mz said:
It would be quite a coincidence if the plane was passing directly overhead. So some squishing is more likely happen. Also shear distortion, maybe in two axes.
Further, if it's flying at 300 m/s, then 10 km takes 30 seconds. There's time to take photos with varying levels of perspective distortion.
We have two sets of images taken at approximately the same location from two different cameras. We'll call Steve's camera the S camera and Dean's camera the D camera.
Most of the attention has been on one of the photos from the D camera. Steve and Dean have released more than one image from each camera. S camera images contain some EXIF metadata about the camera and image.
Images from both cameras show pin cushion distortion that is typical of a telephoto lens that is zoom to near it's limits. You can seem some of the effects of this in each of the images, this is highlighted in one of the images below. Note that the actual effects of this type of distortion are across the whole lens. Pincusion distortion will affect the image little in the center, increasing dramatically as you move out from the center. Pincushion distortion will make straight lines or edges appear concave.
D images do appear to show high order distortions distributed unevenly across the lens. These kinds of distortions can become more invasive at high levels of magnification with a telephoto lens. This is again very common even in high end cameras and telescopes, and these types of distortions are one of the reasons that reconnaisance satellites use ground calibration targets. It is difficult to quanitfy and map the higher order distortions in Dean's images without a calibration target image. Identifying discrete areas of high order distortion within the D images is especially challenging given that the majority of the image is featureless and full of noise.
Both the pincushion distortion and the high order distortions are caused by the geometric of the light path - the lenses. For that reason they are generally classified as geometric distortions.
That said, there do appear to be a number of consistent visual artifacts from the D camera in the released images. These are highlighted below as lens artifacts, though these are only the most easily identifable instances. These appear to areas on the lens or mirror that have accumulated dirt or partially opaque matter. The distribution across the lens follows a pattern suggesting the leftovers of fluid droplets across the curved surface of the lens.
Images from the S camera do not have these visual artifacts. The S camera was used to photograph the same object that has been the subject of so much attention, but those images do not appear to show the same concave edges. Instead it seems to show a straight, or slightly angled trailing edge and some degree of specular reflection from the leading edge. Other photos in the series show a distinct specular reflection from the leading edge of each of the three objects.
You can read a little more about these kinds of distortion here (this is fairly digestible):
http://scien.stanford.edu/pages/labsite/2007/psych221/projects/07/geometric_distortion/project.htm
...Or visit your local physicist.
Adobe actually now has a whole suite of integrated tools for dealing with the geometric distortion caused by lenses. I can't speak to wether these are worth trying.