The model was based entirely on the MGS photo - I didn't even have a motive originally to compare it to the Viking photo. I created a black and white depth map over the picture, based on the depth suggested by the shadows and apparent contours. I then imported it as contour information into a 3D modelling software, where the resulting 3D object was lit roughly from the correct angle. Next I corrected the depthmap many times, until the 3D shape somewhat resembled the photo (naturally lacking all detailed textures and smaller bumps and grooves):

Note that the MGS shot was taken clearly from the left side, instead of almost directly from above, like the Viking image. This is one of the main reasons why the two images look so different.

When I had already made the model, I could not miss the opportunity to check out how it would look if it was lit according to the old Viking image. Considering I had not used this at all as reference for the model, the rendered image was surprisingly accurate - actually even more so than with the MGS image!

Remember: This is EXACTLY the same shape - just the lighting and viewing angle was changed! (I must confess the regular triangle shaped shadow is a bit puzzling though....)

I also used the MGS image as a texture map for the model, then rendered an animation which shows the formation in rotation (top of this page)

continue to "Any new suggestions?"