Radar And Mariner 10, Best possible mapping, pre-Messenger Options

[JRehling]

[...]

[peter59]

I recognize that many of you are waiting for new images of Mercury, but you can't also see old images.


Tape MVE_033 Image 091 (29 November 1973)

[tedstryk]

I have been playing around with some Mariner-10 shots. I experimented with color, but Mariner 10's camera had such a weird filter selection that, when combined with calibration issues, make things really difficult.



Here is a cleaner view of the south polar view I did a while back, both in original form and from a high pass filter. There is a larger mosaic, but since it depends on a heavy amount of reprojection, I decided to stick with the wide angle data, as Mercury's topography makes significant reprojection obvious.



Here is the 3rd encounter Discovery Rupes mosaic, with a small gapfill from stacked first-encounter images to fill the sliver between the two lower images. Also, I did a stack of the "Discovery Dome" area using third encounter images.

[JRehling]

[...]

[tedstryk]

True-color is such a relative thing....I think even Messenger stuff might go through a few variations as we have seen in color images from many early missions....the problem is the lack of a standard for subtle balancing work. By the way, if anyone downloaded the image right after I posted it, I posted the wrong file...it has now been corrected.

[JRehling]

[...]

[tedstryk]

What I am trying to say is that it might take a while before images can be calibrated to the point of providing color of the quality of, for example, Cassini images. For the other planets, we at least have good earthbased color data. Mercury images, on the other hand, are so damaged by atmospheric refraction and (in daylight) scattered sunlight that they don't provide much of a reference point. In other words, I can see the color images being produced two years from now looking very different from the ones we will see in the next few weeks.

Incidentally, Mariner 10, while having odd filters, had a great camera, but unfortunately, as has been mentioned before, the camera pointing resembles footage from an drunken camera man. That makes the images very hard to assemble, especially using color. I always wondered why Mark Robinson and his team only produced a color ratio map of a small area. After trying to work with the data, I understand.

[ugordan]

I personally think any color uncertainties in the first images will be smaller than any comparative "ground-truth" later, onboard spectrometer corrections will suggest. Mercury, after all, isn't likely to be a very spectrally complex place like Jupiter, Io or Saturn. It will probably have gentle spectral slopes and that's where even discrete filter MDIS composites can come very close in "true color" to what a more precise integration through many spectral wavelengths (approximating the actual spectral curve to a good degree) would give. Messenger carries a spectrometer, although apparently not an imaging one. MASCS will probably be good enough anyway, just like Cassini's VIMS can be used at Saturn.
The MER rovers don't have that kind of training tool and all their surface "true" color views are basically generated on the basis of being the most probable result of solving some integral equations IIRC.

I'd like to know if Cassini image products were trained on groundbased imagery (I know my first reactions to the first Saturn color shots were how different it looks than from other earthbased views I saw) because the vantage points were different and direct color calibration would be difficult due to phase angle effects etc. There is certainly variation in color results even today from CICLOPS image releases, although less than before. I'm saying that with good spectrometers onboard you don't even need groundbased observations for good color quality and this might very well turn out to be the case with Mercury.

Color me happy if Mercury doesn't turn out as spectrally grey as our own Moon. Even a brownish hue is some personality.

[tedstryk]

I think the lack of color is the problem. With color variations being so faint, balancing is made very difficult. For "natural" views, this may be only subtly visible, but it may effect the warm/cool color appearance, which is why I say there may be a lot of variation at first (it really depends just what color Mercury really is to see whether this materializes). I have never thought it too terribly important to replicate the human eye in planetary images - the RGB response of our eyes is so arbitrary for planetary targets that unless we are generating graphics for astronaut training, using other wavelengths that better highlight a world's compositional variations is a great technique. However the Mariner 10 filters set is just weird. It does have a dedicated yellow filter, but it was seldom (if ever) used. There is so much overlap between the filters used here (UV = blue, blue = green, and "minus UV" = red) that isolating three channels is quite treacherous.

On a more abstract level, I often wonder how much variation there is between different sets of eyes. I have gotten into long debates with people who are looking through my telescope and could swear it has a cool hue while others swear it has a warm hue. I have noticed that on many nights, it looks cool in my left eye and warm in my right eye. Perhaps this is what gives me my opinion on color.

[JRehling]

[...]