All lenses or lens systems are quite accurate when the light is passed through them on axis or close to the axis. For this reason an abbe condenser and an achromat condenser can perform quite similarly when the image is viewed within a section of the center of the field. With non plan objectives, and those with poorer off axis corrections, the inability of a two lens condenser to correct for off axis aberrations isn't all that apparent because the aberrations of the objective itself are more evident. When microscopes were fitted with standard achromats , an abbe condenser was fairly acceptable. It was only when the same microscope was fitted apochromats, that more stringent recommendations for more highly corrected condensers were voiced, because the off axis aberrations of the condenser could cause the off axis performance of the objective to suffer. There might also be a need for a higher N.A. of condenser too but that is a separate issue. Any condenser design can be built to whatever N.A. is required, it is just that over time certain norms have evolved.
With the evolution of plan and wider field optics, better off axis performance of condensers became required, even when used with moderate N.A. achromats . A conventional 2 lens abbe will initiate rather severe aberrations off axis and especially at the perimeter, as the field increases. Defocusing the illumination beam helps somewhat but ultimately better corrections at the exterior of the illumination beam are necessary for the condenser to support well corrected plan optics accurately.
A solution was to add a third and sometimes a fourth lens, which preferentially corrects the beam off axis; usually referred to as an aplanat. Aplanats were used from the first w.w. on , usually in connection with achromatic condensers but sometimes as an amendment to a simple abbe condenser too.
More recently, some companies used an aspheric bottom lens in in place of the standard spherical lens. Known usually as abbe aspherics, the bell curve shaped bottom lens goes a long way towards fixing those distortions that plague an abbe's off axis performance. In overall performance, an abbe aspheric can rival an achromat aplanat of the same N.A.
A further idea, which goes back some time was to simply make the lenses of greater diameter. In this way, the "cleaner" part of the illumination beam ; the portion passing more central in the lens is wider and therefore more coherent. PZO for one, did this with a simple 2 lens condenser, which passes a wider aberration free beam to the objective.
Much later AO made an abbe aspheric with very wide lenses, which when oiled provides surprising peripheral performance.
Achromatic aplanats continue to be the standard and are usually offered in N.As. of a higher number than smpler condensers, in order to support better corrected objectives. I have never seen an achromat aspheric but such could exist.
The performance difference of various condensers is critically matched to the objective being used. An uncoated, unoiled, 2 lens abbe is going to reduce the performance of a 1.4 N.A. planapo a lot more, than it would a 1.25 achromat. An oiled 1.4 N.A. achromat aplanat isn't going to take the 1.25 achromat above itself but it will maximize it's potential, something an oiled 1.25 abbe won't likely do.
Are the differences startling? In some cases, with samples that have lots of difficult to resolve details; yes. In other cases, not so much but one of the overall effects of using a poorly corrected condenser isn't so much the loss of resolution, it is the blurring of margins and distortion in the depth of field. Diffraction can become prominent. Many people look towards the center of the field and crop photos to show the performance of a condenser but there is coherent condensation of the light mostly at the center . It's increasingly off axis and more towards the periphery, where a condenser's defects more easily show up. Well corrected plan optics demand more of a condenser.
And then there is the ongoing debate about whether a dry condenser completely reduces the N.A. of the objective to the level of the condenser , like a limiter switch. You can find numerous , seemingly qualified comments to support this theory. Another was copied over in this thread. The Rayleigh criterion for resolution says, this is not so. The objective N.A. will be reduced by a factor but not limited to the N.A. of the condenser because the objective N.A. is entered in as a separate value, that value being variable based on the choice of the objective not based on the choice of the condenser.
However, here is a little test one can use to prove this to themselves, if you have a high N.A. objective( 1.25 or over would be good), that has an iris diaphragm. Using the Rayleigh criterion for resolution : R = 1.22lambda/N.A.objective+N.A.condenser the objective N.A. can be changed to whatever one chooses. So, if the objective has an iris diaphragm and a max. 1.4 N.A., and the condenser is .90 N.A., using 500 nm light, the limit of resolution will be
265.2 nm, if the iris diaphragm is wide open.
If you close down the iris to .90 , the limit of resolution will be 338.9 nm. Most people can probably see this difference when viewing fine structures. If the objective N.A. were to be limited by the condenser N.A. there would be no difference , when the iris was closed.
Take your microscope , put on a fairly hard to resolve diatom sample or whatever you choose that is difficult, and using your dry condenser, and an oiled objective, proceed to close the objective's iris while viewing. If it has the N.A. marked, you should be able to read it at the point where you notice a resolution drop and when the iris is closed to .90. What you will find is that it takes a while for the resolution to fall off. If the objective were already limited to .90 or .95, whatever your condenser is, the resolution would begin to decline almost immediately.
Another way to do this, if you do not have an objective with an iris, is to take your chosen slide and compare an oiled objective with a high ( 1.25 or over) N.A. used with a .90 condenser to a .90 objective from the same series used with a .90 condenser. You will find that the resolution with the oiled objective to be superior to that of the .90 objective. Obviously this is dependent on having this sort of objective pair but in some cases people have them.
It is pretty apparent from many angles, that a condenser of a lower N.A. than that of the objective reduces the effective N.A. of the objective but doesn't limit it to that of the condenser. N.A.