I was wondering if, for creating a well collimated light beam, the spatial coherence play a role. In particular, I understand that a laser beam can be well-collimated with respect to light coming from an incoherent source (such as an LED), but it is not clear to me if this is due only to the much lower divergence of the laser beam, or also due to its spatial coherence.
Some examples may help to understand this. A point light source emits in all directions, if a lens is placed such that the light is at the focus of the lens the emerging beam will be collimated, all rays perfectly parallel. (There is no such thing though as a perfect point or a perfect lens.) An LED emits in all directions, it is not a laser cavity. A laser has a near perfectly constructed geometry with precision mirrors placed at either end, the beam inside the laser cavity is highly collimated, when it emerges (example laser diode) the beam diverges rapidly due to diffraction, all laser pointers actually use a small lens close to the cavity (at the focus point) to collimate the beam. Spatial coherence is not a requirement of the source for collimation. A good question would be does collimation of a beam improve spatial coherence? .....but knowing the degree of coherence of a beam is complex, lots of math, and it becomes an exercise in correlation (see wikipedia).