Nice images, James. For your Winjupos projection, how did you align the reference frame unless you had an image containing the entire lunar disk? In the absence of a full lunar disk image, I suppose you could superimpose your image panel onto an appropriately scaled simulation from the NASA libration page, and then align that to the reference in WInjupos? I bet there is some considerable leeway for error in the alignment that would really only be noticeable if you spot-check the latitude and longitude coordinates in your resulting map to LRO data. Otherwise, mistakes in the projection would probably be impossible to notice, and your version looks good.
Yes, I just scaled and rotated the full-disk output from the NASA Visualization service and then overlayed my image onto that. I used Photoshop to do the overlay with my image and it was fairly quick to do (looking at the difference blend). I think the final overlay was within a few pixels (better in some places, with combined error, rotation and horizontal and vertical translation). I measured the rotation and scale just once, then did the rotation and then the scaling (the latter took a while, since it produced a 50M pixel image). I also tweaked the rotation (by about 0.2 degrees) just before the rescaling to better match my original measurements in Photoshop.
Of course, this also means you can do this kind of projection mapping on practically any lunar image, no matter how small the field (although you are somewhat limited by the 16000 pixel-wide output from WinJUPOS, however, I think there may be a way around the latter). Also, the NASA simulation includes the position angle of the moon's north pole, so you can accurately adjust the rotation to keep north directly up (the angle was 7.9 degrees on the night that I took my image). The NASA simulations appear to go back to 2011, so lots of potential to match past images.
Below is the final overlay (reduced greatly in size to conform to the CN limits), with my shot of Gassendi on the lower left (showing up as grayscale against the color of the NASA simulation). This was then imported into WinJUPOS to produce the projection map. Following that, I went back to Photoshop with the image map produced by WinJUPOS and rotated and cropped out everything but my original image. One further thing that could be added is to use the color information from the NASA simulation to colorize a mono image. I'll have to see how that works, although that would mean combining my own data with that from NASA which seems a little bit of a "cheat."
Edited by james7ca, 14 July 2019 - 11:16 PM.