QUOTE (elakdawalla @ Mar 3 2015, 07:04 PM)
![*](http://www.unmannedspaceflight.com/style_images/ip.boardpr/post_snapback.gif)
I tried to match PIA18920 animation frame 1 to PIA19185 with stretching followed by rotation, but that isn't enough to make the two frames match.
Agreed, Emily, I also tried that. This makes it look like more was done than just rotation and vertical stretch - some weird geometrical distortion was also done.
I wonder if it might be that the two frames are not actually the same, but separated by very close to one Ceres rotation. That might explain some apparent geometrical distortion, but still would not explain the vertical stretch.
QUOTE (Gerald @ Mar 3 2015, 06:33 PM)
![*](http://www.unmannedspaceflight.com/style_images/ip.boardpr/post_snapback.gif)
The mask is probably used to mask noise/dark frame outside the shape of Ceres.
Respective shape: I needed to adjust the diameter ratios to 920/975.
Images may be pincushion distorted a bit, but without knowing the position and rotation of the images, that's hard to infere.
But the noise is very low, as you can see by comparing masked and unmasked black. And a levels adjustment could have dealt with the small noise level more simply.
As for pincushion distortion, we don't have the full frames, but if, as I'd expect, Ceres is roughly near the centre of the frames, then I'd expect negligible distortion for a 222 pixel diameter region.