I have given this information so many times, to so many people … it really gets old. But it is essential, basic, must-know material so here it is, one more time.
Aside from shape and tint, the position, strength and angular width of the light source causing a shadow has three distinct effects. Note that these are independent of one another. The cause of each effect has no bearing on the cause of the others.
Stretch and compression are caused by the non-linear difference between the location where a shadow strikes the receiving surface when it is close versus when it is farther away. It causes stretch (non-uniform elongation) when the light source causing that shadow is nearly parallel to the receiving surface – and compression when the light source is nearly at right angles to the receiving surface.
Think about the transition from right angle light source to parallel light source. Also think about drop shadows from different heights (as opposed to cast shadows) in those conditions. See if you can figure out why you have stretch and compression.
Fade is visible in a shadow when the primary light source (the cause of the cast shadow in question) is not much brighter than the secondary light sources in the scene. What you are really seeing is overlapping shadows, but the secondary light source shadowing outside of the cast shadow is overwhelmed by the brightness of the primary light source. When the primary light source is much brighter than the secondary light sources, fade within a shadow will be so slight relative to that main light source, that is won’t be noticeable.
This is very easily studied in your own home. Not only will you usually have multiple lights on at the same time, but light colored walls make for quite strong bounce fill (as secondary sources). Look around you and think about what you see.
Edge blur is caused by the angular width of the light source relative to the object in question (how big the light is combined with how close it is). Light coming from one side of that light source casts a different shadow from light coming from the other side of that light source and all points in between cast slightly different shadows. Thus you get edge blur from wide light sources and no edge blur from point light sources.
Once more, it’s especially easy to study edge blur indoors. The sun, because if its extreme distance, is effectively a point light source (on a clear day; in haze it smears to have some angular width).
