Mike Sibley Comment Posted Nov.19th, 2012, viewed 59 times
A good try, Alexey, and you're almost correct - but not quite :o)
The tree's shadow will signal the overall lighting direction, and describe the three-dimensional form of the rocks that it falls on. Your shadow mimics the shape of tree itself, which suggests the shadow is on a flat vertical surface. But it isn't...
As you shade the shadow, you need to be aware of the shapes of the rocks and to work out where the shadow would fall. Your shadow is OK in the centre of each rock, but it just continues upwards from rock to rock. In reality, the shadow would dive into every crevice and joint. See image below for a explanation. I've just tried to show your the direction of the shadows. They should of course melt into the existing cast shadows between the rocks.
The human brain seeks understanding and will look for the most obvious clues first. Once found, it will happily accept everything else as conforming to the most dominant clue - and it will find reasons to explain anomalies, such as the presence of reflected light. But it will have problems understanding a flat shadow falling on very bumpy rocks... is it the shadow or the rocks that it's misunderstanding? In the end it will give up and the viewer will move to the next image.