Clear collector lens vs opal/ground lens for light source?

DaveBH · Post by **DaveBH** » Thu Dec 27, 2012 12:04 pm

I'm finding I get a much clearer image with an 'opal' (actually, with the convex surface 'ground') glass collector lens than with just a clear lens, though to my understanding this was/is a low-tech solution found in simpler 'non-Kohler' microscopes.

With the clear lens, my (halogen) filament is focused (properly, as I understand it) somewhere near the back of the objective. However, without the diffusion caused by having a matt surface in the light path, the image suffers from what I can only describe as mild 'glare', or a sort of over-bright distortion at the edges of objects. (I have wondered whether what I'm seeing is 'rising heat' distortion from the halogen bulb being directly under the stage. It does get warm, but not what I would call hot).

I think I've read where Charles Krebs found something similar, to the effect that 'somewhat less than ideal' Kohler illumination (having a diffusing element in the light path), had, in practice, some advantages. (Please correct me if I understood this wrongly, and in that case, my apologies to Charles).

I would be interested in any comments on this - thanks!

DaveBH

rjlittlefield · Post by **rjlittlefield** » Thu Dec 27, 2012 2:03 pm

From a theoretical standpoint, you should get the cleanest image when the objective aperture is uniformly filled with light. If the lamp filament is sized and shaped so that it fills much less than the whole aperture, this may be responsible for some of what you're seeing.

If the bulb has been replaced, check to be sure it's the same type that the manufacturer intended.

--Rik

rjlittlefield · Post by **rjlittlefield** » Fri Dec 28, 2012 12:23 am

Dave,

I've been working on a reply to your post from this afternoon. The post itself now seems to have disappeared(?), but I hope the following information is of some help anyway.

DaveBH wrote:I'm still curious about how you (and Charles) see the role/use of a diffuser in the light (source) path. Is it merely an imperfect cure-all' for an non-ideal Kohler setup?

I have a feeling this is going to prompt some vigorous discussion, but I think a good case can be made that Köhler's setup was invented as an imperfect cure-all for non-ideal light sources.

As I see it, Köhler's main contributions were to 1) provide a brighter image from the same light source, and 2) provide perfectly featureless illumination in the subject plane regardless of how highly structured the actual light source is. Contribution #1 comes from no light losses due to diffusion, and contribution #2 comes from focusing the light source at the aperture of the objective.

But there's a catch. The fact that Köhler focuses the light source on the aperture of the objective means that if the light source has strong structure or the focusing is imperfect, then you don't get full benefit from the objective. It's basically the same as you had inked out whatever parts of the aperture are not illuminated and stained whatever parts of the aperture show color fringes. As a result, one source that I respect a lot says that

The basic problem is that this method requires the use of a rather elaborate light source.
...
What is required to implement strict Kohler illumination is a light source with the following special characteristics:
(a) A compact-filament lamp, preferably one that is "centerable" to the optical axis of the light source. This is to minimize subsequent distortion of the filament image.
(b) A well corrected optical system (not just a simple "condenser" lens), which can be focused. This is required to project an accurate image of the lamp filament into the front aperture ("front focal plane") of the microscope condenser.

The same source notes that

For either method[critical vs Kohler], improved results with high-NA objectives may be obtained by replacing the basic Abbe condenser with one that is more fully-corrected -- and this is the subject of the following two sections on Condensers...

The source I'm quoting here is D. J. Jackson in "Better Microscopy / Volume 1", pages 48-51 on Kohler vs Critical vs "simplified Kohler". If you don't already have a copy of this book, I strongly recommend picking one up. His several works are quite affordable and are packed with information.

Adding one more summary comment from Jackson (page 48 ),

The underlying problem with all this is simply that the Kohler method is now become "oversold," to the exclusion of some viable alternatives.

Bottom line, I'm not surprised that you're getting better results with the diffuser and I am not bothered by its presence.

DaveBH wrote:Does an LED light source, with or without a diffuser, perhaps make the Kohler theory something of a moot point?

I don't think so. Again, the point of Kohler (and diffusion) is to enable a structured light source to produce uniform illumination in the plane of the subject. Periodically it occurs to me that an LED die might be a sufficiently homogenous emitter that it could be used with true critical illumination, where the light source is focused into the plane of the subject. However, I just now shot the following image of an LED from some strip lighting that I recently bought for use as under-counter lighting. Sad to say, at least for this LED my hopes for homogeneity were dashed:

--Rik

Edited to add: more information about the LED image can be found HERE.

DaveBH · Post by **DaveBH** » Fri Dec 28, 2012 2:18 am

Rik, thanks very much for your thoughtful and informative reply.

My apologies for 'disappearing' the post! I'd gone though so many turn-arounds in my thinking on this that it no longer made sense to me, or posed any relevant questions. I intended writing a new one, and was cramming in some reading when you beat me to it 8>)

So would you say that following 'strict' Kohler theory has become more an expected principle of design, than an absolute necessity?

And my simple diffuser (why is it shaped like a lens(!), and possibly a few more watts, might be all I need. Sounds remarkably like the K.I.S.S. principle, which should make me very happy - yet I can't help feeling a bit bad that I still don't quite understand Kohler... 8>/ Never mind, I'll keep at it... and will probably have many more questions.

thanks again Rik

DaveBH

rjlittlefield · Post by **rjlittlefield** » Fri Dec 28, 2012 12:55 pm

So would you say that following 'strict' Kohler theory has become more an expected principle of design, than an absolute necessity?

In my view having Köhler is definitely not an absolute necessity. I'm not even sure it's an "expected principle". Frequently I get the feeling that it's an advertising bullet -- that if you're selling a microscope you have to say Köhler even if the optics really aren't.

On the other hand, I definitely think it's a good thing to understand Köhler. If nothing else, understanding Köhler helps to also understand other systems that are not strictly Köhler but nonetheless work very well.

--Rik

DaveBH · Post by **DaveBH** » Sat Dec 29, 2012 12:51 pm

Rik, I might just understand Kohler a little bit better now... After experimenting with different light-source lenses, I managed to get the filament image focused simultaneously at the plane of the condenser diaphragm and at the rear of the objective. With that done, the earlier 'distortion' was gone.
I'm guessing there was some kind of interference going on between the specimen and light source planes of focus(?)

Light intensity, swapping lens for diffuser, seems about the same. However the diffuser gives a 'softer' and more even background, without noticeably degrading image resolution (though this will need further checking).

Must get on with the camera and flash set-up 8>)

cheers
DaveBH

rjlittlefield · Post by **rjlittlefield** » Sat Dec 29, 2012 1:34 pm

DaveBH wrote:I managed to get the filament image focused simultaneously at the plane of the condenser diaphragm and at the rear of the objective.

When the condenser is working perfectly, its aperture diaphragm will be focused at the aperture of the objective (usually described as "rear of the objective"). If you then focus the filament image on the condenser diaphragm, it will automatically be focused at the objective aperture also. This is part of the "conjugate planes" concept [ref].

Light intensity, swapping lens for diffuser, seems about the same.

It's interesting that the light intensity seems about the same. The systems I'm used to have mirrors and lenses that send most of the light directly into the condenser when no diffusion is present. Adding diffusion then cuts the intensity by causing much of that light to get diverted onto paths that get lost. I'm having trouble imagining how that could not happen. Where is your diffuser placed, with respect to the lamp?

Must get on with the camera and flash set-up

I expect you'll have an easier time adding flash with the diffuser in place. Without the diffuser, you'll have to choose whether to focus the flash or the filament, and in either case what you see with one will not be the same as you see with the other. With the diffuser, they should be the same or at least much closer.

--Rik

DaveBH · Post by **DaveBH** » Sat Dec 29, 2012 2:36 pm

rjlittlefield wrote: ... It's interesting that the light intensity seems about the same. The systems I'm used to have mirrors and lenses that send most of the light directly into the condenser when no diffusion is present. Adding diffusion then cuts the intensity by causing much of that light to get diverted onto paths that get lost. I'm having trouble imagining how that could not happen. Where is your diffuser placed, with respect to the lamp?
--Rik

The lens was positioned over the blue filter,which sits in the top of the lamp housing. With no lens, the diffuser is underneath the blue filter, it's convex side then being very close to the halogen bulb. This 'diffuser' has a large convex side, which is frosted/ground, but is flat and polished on the other side. I'm not sure why it's lens-shaped(?) In it's original application in the same lamp housing, the convex/frosted side was facing up. It ended up facing downwards during the trial and error process, and was left that way 8>) So much for my 'scientific' approach(!)

Maybe the closeness of the diffuser to the bulb has something to do with it being relatively efficient. I don't know why it's lens shaped at all, although come to think of it, having the polished side next to the bulb rather than the frosted side might make more sense. Is there a theory that might fit this(?)

DaveBH

rjlittlefield · Post by **rjlittlefield** » Sat Dec 29, 2012 2:53 pm

Maybe the closeness of the diffuser to the bulb has something to do with it being relatively efficient.

Could be. I haven't worked through the details, but in general it seems like the only way that adding a diffuser can be a wash is if it ends up redirecting into the condenser as much light as it's redirecting away from it. Being close to the lamp, surrounded by light bouncing all around the lamp chamber, would put it in the best position to do that.

I don't know why it's lens-shaped at all(?)

Well, a perfect diffuser at a constant distance from a nondirectional light source will be equally bright everywhere. That requires a spherical diffuser. Contrast with a planar diffuser, which will be brighter wherever it's closest to the source. Maybe this is the effect they were going after.

--Rik

DaveBH · Post by **DaveBH** » Sat Dec 29, 2012 3:09 pm

Yes, a 'spherical diffuser' over/around a 'point source' makes sense. I'd better consider turning it right-side-up again(!) but closer to the bulb (or LED).

DaveBH

rjlittlefield · Post by **rjlittlefield** » Sat Dec 29, 2012 4:01 pm

Hhmm... If it's closer, then it's not spherical and centered so it's not going to be uniform. Given that what you have works, it's not clear to me that this is worth spending time on.

--Rik

DaveBH · Post by **DaveBH** » Sat Dec 29, 2012 4:33 pm

Ha! I'm wondering now whether the whole thing is worthwhile! But I'm having fun, and learning (even if it's 'what not to do' 8>)

DaveBH

Charles Krebs · Post by **Charles Krebs** » Sat Dec 29, 2012 5:49 pm

Dave and Rik,

Arriving a little late to the party here.

It would be nice to have some idea of the microscope, and the components used in the light path of Dave's microscope. "Stock" components and set-up, or highly "customized"? Without this I can only add a few generic thoughts to what Rik has already mentioned.

Practically no microscope made in the past 30 years or more offers bulb position adjustments with the basic in-base lighting. Pretty much everything is described as "pre-centered". Whether this is good or bad depends on your perspective I suppose. It certainly eliminates some considerable and problematic misalignment by folks who do not understand, or do want to be bothered with a careful alignment procedure. Those who don't like it say that it could (?) result in bulbs that are sometimes not centered absolutely perfectly, which could cause some unevenness in illumination. This lack of bulb adjustment seems to bother some of the old-timers (term used very respectfully

) who became used to having it. So you will sometimes hear them refer to the older scopes as have "true Koehler" or "real Koehler". Most of the microscope with pre-centered bulbs do incorporate some diffusion someplace. Sometimes it is a separate "filter" or occasionally the collector lens itself will have a "frosted" surface. My understanding is that the reason for this is twofold... it evens out illumination caused by a slight misalignment, and it breaks up the structure of the bulb filament so that it is not apparent in the image. (This is a bigger issue with lower power objectives. Having the image of the bulb filament properly magnified, positioned and focused also minimizes problems in this regard.) I doubt this diffusion has any negative effect on the final image quality. But the image of the bulb filament should be centered and focused on the condenser aperture plane. So if you do have bulb adjustments, the diffusion should be removed (where possible) when you are adjusting the bulb position and focus to make the adjustment easier. If you have a "custom" lighting arrangement or components, you want to be sure that the collector lens and any other optics involved between the bulb and condenser actually do form an image of the filament onto the condenser diaphragm, and that the image is large enough to completely fill (or very nearly so) the condenser aperture when it is fully opened. A filament/optical mismatch could cause a problem here. You can close the condenser aperture blades and place a mirror to the side and under the condenser, angled so that you can see this filament image on the condenser aperture blades as you make those adjustments.

These days Koehler adjustments consist of centering the field diaphragm in the field of view, and focusing the condenser up/down until the field diaphragm blades appear in focus simultaneously with a subject on a slide.

Glare or flare is controlled by properly setting the two adjustable diaphragms. The field diaphragm should be closed down to a point where it is just outside the field of view. If the condenser is capable of providing a cone of light larger than the NA of the objective used, the condenser diaphragm should be closed down at least no larger than the full NA of the the objective. Most of the time you will probably go a bit smaller, 70-80% of the full NA, to get the best compromise between contrast, resolution and DOF.

DaveBH · Post by **DaveBH** » Sat Dec 29, 2012 6:43 pm

Hi Charles - Thanks for your input. There's some pics of the beast here:

http://www.photomacrography.net/forum/v ... 367#120367

I'm working on it as we speak - was thinking to invite some critters in for a visit over the weekend 8>)

Back to it ... will check in later.

DaveBH