Mariner 7 far encounter imagery and the horrors of analog data...
For most of the far encounter (save a short bit at the end, which was all digital but with so much of the image missing as to render the frames extremely hard to use), the images were only sent in an unfiltered analogue format. Thus, while similar in pixel size to many images taken by Hubble and by other approaching spacecraft, the images tend to be indistinct. Worse, the crispness is very inconsistent. Features of equal contrast within darker colored areas of mars are less distinct than in bright areas, and features near the terminator are less distinct than those in more brightly lit part of the disk. Higher contrast features are sharper than lower contrast features. Thus, while one can get a better result than this version from JPL's website (probably an original press-release image), it is next to impossible to get it to look like a modern image.
the digital color images (I need to rework this - I was focused on reconstructing the missing parts of the image and didn't make any attempt to do even relative calibration of the color) and on doing a cleanup of all the images from the approach. Thomas Romer assembled my cleaned up images in to some neat movies that look like Nickelodeon movies from over a century ago (there is also some quality loss due to the use of the gif format) - you can view Mariner 6 and Mariner 7 by following the links. I have made several attempts to make improvements on the images, but unlike my work with the near encounter images, I have had trouble developing techniques that meshed with the spongy oddities of the far-encounter analog-only images. This version, done in 2007, was one of my most successful attempts, but it still looks off.
Recently, after the images from ESA's Rosetta spacecraft's flyby of Mars in 2007 were released, I felt the need to take another crack at this dataset. I also attempted to extrapolate from the color data Mariner 6 and 7 obtained for some areas of the planet to colorize the images. I first worked on the famous image of Phobos (the dark blip near the top of the image) passing over Mars . The result, while good considering the quality of the data, highlights a serious problem - high contrast features stand out - note how sharp Phobos itself is - while low contrast feature are lost, with the problem being inconsistent throughout the image. There were also problems with the color overlay.