episcopic adapter

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

viktor j nilsson
Posts: 423
Joined: Fri Mar 01, 2013 1:43 am
Location: Lund, Sweden

Re: episcopic adapter

Post by viktor j nilsson »

Pau wrote:
Mon Jul 05, 2021 9:48 am
To be more exact, the "epicondenser"* optics must focus the light source at the objective rear focal plane and fulfilling it, like in transmitted Köhler

* this is the usual name although the main condenser in this case is the objective itself
Yes, I was a little confused about Rik's explanation. I've thought about it in terms of focus and magnification, rather than angles. Or specifically that the filament should be in focus at the rear focal plane, and the collimating optics should magnify the image of the filament so that it covers the entire width of the rear aperture. Which is pretty big in a 10x 0.50 objective, and pretty small in a 100x 1.40 objective, so it's unlikely that one solution is perfect for all objectives. Rik, is this consistent with the way you think about it, just phrased in a different way?

rjlittlefield
Site Admin
Posts: 23559
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Re: episcopic adapter

Post by rjlittlefield »

Pau wrote:
Mon Jul 05, 2021 9:48 am
must focus the light source at the objective rear focal plane and fulfilling it, like in transmitted Köhler
Hhmm... I think there are some hidden assumptions in that characterization.

Consider using the sun as a light source, with a single long lens used to focus an image of the sun at the objective rear focal plane.

Now make the diameter of that lens very small, approaching a pinhole.

It seems to me that this still matches the words, "focusing the light source at the objective rear focal plane and fulfilling it".

But still, all the light entering the rear of the infinity objective will be coming from essentially one direction, and as a result all of it will be directed to a small area of the subject, approaching a single point. From the standpoint of the subject, at that point the entire objective will be evenly bright, because the rear of the objective is filled. But at other points, the entire objective will be evenly dark, because all the incoming light got directed somewhere else.

So, it seems to me there must be something in the actual requirement that specifies a sufficient angular spread for illumination entering the objective.

What have I missed, in this line of thought?

--Rik

viktor j nilsson
Posts: 423
Joined: Fri Mar 01, 2013 1:43 am
Location: Lund, Sweden

Re: episcopic adapter

Post by viktor j nilsson »

reflectedfigure5.jpg
Here's an example of a Köhler epi illumination setup to illustrate how I think about it. From http://zeiss-campus.magnet.fsu.edu/arti ... ected.html

If you reduce the size of the aperture diaphragm, you clip the image of the filament at the rear focal plane. This will reduce the NA of the illuminating beam (the back focal plane is no longer entirely filled), which will make the illumination weaker, and, I assume, reduce resolution. But the size of the illuminated field at the object plane will not be affected. But if you reduce the size of the field aperture, you will shrink the size of the illuminated field at the object plane.

Which of these aperture planes is equivalent to the pinhole in your thought experiment?

Pau
Site Admin
Posts: 6049
Joined: Wed Jan 20, 2010 8:57 am
Location: Valencia, Spain

Re: episcopic adapter

Post by Pau »

Sorry, I'm not at all sure of fully understanding your ideas, as you likely know my Maths and Optics knowledge is less than basic.

With my fluorescence epi-illuminator I can see the LED image focused at a plane that I assume is the rear focal plane in a window exactly designed for this purpose (sorry for the horrible phone picture)
led window.jpg
rjlittlefield wrote:
Mon Jul 05, 2021 10:33 am
...
So, it seems to me there must be something in the actual requirement that specifies a sufficient angular spread for illumination entering the objective.
Yes, As far as I understand it, a focused image implies angles of light but it is a very low resolution image of the light source and the angle of the light entering the objective rear plane is small
Maybe the lacking factor in your mental experiment about what the subject sees could be the objective acting as actual condenser: of course the light rays hitting the subject make angles, wider with high NA objectives, of course.
Pau

rjlittlefield
Site Admin
Posts: 23559
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Re: episcopic adapter

Post by rjlittlefield »

viktor j nilsson wrote:
Mon Jul 05, 2021 10:54 am
Which of these aperture planes is equivalent to the pinhole in your thought experiment?
Good question!

Let me see if it helps to "think out loud".

With an infinity objective, each point in the focused object field corresponds to a bundle of parallel rays behind the objective. The angle of that bundle, with respect to the optical axis, determines the location of the corresponding point in the object field. Of course the correspondence is bidirectional -- light entering the rear of the objective will follow the same path as light entering the front of the objective, just going in the other direction.

So, object side NA depends on the amount of objective rear aperture that is filled with light at any particular angle, and illuminated field size depends on the range of angles present in the illumination.

I wrote earlier that, with the pinhole'd lens, "all the light entering the rear of the infinity objective will be coming from essentially one direction". That is not quite correct. In fact, rays will be entering from a range of angles, whose lower limit corresponds to the angular width of the sun in the sky. However, at each point on the rear of the objective, all the rays will be coming from a single angle. If the lens in the illumination path is made wider (so, not a pinhole), then for each point on the rear of the objective light will be coming from a wider range of angles, and also a larger area of the objective rear plane will be illuminated by light at any particular angle.

Having talked this through, it seems to me that an iris at the lens location in the thought experiment has some aspects of both the field stop and the aperture stop in the Köhler setup. Maybe it's like an aperture placed somewhere between the two perfect positions of Köhler.

--Rik

rjlittlefield
Site Admin
Posts: 23559
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Re: episcopic adapter

Post by rjlittlefield »

OK, after a night's rumination, let me try this again.

Earlier, I wrote:
Pau wrote:
Mon Jul 05, 2021 9:48 am
must focus the light source at the objective rear focal plane and fulfilling it, like in transmitted Köhler
I think there are some hidden assumptions in that characterization.
...
it seems to me there must be something in the actual requirement that specifies a sufficient angular spread for illumination entering the objective.
So now to explain more fully...

In the Köhler configuration, the "sufficient angular spread" that I'm concerned with are the light rays that are controlled by the field diaphragm.

The Köhler configuration is special in that it precisely factors the controls into one diaphragm that adjusts the NA and a second diaphragm that controls the field, with no crossover between those controls. Using transmitted Köhler that is perfectly focused and both diaphragms precisely set, every light ray that gets through the condenser will end up contributing to the image.

In illumination systems that are not perfect Köhler, the controls are not so precisely separated and there are likely to be stray rays that get through the condenser but do not end up contributing to the image. Nonetheless such illumination works fine as long as the objective aperture is fully filled, the field is fully illuminated, and stray light is blocked or absorbed so that it does not cause flare.

In soldevilla's image at https://photomacrography.net/forum/view ... 65#p275065 , the strong vignette suggests to me that edges and corners of the field are not being illuminated.

If they are not, and his setup is Köhler, then we can say that the field diaphragm is set too small. Of course, depending on the lenses involved, it may not be possible to set the field diaphragm large enough. In the Zeiss diagram, the field diaphragm cannot be set larger than the diameters of the lenses that it sits between. If those lenses are too small, then there will be no way to illuminate the entire field.

If soldevilla's setup is not Köhler, and the edges/corners are not being illuminated, then we cannot so precisely point to a single place and call that place the problem. But we can say, in general, that the illumination is missing the light rays that would otherwise illuminate those edges/corners. In an infinity system, those missing rays correspond to very specific missing angles -- in fact the same angles that the objective would produce when imaging those edges/corners.

In some cases the problem of missing rays can be solved effectively, though without elegance, by defocusing the optics or by adding some diffusion in the illumination path. As an extreme example, the following arrangement will work just fine to illuminate the subject field at full NA -- assuming that you can control the stray light!

HighlyDiffusedTTLIllumination.gif

Does this help?

--Rik

soldevilla
Posts: 684
Joined: Thu Dec 16, 2010 2:49 pm
Location: Barcelona, more or less

Re: episcopic adapter

Post by soldevilla »

These types of threads always end up circulating towards topics that I don't understand anything about.

Obviously, my lighting is not Köhler. It is a 3W LED head, I think, which falls short anyway because it is a system that wastes an impressive amount of light. But something has suggested to me what I have tried to read ... The light beam that reaches the rear of the lens should have the same "diameter" as the light beam that the lens sends towards the camera, in order to fully illuminate the field. This, with the x4, is clearly not the case, and therefore there is an important vignetting. I have tried putting a diffuser on the illuminator outlet, but all I get is having to extend the exposure much longer, to several seconds. And the illuminated field has hardly gotten bigger. This weekend, if I find some free time, I will try to test it with a x10 and a x20.

Scarodactyl
Posts: 1617
Joined: Sat Apr 14, 2018 10:26 am

Re: episcopic adapter

Post by Scarodactyl »

I'd definitely encourage you to pick up a cheap wafer on eBay or wherever. It's not just better for testing, they also reqlly shine under episcopic illumination, so they should be more fun to work with.

Post Reply Previous topicNext topic