Illumination principle used to convert an inhomogene illumination to a homogene Illumination without too much light loss

The main components of a Koehler-Illumination are the “condensor” that maps the illumination (for example some LEDs) to the location of the second component, an Iris, often a manual Iris.
The “Collimator” maps the Iris (and therefore the images of the LEDs) to infinity. So the light will leave the collimation parallel.
Because the LEDs have a certain non-zero diameter, so have it’s image at the Iris. As the result the light from each arbitrary small point leaves the collimatior parallel, but at an Angle.
The sine of this off axis angle is called numerical aperture “NA” (in air) .
The speciality of the Koehler Illumination is, that the light of all these parallel light beams meets at (ideally) one position, markered as “best mix” in the graphics. For best results the real world object to be illuminated shall be at this best mix point.