Microscope objective for low magnification photos

[Tonikon]

Hi everyone,
is not a mistery that to use a good microscope objective (without eyepiece) to do high magnification photos can give better results than most of common macro lenses.
But this is surely true for very high magnification (due to the diffraction problems that afflict Luminar, Photar and similar lenses), whereas I don't know anyone that use microscope objectives (ever without eyepiece) for low magnification photos.
For example, a 2.5x or a 4x objective.
Recently, I have purchased on eBay an Olympus MPlan 1.3x 0.03 for metallographic microscopes, wich gives 1:1 with a working distance of about 50mm. It produces pics with excellent center sharpness; borders are afflicted by a moderate field curvature, but using a stack software I can obtain perfect images corner to corner!
So, I'd like to use a 3-5x microscope objective to fill the gap between this Oly 1.3x and the Nikon U10 (another excellent performer). Naturally, I need an objective wich covers at least the Nikon DX format at its nominal magnification.
While we're at it, how can I calculate the real focal lenght of my Olympus MPlan 1.3x? If I divide 160/1.3 I'll obtain 123mm of focal lenght, but this is not in good accord with a WD of only 48mm at 1.3:1. Moreover, 0.03 means about f/16 and 130/16.6 gives about 7.5mm of entrance pupil (right?), and my Oly has only 6.5mm of entrance pupil (while exit pupil is about 10mm).

Ciao

Toni


Butterfly wing - Olympus MPlan 1.3x 0.03

[lauriek]

I don't know the answer to your question, but I've been in the market for a nice 5x objective for a while. There unfortunately doesn't seem to be a 5x version of the Nikon CF M Plan range, which is what I'd ideally like. Hopefully someone will suggest an alternative!

I know the Oly 38mm/f2.8 or 38mm/f3.5 bellows lenses are about the right magnification, but these, especially the f2.8 model seem to be quite expensive at the moment. (The f3.5 model is quite like a microscope objective and has an RMS thread, requiring a PM-MTOB to mount on an OM bellows/mount.)

[Harold Gough]

I know the Oly 38mm/f2.8 or 38mm/f3.5 bellows lenses are about the right magnification, but these, especially the f2.8 model seem to be quite expensive at the moment. (The f3.5 model is quite like a microscope objective and has an RMS thread, requiring a PM-MTOB to mount on an OM bellows/mount.)

See here for details:

http://www.alanwood.net/photography/olympus/#lenses

This search term will find them for you on eBay:

(olympus, zuiko) (20mm, 38mm, 20, 38 ) macro

Delete the 20mm and 20 if you don't want to see them but they rarely turn up anyway. In any case, delete the space before the final bracket, as without it we get a "cool" Emoticon in this forum!

Harold

[mgoodm3]

I do a lot of imaging in the lower magnification ranges and am also interested in low-mag objectives.

4x objectives tend to work best in the 4:1 - 7:1 range for me. when you drop the amount of extension too low you start getting serious spherical aberration in the images.

I have a Leica Photar 25/2 and that takes extrememly nice images in the 3:1 - 7:1 range. I think a 50mm photar would likely be nice. No experience with the others in that range.

In the 2:1 range the best that I have found so far are stacked lenses. I use a Nikon 105/4 with a reversed Nikon 50/1.4.

I really need to play with some low power objectives - just gotta find a suitable one that will work well without a tube lens or an eyepiece.

[Charles Krebs]

If you really want to use objectives in that range there are (were) a couple of CF M Plans (finite, 210mm tube length) to watch for.

One was a 2.5/.075, with a working distance of 11mm. Focal length is 76.68mm. (See last paragraph for important note on this one).

One was a 5/0.1, with a working distance of 20mm. Focal length is 37.64mm.

A reason that these may be of less interest than their higher magnification brethren is that there is no real compelling reason to use them. At the magnifications they provide, there are many more conventional choices that may even offer advantages. The key is the aperture. If you go by the 1/2NA relationship to determine f-number, the 2.5/.075 has a aperture of about f6.7, while the 5/0.1 would have an aperture of of about f5.

So you have a 77mm f6.7 and a 38mm f5. It is not too hard to find optics (designed for this type of photography) in that focal length range that have maximum apertures greater than that. If these lenses were "stopped down" to the the equivalent of the objectives I'm not sure there would be any advantage to the microscope objective. Also, in this magnification range it is often nice to have an adjustable diaphragm because it is possible to take a "single frame" picture stopped down a bit more than the fixed objective aperture.

I have the 5X M Plan. Haven't used it in a while. Some time ago, when I "tested" it against the other optics I use in that range there did not seem to be any real advantage. As I recall it was as good, but not better. (Perhaps I should re-visit it and take another look). One objective that I would still love to try is the Nikon CF N Plan Apo 4/0.20. This is a "biological" objective with a working distance of 15mm.

While it is possible to find more conventional lenses for these magnifications, consider the situation at the higher magnifications. The 20/0.40 CF M Plan ELWD has a focal length of 11.1mm, working distance of 10.5mm, and (roughly) an aperture of f1.25. I know of no "conventional" macro lens that can match this.

There were actually 1X and 1.5X CF M Plans. A very important note on these (and the 2.5X mentioned above) is that Nikon notes that they will vignette with UW eyepieces. These eyepieces "see" a field of 26.5mm. So I would expect them to also vignette considerably on any SLR that had a format diagonal around 26.5mm and greater. (Might be OK on the 4/3 format like Olympus, but would likely be an issue on any other DSLR).

Charlie

[Charles Krebs]

... one additional note....

We've talked about this elsewhere, but it should be stated again. Most "finite" microscope objectives were designed to be used with eyepieces that corrected certain chromatic aberrations. The only ones I am aware of that were (supposedly) chromatically corrected within the objective (ie: no need for corrective type eyepieces) were the Nikon CF's. Some of the "non-CF" objectives I have tried showed significant chromatic aberrations, especially toward the field edges. Obviously I've only "sampled" a few, but it is something to be aware of.

"Infinity" objectives really require a tube lens. With some manufacturers there are chromatic corrections taking place in this tube lens.

[mgoodm3]

I have a standard Olympus 4x/0.10 plan infinity-corrected objective that works very well on a bellows without a tube lens. Field is quite sharp to the edges.

[Charles Krebs]

mgoodm3,

That's good to know. I think Craig has an Olympus 20X infinity that he has tried. Not really sure what his experience was.

The only "infinity" objective I have tried this way an older American Optical 4x. Good on the microscope, bad on the bellows! As I mentioned to Craig in a message earlier this spring, you can get an image with an infinity objective on bellows (without the tube lens), but I would not hazard a guess as to the image quality. With the one I tried the coverage and resolution were fine, but the chromatic aberrations (when used on a bellows) were brutal.

Sounds like a "try it and see how it works" situation.

[augusthouse]

The 20x infinity-corrected objective that I was using for a brief time was a Nikon CF; it had a standard working distance and was rather tricky to work with on a bellows - not ELWD or SLWD. I will locate some images that I took with it. I have since sold the objective. It was an interesting piece of glass to poke at things, but belonged on a scope.

There were a couple of recent discussions concerning the use of infinity-corrected objectives on bellows (I will locate the threads and insert links); but finite is the way to go.

One objective I do have, and I know some others also have, is the Carl Zeiss Plan 2.5/0.08 160/-. It is an exceptionally well crafted objective (as you would expect from Carl); but it has been a while since I last used it (shame); however, this thread and Toni's question has reminded me of its presence in my collection. It is going to be given a mission.

As I recall Toni's setup he makes use off a turret, as shown in this thread.Here.

*later note: Carl Zeiss Plan 2.5/0.08 160/- exhibits CA and vignetting.

* edited to add observation

Craig

[rjlittlefield]

While we're at it, how can I calculate the real focal lenght of my Olympus MPlan 1.3x? If I divide 160/1.3 I'll obtain 123mm of focal lenght, but this is not in good accord with a WD of only 48mm at 1.3:1. Moreover, 0.03 means about f/16 and 130/16.6 gives about 7.5mm of entrance pupil (right?), and my Oly has only 6.5mm of entrance pupil (while exit pupil is about 10mm).

I think this question has not been answered yet. Let me give it a try.

Calculating the real focal length of this objective is very difficult and cannot be done from only the information provided here.

If the objective were a "thin lens" (which it is not), then focal length could be calculated as 160/(m+1). A thin lens would have a focal length of 160/2.3 = 69.6 mm and would give 1.3X when focused on a subject 123 mm away. (160/123 = 1.3)

Obviously the objective focuses a lot closer than that.

To focus closer without providing higher magnification, the objective contains a collection of both positive and negative elements that make it act sort of like a wideangle lens. The area viewed is larger than you would expect based just on the distances involved.

Such an objective can be modeled as a "thick lens". But thick lenses have more parameters than just focal length, and there are many combinations of parameter values that could produce 1.3X under the same conditions.

That is why we cannot calculate the focal length of your 1.3X NA 0.03 objective.

What we can tell (more or less!) is what sort of ordinary lens would correspond to it.

The equation that f-number = 1/(2*NA) is only approximate, and it is not very good at low magnifications.

If you slog through the algebra, you find that a much better approximation is 1/(2*NA) * m/(m+1).

In other words, there is an m/(m+1) correction factor. For large m, the difference is small; for small m, the difference gets larger.

Using the better approximation, we see that your 1.3X objective actually corresponds closer to f/9.4, rather than the f/16.7 that comes from 1/(2*NA).

So, a roughly 70 mm lens that works well at f/8 should be similar to your 1.3X NA 0.03 objective, except for having a longer working distance.

With the m/(m+1) correction, the numbers that Charlie mentions change a bit. The 2.5/.075 corresponds to f/4.7, and the 5/0.1 corresponds to f/4.1.

However, the difference is not enough to change the point he makes, that there are conventional lenses also available that work well at these magnifications and apertures.

--Rik

[Charles Krebs]

RIk...

much better approximation is 1/(2*NA) * m/(m+1)

I like that! Thanks. Makes more sense of the "real word" experience we saw a while back when it was obvious that there was less DOF than the simpler relationship would lead us to believe.

[dmillard]

The equation that f-number = 1/(2*NA) is only approximate, and it is not very good at low magnifications.

If you slog through the algebra, you find that a much better approximation is 1/(2*NA) * m/(m+1).

In other words, there is an m/(m+1) correction factor. For large m, the difference is small; for small m, the difference gets larger.


--Rik

Thanks Rik -

That equation almost satisfies the f/stop of my Nikon CF N Plan 4/0.13 when measured against a 35mm f/2.8 Canon Macrophoto lens (approx. 3.14 rather than the 3.84 indicated by the simpler formula 1/2NA).

One objective that I would still love to try is the Nikon CF N Plan Apo 4/0.20. This is a "biological" objective with a working distance of 15mm.

Charlie

After seeing the impressive performance of the Nikon CF N Plan 4/0.13 objective last spring, I put a wanted listing on LabX for the Apo version, and also put this lens on a saved search list on eBay, with unfortunately no luck from either venue.

[Tonikon]

.....There were actually 1X and 1.5X CF M Plans. A very important note on these (and the 2.5X mentioned above) is that Nikon notes that they will vignette with UW eyepieces. These eyepieces "see" a field of 26.5mm. So I would expect them to also vignette considerably on any SLR that had a format diagonal around 26.5mm and greater. (Might be OK on the 4/3 format like Olympus, but would likely be an issue on any other DSLR)......

Thanks for your complete and detailed informations (as usual, however), Charles. Not good notices... the CF 2.5X will vignette on my Nikon DSRL, and probably it will show not good corners...

.....As I recall Toni's setup he makes use off a turret, as shown in this thread......

Yes, Craig. I use a modified microscope and so I prefer to use microscope objectives: they have a comfortable work distance.
For example, my Oly MPlan 1.3 gives 1.3:1 with about 50mm of WD, while a Luminar 63 requires much more distance and it is unusable with a normal microscope stage.

.....If the objective were a "thin lens" (which it is not), then focal length could be calculated as 160/(m+1).....

Gosh....I have lost an "m"....

.....So, a roughly 70 mm lens that works well at f/8 should be similar to your 1.3X NA 0.03 objective, except for having a longer working distance.....

Thank you very much, Rik...now all is clear. I use currently a lot of conventional macro lenses for 1X-5X range, as a Nikkor AFD 200 Micro f/4 ED, a Macro-Nikkor 65mm f/4.5, a Rodenstock Rodagon 50mm f/2.8 and others, on a Nikon Bellows PB-4.
But my problem is to achieve 2-5x on my modified microscope, for vertical shots. Infact, using one of these conventional macro lenses on my modified microscope is impossible, because they are not RMS.
Moreover, even a Leica Photar 50mm f/4 or a Canon 35mm f/2.8 or a Minolta 12.5 f/2 (that are RMS mount) area unusable, because they are a bit fat and there is not enough space on the nosepiece (actually, I'm using a better microscope with a 5-position nosepiece).
Probably, I have to search for a Luminar 40. But years ago, I have tried a Luminar 40 (Zeiss Winkel old series) and I did not liked it (too much field curvature).
Because of all this, I have thought to a 3.5-5x microscope objective.

Ciao and thanks to all

Toni

[rjlittlefield]

much better approximation is 1/(2*NA) * m/(m+1)

I like that! Thanks. Makes more sense of the "real world" experience we saw a while back when it was obvious that there was less DOF than the simpler relationship would lead us to believe.

It is part of the explanation. But most of the observations we discussed in http://www.photomacrography.net/forum/v ... .php?t=424 were at 8-10X magnification, where the m/(m+1) correction factor is only around 10%. There is also a small correction because tan(a) is only approximately equal to sin(a). As I wrote in the earlier thread, "at 10X and NA 0.25, I calculate the difference to be only about 15%, much smaller than what I think I'm seeing in the pictures."

I have been chipping away on those earlier results, and I'm pretty sure that a more important part of the explanation lies in pupillary magnification factor.

For example my Olympus 20 mm f/2 bellows lens is labeled "f/2" because it has an entrance pupil of 10 mm for light entering from infinity. This is a standard way to make the measurement. But the lens also has a PMF of roughly 0.7 (based on by-eye measurements using a caliper). When this lens is extended to give 8-10X, the PMF makes it act more like a thin lens at f/2.8, rather than the f/2.0 that you might expect based solely on the markings. This is a whopping 40% difference.

What I find fascinating about this is that the Olympus 20 mm f/2 bellows lens is specifically designed to be used at rather high magnifications, nominally 4.2X - 16X. But this is exactly the range where the effective aperture is dominated by PMF. Olympus doesn't quote PMF.

In other words, that "f/2" rating quoted by Olympus simply does not apply under the conditions of intended use, and they don't provide a corresponding rating that does. If I were feeling grouchy, I might call this "misleading advertising". But since I'm not, I'll just point out the facts and once again emphasize that actual testing is a lot more reliable than computations based on published specifications.

--Rik

[dmillard]

Toni -

These images may be of interest to you:


Red Admiral (Vanessa atalanta) forewing at 6.4X, Nikon CFN Plan Apo 4/0.20, D200, 30 images at .01mm stacked in HF


Same as above, but with Nikon CFN Plan Achromat 4/0.13

I finally got a Nikon CFN Plan Apo 4X objective, and ran some tests today. The original object field was 3.68mm wide, determined by measuring 3.5mm on a glass micrometer slide through the viewfinder, and then dividing this number by .95 to derive the full coverage. The field width displayed here represents approximately 1.25mm on the wing (I need to learn how to insert scale bars!). Both images were taken as JPEGs, and were also sharpened slightly (35% @ pixel radius 0.7) after resizing for web display.

The loss of resolution in the 4/0.13 image results from the smaller effective aperture. At the nominal 4X magnification, apparent differences between the lens performances are very slight with my camera: both objectives produce images superior in color, contrast, and clarity than a 35mm Canon Macrophoto lens (despite its larger maximum aperture than the Achromat), and give markedly better results than a 40mm f/4 Luminar. However, the short working distance of about 16mm makes the Nikon objectives best suited for relatively 2-dimensional subjects (like butterfly wings : ) ).

David

Edited for clarification