Originally posted by KSPigpen
It seems to me as though if it were artifacts of compression, they would affect the border of the craft in a more elliptical manner, corresponding to
the the relative shape of the craft....rather than a clear rectangular border...
But .jpg compression works with square pixels and compresses them in square groups. The compression algorithm will look at a group of pixels.
Depending on the level of compression, that group could be big or small. From my understanding, the algorithm looks for similar coloured pixels and
makes them the same colour, so that there is less colour information to deal with, thus a smaller file size.
The "UFO" photo looks like it was being compressed in groups of 8x8 pixels. The sky is rather easy for the compression algorithm to deal with as
there isn't too much colour variance to deal with, just different shades of blue. But when it comes to the "UFO" or the tree branches, it can't
compress it as smoothly as there is more colour information to deal with. So those groups won't be compressed in the same way and the groups of 8x8
pixels that are around the "UFO" or trees won't match those that are only of the sky.
Hopefully I explained that correctly and clearly.
but to argue the other side, why would someone do a rectangular cut and paste in PS or Gimp, or whatever else when it would be just as easy to
get a cleaner border on the inserted craft? Laziness?
Exactly. When I or my wife edit photos, there aren't obvious signs of editing left behind. You'd have to be incredibly lazy otherwise.
[edit on 21-9-2009 by jra]