Why recommend polarizer, 1/4 wave plate for Mitutoyo 1x,2x?

[Chris S.]

A small puzzle:

On page 16 of Mitutoyo’s Microscope Units and Objectives Catalog, there are footnotes on the 1x and 2x objectives' specifications:





It seems odd that Mitutoyo recommends using a polarizer with the 1x and 2x objectives. Apparently the intent is to enhance contrast (clip below is from page 12 of the catalog):



But while I consider polarizers very useful things, I've never thought of them as enhancing contrast per se (although they can seem to, by cutting certain reflections that can reduce the appearance of contrast--but this effect is not limited to low-magnification objectives). So a quick test was in order.

Here is the result--a pixel-level crop of a composite of two photographs taken with the Mitty 2x objective (of very fine print from a laser printer). One side (either left or right) is taken with a polarizer, the other side without. I flipped the right-hand image horizontally to "book-match" it with the left side, which I think makes certain comparisons easier.

Can you tell which side is polarized? I can't. If there is any increase in contrast from the polarizer, I don't see it. (To avoid introducing bias, I'll identify which side is which at the bottom of this post, rather than here.)



Technical details: I didn't use a Mitutoyo polarizer, but an Edmund Optics Techspec High Extinction linear polarizer that I've found to perform very well and indistinguishably from the Olympus polarizer (aka analyzer) that I use with RMS objectives. I strongly doubt there is any magic in the Mitutoyo polarizer that this one is lacking. The Techspec polarizer is, naturally, mounted between the objective and tube lens. (And I of course adjusted my flash by two stops to compensate for the light loss from the polarizer. And I put a curves adjustment layer on each image to get fill the unused portion of the histogram to the right and left sides; these curves were very similar for both pictures, though not quite identical. They were both fairly modest adjustments, and didn't change the overall impression.)


So a polarizing filter alone does not--at least in this test--change the output of the 2x objective. But notice that the recommendation is to use it with a 1/4 wavelength plate. This recommendation is not given for the 1x objective (which I don't have), but the following diagram probably explains why, as a 1/4 retardation plate is supplied on the front of the 1x lens.



I have no experience with 1/4 wave plates, other than the ones built into the circular polarizers I use for camera lenses. What is the retardation plate's purpose here? If it were placed behind the polarizer, it would turn a linear polarizer (if that's what the Mitutoyo polarizers are) into a circular one. I could see that being potentially useful for certain cameras that might be attached to a Mitutoyo microscope. But at least in the case of the 1x objective, it's in front. What would that do?

I don't have (or want) the Mitutoyo 1x objective, but I do have and use the rest in this series, other than the 7.5x. In case it provides a clue to the "polarizer plus 1/4 wavelength plate" recommendation, the 2x has one peccadillo compared with its higher-magnification brethren: It displays significant, though easily corrected, transverse chromatic aberration away from the center (shown here). Since objectives from various brands seem to make substantial compromises at about 2.5x and below, this is probably a hard range to design for in a system that requires parfocality with much higher magnifications. So I don't take an easily corrected aberration as much of a fault in this case.

Can anybody suggest why Mitutoyo recommends a polarizing filter and 1/4 wave plate with the 2x objective? Can you explain the effect of having the 1/4 wave plate out on the front of the lens?

Thanks!

--Chris

------------------------
The polarized image is on the left, unpolarized on the right.

[Pau]

I guess that recommendation would only be relevant for epi illumination through the lens, not for external illumination.

I think this is what Zeiss calls Antiflex:
http://www.photomacrography.net/forum/v ... t=antiflex

[rjlittlefield]

Pau's explanation sounds good to me. In that case the linear polarizer plus the quarter-wave plate combine to make a circular polarizer, which blocks out-and-back light reflected off shiny and metallic surfaces. (Google circular polarizer mirror test.)

--Rik

[Chris S.]

Thanks, Pau and Rik. I think you're right.

It was interesting, reading about Zeiss's Antiflex. If I ever get around to adding through-the-lens illumination to the Bratcam (it's been on the drawing board for some time), it might be handy to have a 1/4 wave plate available. And of course, after reading about "circular polarizer mirror test," I had to try it. The effect is every bit as dramatic as they say.

--Chris

[Joseph S. Wisniewski]

I think you're misreading slightly. I'm guessing that, for normal photography or observation, they are not recommending that you use either a 1/4 wave retarder or a polarizer.

I think what they're trying to say is that, when you need linear polarization, you should use this particular polarizer behind the objective, probably because their standard polarizer is just a bit too small (oops!) or maybe too thick to operate behind the low mag objectives without inducing aberrations (you can pile dang near anything behind high mag objectives) and when you want circular polarization for the 1x, you should use this particular retarder in front of the objective, not a normal retarder behind it (guessing thickness, again).

[DQE]

Isn't there an issue with respect to improving the contrast via polarizers if you are using highly diffused light? My rule-of-thumb theory has been that diffused light (as from a highly diffusing sheet of paper) is unlikely to have much polarization left in the reflected light.

Thus, I wouldn't expect polarizers to matter much with this specific subject.

[Chris S.]

Phil, the lighting in the test shot wasn’t actually diffused, though it may appear otherwise. It was a single, undiffused Nikon SB-800 flash placed about 12 inches from the paper subject, to camera left. The light is falling on the front of the subject, from a point about 45 degrees off the subject plane. (Pretty much as basic as lighting gets.) While paper can of course act as a strong diffuser to light that is passing through it, in this case, the light is bouncing off it.

But to your main point, I didn’t expect the polarized shot to look much different, either. I thought that Mitutoyo’s statement indicating that a polarizer is “also suitable for enhancing contrast of low-magnification objectives,” was, as written, wrong. But since a test was quick and easy to perform, I elected to test the claim before moving on. Polarization can of course work wonders for certain lighting regimes and subjects—but Mitutoyo’s claim here seemed to be that it improved certain lenses. That’s what raised my eyebrows. If I were to set up a situation in which polarization made a noticeable difference, that situation should produce that difference with any objective, low magnification or high.

Joseph, I think that if you take another read of Mitutoyo’s words, you might see it differently. They are not recommending a different polarizer for these objectives—just saying that one should use the “appropriate polarizer for the microscope used,” along with an integral or added ¼ wave plate—for these two objectives.

. . .and when you want circular polarization for the 1x, you should use this particular retarder in front of the objective, not a normal retarder behind it (guessing thickness, again).

But if the retarder is in front of a linear polarizer, the light coming out the back of the linear polarizer is still linearly polarized, right? To get circular polarization, doesn’t the retarder have to come after the linear polarizer? I certainly think so.

My sense is that Pau and Rik have nailed it. This seems to be similar to the Zeiss Antiflex, but poorly communicated by Mitutoyo. In a through-the-lens lighting scenario, the light would pass through a linear polarizer, then through the objective, through a ¼ wave retarder, bounce off the subject, return through the ¼ wave retarder, through the objective, and through the linear polarizer. I get a little fuzzy at this point, and want to try it myself, but my thought is that this arrangement is effectively a form of cross-polarization. Since X-pol is some I use a lot, and since I do intend to add epi-illumination to my rig, this may well be something I want in the bag of tricks.

I do wonder if the technique is truly limited to low-magnifications. Zeiss isn’t entirely clear on this with Antiflex. When I get around to trying it, I’ll certainly see what happens at higher magnifications.

One small oddity: My searches for the Mitutoyo “¼ wavelength plate A (02ALN370)” mentioned in the quoted literature come up with only hits for the literature I quoted it from--no further specifications or offerings of availability. Not that I greatly care to find the Mitutoyo part—when the time comes, I’ll likely get whatever cast-off ¼ wavelength retarder I can find at a reasonable price, and adapt it. But the omission everywhere else makes this mention seem a bit like a fluke.

Cheers,

--Chris

[Charles Krebs]

From: http://www.microscopyu.com/print/articl ... print.html

Polarizing components are utilized to eliminate internal reflections from optical elements, and other sources of glare, that would lower the image contrast. The primary polarizers are placed between the light source and mirrors to polarize light entering the zoom body. Analyzers (or secondary polarizers) positioned above the half-reflecting mirrors eliminate undesirable reflections before they reach the eyepieces. In order to allow image-forming light reflected from the specimen to pass through the upper polarizers to the eyepieces or camera attachment, a quarter-wave retardation plate, which functions as a de-polarizer, is mounted over the front lens element of the common main objective. In use, the retardation plate can be rotated to an angular position to ensure the brightness and contrast of the image is optimized for the specimen being studied.

From: http://www.zeiss.de/c1256b5e0047ff3f/Co ... 6800503e12



From: http://zeiss-campus.magnet.fsu.edu/arti ... ected.html

In cases where objectives of very low magnification are used in reflected polarized light, a rotatable optical plate (termed an Antiflex cap) consisting of a one-quarter wavelength lambda plate is placed on the objective front lens element to block reflections from the objective itself. The Antiflex method is also particularly useful when the specimen has very low reflectivity, such as would be observed in coal samples.

[Charles Krebs]

Chris,

I knew I had recently read a good explanation about this, but could not find it last night. I think this was it:

http://tech.groups.yahoo.com/group/Micr ... sage/72469