You are right about the on-axis performance, aspheres are corrected for this and with a resonable quality performe quite well. The off-axis performance is not at all comparable to microscope objectives. Having a plano-convex lens there is no degree of freedom left to correct for field aberrations. Using the second surface and adding a curvature gives you some correction, but still we are only talking about maybe 30µm FOV, which is way lower than for an objective. There is a reason why there are 10 or more lenses in such objectives.

Designing  a new objective one could save a few lenses by adding an asphere or two.

At first, aspherical lenses not so cheap as spherical. At second, 2-3 spherical lenses can be equal one aspherical lenses. It means if you know how calculate simple coma-corrected objective (in zemax for example), you can make it without aspherical lens. So if you wanna design microscopic objective you can decrease number of elements (necessary for correction aberration) by using aspherical lenses, but you always should make aberration calculation for determination of surface parameters of your lens. Simply put, it not easy to take some random lens and put it instead cost commercial objective. But if you requirement for aberration is not high you can try!)

You are right about the on-axis performance, aspheres are corrected for this and with a resonable quality performe quite well. The off-axis performance is not at all comparable to microscope objectives. Having a plano-convex lens there is no degree of freedom left to correct for field aberrations. Using the second surface and adding a curvature gives you some correction, but still we are only talking about maybe 30µm FOV, which is way lower than for an objective. There is a reason why there are 10 or more lenses in such objectives.

Designing  a new objective one could save a few lenses by adding an asphere or two.

Michael,    Given microscope objective and aspheric lens (singlet) being aplanatic (well corrected for spherical aberration and coma), I have confirmed why microscope objective is superior in off axis performance as Ulrike pointed out.  Please see attached ppt file, where four element monochromatic microscope objective is evaluated (Fulcher discussed about the design).   As you will see in the third page, the design is well corrected for not only SA3 and CMA3 but also PTZ3 (Petzval sum).   Petzval sum contributes field curavature, which is primary off axis aberration.   It can be reduced using a thick meniscus lens.  Two meniscus lenses are included in the design in addtion to two thin lenses in front.  Higher NA micsroscope has more.    However, aspheric surface does not influence Petzval sum at all.  

Hello,

I would ask what precisely application is? I am not a specialist of microscopy but do a lot of experiments with relatively high magnification down to ~10um resolution. There are important factors - group velocity dispersion; magnification (single set of elements can be used for imaging into camera or other optical system-device-stage); specific wavelength out of objective design; waist inside objective what causes damage of optics. Working with fs pulses there is no possibility to use real objectives as they are thick - dispersion is enormous. Therefore objectives are out of use in deep UV (extremely pricy), mid and far IR (out of design range and loses due to improper coatings), for really short fs pulses (elongation due to dispersion) and high intensity pulses (self-focusing, self-modulation and damage). Everything is true for what I am doing. Compromise - few thin lenses. Obtained quality is definitely lower but final result of the experiment is always better.

So, what is the reason to reject "in factory designed" objective? When it is known - solution can be found :)

Sincerely,