. The way Pancam DN encoding works, how the raw stretch is applied to both darks and brights in each image, the huge amount of dust on the lens, often significantly varying surface lighting conditions and lacking any stable color reference point in the frame, it's really impossible to claim any accuracy using the raw images. I'm not terribly satisfied with how that image turned out, actually.

Nigel has a point about adaptation to the general hue of the scene, but that doesn't mean color variation is lost to the eye as well. If anything, it should be more noticeable once the overall hue "bias" is removed. Yet I often see images posted where people focus on the overall hue and in the process accidentally destroy color variation in the scene, making the product almost look like colorized monochrome.