Shack-Hartmann image of an aspherical lens with fl=50 mm taken with OMI lenslet arrays with fl=22 mm and fl=11 mm.



We explain the importance of using Shack-Hartmann lenslet arrays with long focal length using a simulation computed with our Sensoft simulation package.
10 nm of Coma are added to a perfect wave (Standard Zernike coefficients).
The spot displacements on the normalized pupil are computed for two different cases.
Assuming our standard lenslet array with focal length = 22 mm.
Assuming a generic array with focal length = 5 mm.
The result shows that the length of the displacement is proportional to the focal length. It is therefore easier to detect Coma when fl = 22 mm.
Spot displacements due to Coma aberration (PV = 48 nm) that one can expect using Shack-Hartmann lenslet array with focal length fl = 22 mm (our standard choice, left) and fl = 5 mm (right).

An additional simulation is performed by adding 0.2 micron random noise to the spot displacements for the two cases.
In the presence of modest noise, with fl=5 mm the Coma displacements disappear inside the noise, while for fl=22 mm the Coma pattern is still visible in the displacements.
Adding random noise of max 0.2 um. For lenslet array with focal length fl = 5 mm (right) the displacements are inside the noise while for fl = 22 mm (left) the Coma pattern is still visible in the displacements.

Deviation on camera chip is proportional to focal length.
Deviation on camera chip is proportional to focal length.
Because of that, we improve the sampling by a factor ~3X using sophisticated mathematical interpolation.
Gives unstable and unreliable results since it is very poor in resolving aberrations especially in presence of noise and also causes thermal instability as the lenslet array is too close to the camera chip and often INSIDE the camera body.