Ferenc Éger schreef op vr 18-12-2020 om 02:50 [+0100]:

Nice work, I agree that when using a LED matrix array it corresponds with handling each point as independent light source (this contrary to the use of micro-mirrors where off-axis would still be possible). It thus corresponds to optimization of the optics for what in lithography is called isolated features. What does stay the same is that optimization of the lens NA is a trade-off between DOF (depth-of-focus) and resolution + sensitivity to light intensity variation.

The light intensity variation is caused by difference in intensity between different LEDs, the variation in intensity of a single LED over time and the non-uniform intensity over the field of the lens system. The first could in theory be corrected by varying the on-time of different LEDs although then the resolution in time has to be higher than the total 'amount' of the light in the minimum period.

The focus variation is determined by the quality of your leveling system, the flatness of the top surface you want to image on and the non-uniformity of focus over the field of the lens system.

In lithography this is verified experimentally by looking at so called focus-exposure (FE) matrices. This optimization can be done on a built system although money can be saved by elimating the need for high-NA from the start. The cost of lens sytem increases with increasing field size and increasing NA.

greets,

Staf.