Experimenting with how to use a Mitutoyo 10x Plan APO

[Tucson Tom]

A few months ago I found a decent deal on a Mitutoyo 10x Plan APO microscope objective. This week I began trying to set it up and use it. This objective, being an infinite conjugate objective, requires a relay lens, and a 200mm lens is what it is made for. To this end I purchased a couple of 200mm f/4 manual focus prime lenses. One is an FD mount canon 200mm macro f/4 lens (a great lens in its own right!). The other is a M42 mount Takumar screw mount lens. I plan to use these on my Canon EOS 5Dii camera, so I sniffed around on Ebay and purchased some adapters. The FD lens adapter can be purchased with a negative lens to allow it to focus at infinity, or without. For a multitude of reasons, the best thing is to purchase one without the lens, at least for this kind of macro work. The M42 lens is in many ways at an advantage. It will focus at infinity, and manual control of the diagram is already provided for on the lens.

I'll also note that the Canon 200mm FD macro lens is a desirable item and still commands a fairly high price. The Takumar however is an excellent lens and is readily available in the 50-100 dollar range, and you can always get lucky.

But now the main part of this post. I want to do some tests and decide which of these two candidate relay lenses will be best. And for the tests, I need a test subject. My thinking was that a flat subject with lots of recognizable fine detail would be great, and for some reason I thought of a semiconductor die. Being an electronics experimenter I have a lot of options, but chose a faulty Intel 8751 chip I had on hand.



A hammer and chisel and a 2x4, and one sharp whack ...


And it pops right open (it is a sandwich of two ceramic slabs glued together).



Also, to put the Mitutoyo 10x on the front of the lens, I bought two other items on Ebay. One is a disk-like adapter that holds the 26mm threads on the Mitutoyo and allows it to thread into 52mm filter threads. The other is a 52 to 58 mm filter adapter (these come in two flavors and you have to take care to get one of the correct "sex"). I actually have a flock of these filter adapters so I can screw around with all sorts of adaptations.

One of the first things I notice is that both combinations vignette just a little.
I am undecided which is better at this point, they are quite similar. Here are out of focus images with the Canon FD 200 on the left, the Takumar 200 on the right.
This is on a full frame camera, so for many subjects, a person could crop or fix the vignetting in photoshop. On a crop camera like the 7D, this would not be an issue at all.



Here is what we get with the Canon FD 200 macro:



Here is what we get with the Takumar 200 f/4



And here are some 400 x 400 pixel crops (unscaled) with the Canon lens on the left and the crop from the Takumar on the right. If I calculate right, these little crops are 0.256 millimeters (256 microns) on a side.





And just for fun, here is a similar area of the chip photographed with the MPE-65 set at 5x magnification and f/4. Surprisingly, even with a subject this flat, focus varies across the subject and I had to reshoot this area to get a cropped image for comparison. I think the differences in color are due to changes in ambient lighting (tungsten mixing with room fluorescents). The MPE is on the left, Canon FD in the center, Takumar on the right.
The MPE-65 image is an unscaled crop of 500x500 pixels, the two Mitutoyo images were 1000x1000 pixel crops scaled to 500x500. It is not as apparent on these images as it is here on my screen, but the Mitutoyo impressively outdoes the MPE-65 at these magnifications.



I am still deciding whether I like the Canon or Takumar better. For several reasons I would like to dedicate the Takumar to this purpose and keep the Canon 200 macro for a general purpose lens and am trying not to let that bias me.

[rjlittlefield]

Tom, regarding comparison with the MP-E 65, I'd be just plain shocked if your 10X Mitutoyo did not give substantially more resolution on subject than the MP-E 65 does. Very briefly, from the standpoint of the subject the 10X Mitutoyo is running at NA 0.28 = f/1.79 effective, while the MP-E 65 at 5X is running at only about NA 0.149 = f/3.36 effective. With almost twice as wide an aperture, the Mitutoyo has a huge advantage in the diffraction department. The Mitutoyo's cutoff frequency will be almost twice as high, and it will have substantially higher contrast then the MP-E 65 at lower frequencies. See Lenses for use at 4-5X on an APS-sized sensor and Pixels for use at 4-5X on an APS-sized sensor for more discussion and illustrations on that topic.

As for the Takumar versus the Canon, your results look to me like it's about a wash. I'm not surprised by that, since sticking the Mitutoyo on front effectively stops down both of the telephotos to about f/18. I've seen numerous comparisons of various tube lenses, and my general impression is that while differences are unambiguous in head-to-head comparisons, they're seldom big enough to be noticed outside that context.

One caveat is that you really ought to make the decision based on tests done the way you intend to use the lenses. Certain kinds of aberrations become much more obvious in stacked results, so it's a bit risky to choose based on single frames. I don't see any evidence of those aberrations in the results you're showing, but if you plan on stacking, then comparing stacked results would be safer.

--Rik

[Tucson Tom]

[Craig Gerard]

[rjlittlefield]

[Tucson Tom]

[Chris S.]

Tom,

Interesting thread you have here. To my eye, the Canon converging/relay/telan/tube lens looks considerably better than the Takumar. But as Rik said, this may not translate to meaningful differences in real-life shooting. I wonder how it would look, a stacked comparison of the two lenses with one of your micromounts at 100 percent pixel crops, after you've done your best in Photoshop with each stack. Would the differences I think I saw in your images of the Intel chip translate to meaningful differences in the final image?

Like Craig, I think 25 microns is way too much for the Mitty 10x. But I'd go even tighter than 10 micron increments to 4.6 microns. Ten microns will work, but in my experience (with the Mitty 10x on a Nikon D200), 4.6 micron steps confer small but notable improvements. That said, your 25 micron steps were certainly not a disaster--and until you get your motorized stage going, finer steps would be a huge pain.

Broadly diffuse light, as suggested, seems a solid next step. That said, some good micromount photographs I've seen have also involved small additional reflectors to highlight particular facets. Can't say I've tried this--but I bet you will.

Cheers,

--Chris

[ChrisR]

Paper tubes work as diffusers too, and lose less light that a styrofoam cup.
Cut a strip an inch and a half or so wide, run it over a sharp edge and it'll roll itself into a tube. More pressure + sharper edge = tighter tube.

It'll then cling to the sides of the objective - you can slide it up and down.
If you put a flash one side and a shiny ( ally foil/tape) reflector behind you'll get very flat light indeed.
For a posh version run black pvc tape around the outermost end. That holds the tube, and also stops some stray bright light bouncing up into the lens. I've found that helps with shiny minerals, You can adjust the height from which your ring of light comes at the specimen.

While I'm wittering about bits of paper - try black paper behind the specimen to reduce flare.

[Tucson Tom]

[rjlittlefield]

[Tucson Tom]

[rjlittlefield]

[Craig Gerard]

Tom,

Lee Filters and Rosco make good diffuser sheets that can be formed into cylinders, cones, etc. The filters do not alter colour temperature; but don't come with yogurt
I like LeeLUX 400...as a diffuser, but it's good to have a range of options on hand.

http://www.leefilters.com/lighting/diffusion-list.html


Craig
_________________
To use a classic quote from 'Antz' - "I almost know exactly what I'm doing!"

[Tucson Tom]

Chris S said that he thought the Canon had the edge over the Takumar. This was disturbing to me because I really wanted to NOT use the Canon 200 macro as a relay lens, since I like it a lot a a general purpose macro lens (with the FD lens adapter it won't focus to infinity, but that is neither here nor there to me). But I realized that somewhere I had a plan old FD mount 200mm f/4 canon lens that I bought on Ebay, and I managed to dig around and lay hands on it yesterday. A fellow I know said it was probably the sharpest FD lens he ever owned -- so lets see how it stacks up against the Takumar.

I will note that the FD to EOS adapter adds about 9mm of extension. And the Canon EOS body is about 2mm thicker than the old Canon manual bodies -- this adds up to about 11mm of extension that we can't do anything about. I worried about this (thinking that it might be absolutely critical that the focus be set at infinity in order to use it as a relay lens). My impression is that it simply does not matter at all. It does give a bit more magnification as you can see on the following full frame images.



Of course you can't tell much from these, so here are some unscaled crops from the left center of these images.



I should note that the FD to EOS adapter is machined better than the M42 to EOS adapter I purchased. The M42 to EOS adapter has about .0055 inches of slop on the bayonet lugs that engage the Canon body, which allows the lens to wobble just a bit. (In boxes labelled "Pixco" if you want to go with an M42 lens and avoid these adapters).. I was planning to fuss around with shims and perhaps epoxy to fix this - but I think I like the Canon lens better and will make it my standard relay.

Also note that the 10mm of extension gives about 5 percent extra magnification, making the Mitutoyo into a 10.5 x lens (as you can see in the images).

My bottom line after all this is that either lens will do a great job, but I am going to go with the Canon. This is risky to judge from one pair of images like this (albeit shot in as near identical conditions as I can manage) -- but I would say the image from the Canon lens has more contrast and color quality. Resolution wise, I would take either.

[Tucson Tom]

I just wanted to check back in with an example of some results. After changing my lighting setup and getting the white balance right for a cheap tungsten filament "halogen" light I got at Ace Hardware, I am getting results that I am satisfied with. A diffuse light source (using my yogurt cup diffusor) made a significant difference in how photographs of crystals (with specular reflections) respond to focus stacking.
Thanks again Rik for the advice.

My notes on my fussing with white balance may be found at http://cholla.mmto.org/minerals/macro/whitebalance/ for those who might be interested.



This is the mineral Calderonite from the C and B mine in Gila County, Arizona, 2.4 mm field of view. It is a Pmax stack (via Zerene) of 64 images taken at 10 micron intervals.

The next thing I will probably do will be to experiment with the size of the focus steps and see where Zerene falls apart. Do something like 10 micron, 20 micron, 40 micron, 80 micron, and so forth in a geometric progression. Apart from discovering when stacking falls apart, I will also get some experience with what images look like when the stacking interval is too coarse.