.hans wrote: ↑Fri Jul 30, 2021 5:42 pmOn the image side, but on the illumination side the condenser images the aperture iris at higher NA when the illuminated field is large, I believe? For example MicroStar IV system goes up to 5 mm with ~10 mm EFL condenser at the normal height, so NA 0.25? (2.5 mm / 10 mm)
Yes/no/I don't know? For Brightfield, I'd say that the field stop at the eyepiece/camera adapter is the one ruling them all, so the NA is restricted. For scattered light (darkfield etc) I fear that you're right, the field gets larger, so the NA, and spherical aberration nastier.
-
That's getting above my paygrade... guesses:hans wrote: ↑Fri Jul 30, 2021 5:42 pmRegarding telecentricity, this is equivalent to saying that the limiting aperture stops are located at the condenser FFP and objective BFP as usual? Not sure it makes sense to say eyepieces themselves are inherently telecentric, rather projection of the intermediate image near infinity by the eyepiece is telecentric only because the objective BFP is the limiting aperture stop in the system? Actually I have read that newer camera lenses designed for digital (particularly mirrorless systems) are often telecentric on image side so maybe could work as condenser with modest NA?
Yes the condenser FFP - objective BFP should match, but that must be implemented explicitly in the design, doesn't happen naturally. If both are telecentric (same "zero aperture convergence" ) then they match; but can made to match also if they're complementary convergent-divergent; some manufacturer may have chosen the latter road, making mix & match of condenser-objective a gamble, like with compensating eyepieces.
There is no "natural" place for the aperture plane, exactly halfway from the field; you can put the aperture where you want, between field planes.
You can tweak LWD objectives and eyepieces to be pericentric (hypercentric) by imposing an aperture at the "wrong" place with a relay lens:
https://wordpress.com/post/patta1072853 ... ss.com/620
Or tweak high NA objectives to become fisheyes (with defocused Bertrand lens).
https://wordpress.com/post/patta1072853 ... ss.com/445
But, images are not so good with those tweakings.
Because the designers, choose a place for the aperture, then optimize the optics around this position. Then,if we move the aperture away from this position, the lens doesn't work well, extra aberrations chime in. So, eyepieces are optimized for (approximately) telecentric; yes with the objective BFP being the working aperture. But if I place the aperture nearer, image from the eyepiece gets worse.
Camera lenses, are never exactly telecentric image-side; but, true, they need to be less divergent than old film lenses, because digital sensor have narrow "acceptance angle", they won't record light that is coming from a grazing angle, better if it comes almost perpendicularly.
DSLR lenses needed a "back focal length" or "flange distance", because of the mirror; the lens need to be away from the sensor. That forced to use Retrofocus for wide angle lenses, which has the added benefit of projecting light on the sensor from not-so-grazing angles (but still far from telecentric).
With mirrorless, the need for Retrofocus disappeared; so people started experimenting with old wide angle lenses designed for film rangefinder (aka film mirrolress), that do not have retrofocus.
But, then, horrible images; because the sensor is not sensitive to the narrow-angle rays those lenses were projecting. So, the need for telecentricity came back. This time with "telecentric" name (while before, it was called "back focal length").
Manufacturers today are scrambling to make sensors with wider acceptance angle; and mirrorless are likely to accept less-telecentric lenses than old DSLR.
Many confused words, but a couple of images may explain the above very easily