Mitutoyo 50x Sharpness Problems?

[Zack]

Hi all! I have been lurking here for a while and have a question I was hoping I could get some guidance on. I have just purchased a Mitutoyo M Plan APO SL50x (NA = .42), which is my first foray into very high magnification photography. I took the below photo (about 50 images stacked in Zerene), but it just doesn’t seem as sharp as I feel it should be. In terms of gear, I am shooting on a 20.9MP Nikon Z50, APS-C camera, through a Tamron 70-200 2.8 as a tube lens. I use flashes at 1/200s to light the subject, and my setup has virtually no vibration. At this point I feel like possibilities are:

1. This is actually a good, sharp photo and I should be happy with it
2. There is something with my setup that’s introducing blur
3. The sensor on my camera doesn’t have enough resolution to render photos sharply enough, and I should upgrade to something like the Nikon Z7 with 45.7MP
4. At this level diffraction blur is inevitable and there’s nothing I can do about it

I would love some thoughts as to whether it’s one of the above, a combination, or something else that I hadn’t even considered. Thank you!

[Chris S.]

. . . I have just purchased a Mitutoyo M Plan APO SL50x (NA = .42). . . .

4. At this level diffraction blur is inevitable and there’s nothing I can do about it. . . .

I would love some thoughts as to whether it’s one of the above, a combination, . . .

Zack, welcome aboard!

Your image looks about right, to my eye, for an NA 0.42 objective magnified to 50x on sensor. Here you are well into diffraction territory. Can you do anything about it? Probably not, with this lens.

However, you can do something about it if you switch to the Mitutoyo M Plan Apo 50x/0.55 objective, which will give you more resolution. For this reason, the NA 0.42 SL version of this lens, which you are using, is not the preferred Mitutoyo 50x optic for most uses. The Mitty 50/0.55 is more commonly used; there are also a few other objectives around this magnification that forum members use, some with still higher numerical apertures and correspondingly lower working distances.

Nice work with what you have!

--Chris S.

[Zack]

Thanks for the quick response, Chris! I was looking into the 50x .55 as well, but I got a good deal on this one ($800) and I really like the longer working distance; I suppose that's my trade-off for losing a bit of resolution. I don't think it makes sense with my budget and skill level to shell out an extra thousand dollars or more for just a slight uptick in resolution that the 50x .55 would offer. Much appreciated! I just started with photomacrography about 6 months ago, so I'm excited to keep learning and testing my skills

[Sym P. le]

Have you explored the range of settings on your zoom lens? Since most lenses are not optimized about their extremes, reducing the zoom and closing the aperture a stop or two may give you a step up in performance that is more pleasing than just chasing the maximum magnification. The next step, of course, is to fetch a know good performer for a tube lens.

[Zack]

I haven’t tried that! I always thought that I should keep the tube lens fully zoomed in (200mm, about right for Mitutoyo) and at the widest aperture, f/2.8. I’m definitely going to test slight differences in zoom/aperture tomorrow.

If I were to get a 50x/.55 instead, would there be a noticeable difference in sharpers from my current 50x/.42? There is a Mitutoyo BD 50/.55 that I was looking at (shorter WD of 5.2mm, but my Nikon 20x has a WD of 1.3mm and I’m able to get light in there). Wondering if that upgrade is worth the cost. I’m doing this for aesthetics and the hope of enlarging photos to 16x20 or bigger one day, so ideally I’ll get as much resolution and sharpness as possible. Thoughts?

[Scarodactyl]

There will be a noticeable difference with a 50x/0.55 but it's still a bit underpowered in resolution for the magnification. If you can work with a 1mm working distance you should consider more conventional metallurgical objectives too, like the Olympus mplanfl 50x/0.8. Heck, I picked up a generic Chinese clone of the Nikon cf plan bd 50x/0.65 for 35 bucks and it is pretty good, and very easy to get light in once the shroud is removed.

[Zack]

Any suggestions on how to find a Chinese clone objective to try? I am willing to try a few different options for testing and I would rather shell out $35 than $2,000+ on something that wouldn’t be noticeably better. I know Mitutoyo makes a 50x/.75 HR with a decent working distance but I can barely find those anywhere and they are way out of my budget.

[Zack]

I did some testing and compared the Mitutoyo 50x/.42 to a Nikon CFI Plan Achro 20x/.40 that I have. The bad news is that the Nikon has a horrible working distance of 1.3mm, compared to the generous 20mm on the Mitutoyo. However, the Nikon was also about $250, and I'd love some input on the photos below but I am honestly not seeing an advantage of the Mitutoyo in terms of sharpness. Of course, I have to consider the fact that the 20x photos below are significantly cropped to about 3MP and probably unprintable. I would love input from those with more experience than me, but I am not seeing justification in the added cost. These photos are completely unprocessed and were taken with identical camera and flash settings, so it is interesting to note the color and contrast differences.

Mitutoyo, full image

Nikon, cropped to match Mitty FOV

And then the severely cropped versions:

Mitutoyo, close crop

Nikon, close crop

Honestly, the Nikon 20x severely cropped seems even sharper to me than the Mitutoyo... I defer to the experts, though. Thanks, as always, for all your wisdom and input!

- Zack

[rjlittlefield]

I did some testing and compared the Mitutoyo 50x/.42 to a Nikon CFI Plan Achro 20x/.40 that I have. The bad news is that the Nikon has a horrible working distance of 1.3mm, compared to the generous 20mm on the Mitutoyo. However, the Nikon was also about $250, and I'd love some input on the photos below but I am honestly not seeing an advantage of the Mitutoyo in terms of sharpness.

At best, sharpness on subject is limited by diffraction, which depends on the NA of the lens. The Mitutoyo has only 5% larger NA, so that is not a significant advantage.

When using a microscope objective, sharpness is seldom affected very much by sensor resolution.

To see why, note that effective aperture, from the standpoint of the sensor, can be calculated as Feff = magnification / (2*NA) .

So, your 50x/.42 is running at almost effective f/60, and the 20x/.40 is at effective f/25. Both of those are significantly narrower than the diffraction-limited aperture of your 20.9MP APS-C sensor, whose Nyquist cutoff frequency (2 pixels per cycle) is around 119 cycles/mm, corresponding to diffraction cutoff at f/16. (Discussion at viewtopic.php?p=124831#124831 .)

Putting all that together, the point is that your sensor can adequately capture all the optical information in both images, and considering diffraction only, those images will have almost equal sharpness on subject. Your results are consistent with theory.

If you want to capture more detail on subject, you need larger NA. Several of us have taken the path suggested by Scarodactyl, of getting a high-NA metallurgical objective like the Olympus mplanfl 50x/0.8.

All such lenses require removing the outer barrel, which may be just a matter of unscrewing it or may require metal-cutting surgery. The working distances are also quite short, in the vicinity of 1 mm. In my own case, the lens is a finite Nikon 40X NA 0.80 M Plan apochromat, used as shown at viewtopic.php?t=40974 and producing results like shown at viewtopic.php?t=40860 .

These photos are completely unprocessed and were taken with identical camera and flash settings, so it is interesting to note the color and contrast differences.

I'm thinking that these images are stacked, and you mention Zerene, and I know that with default settings Zerene Stacker will do brightness and contrast adjustments, which can also affect color saturation, and even without those adjustments, just the basic PMax stacking method can also affect color/contrast/brightness.

So, I'm inclined to think that these images actually have been processed quite a bit, and in ways that worry me about contrast and color.

It is certainly possible that two lenses would give different color and contrast, particularly with insect specimens which seem particularly subject to some form of metamerism that can cause a strong reddish cast. (See viewtopic.php?t=37664 , third image panel.)

But to make that decision, you would need to look at the source images straight out of camera, before stacking.

I hope this helps!

--Rik

[Pau]

Your image looks about right, to my eye, for an NA 0.42 objective magnified to 50x on sensor. Here you are well into diffraction territory. Can you do anything about it? Probably not, with this lens.

I concur with Chris. As demonstrated in your last images 0.4 is a NA more typical of a 20X objective, even there you're into diffraction territory so at 50X you couldn't expect sharp images at sensor level. I use a 40X 0.80 (0.7mm WD ) and it doesn't deliver a very sharp image at 100% view.

If not really overdone, strong sharpening will improve quite a lot the visual aspect of the final image.
I've applied a 100/3/5 PS unsharp mask to your first image and it looks clearly sharper, test it (I could post it if you want)

[Scarodactyl]

Any suggestions on how to find a Chinese clone objective to try?

I've found two on eBay at that price. There isn't a regular supply unless you want to buy new (not recommended).
You might want to consider a mitutoyo clone like this one, though they're weakly teated and not that cheap viewtopic.php?p=280193#p280193
There is also this objective which has been on eBay a while. It is a standard chinese Olympus clone with that 0.55 NA and a long working distance https://www.ebay.com/itm/Motic-Microsco ... 3511787220
I tried a similar though likely not identical Chinese olympus copy 50x lwd objective and it was quite decent.

[Adalbert]

Hi Zack,
Your photo is pretty good.
I have already compared Mitu 20x/0.42 vs Mitu 50x/0.55
and found that the difference in resolution is not that big.
Best,
ADi

[Zack]

Rik - the "diffraction cutoff" that you mention - does that mean that anything beyond f/16 will appear equally soft? Is the solution to get a camera with a larger sensor or more MP? You also mention that you use a finite lens and I noticed that it asks for a 210mm tube lens. Would my 200mm Tamron lens suffice? Or do tube lenses need to be exact measurements in order for a finite objective to function as intended?

Pau - thanks for checking those numbers for me. I tried to apply sharpening as well and it made the result what I would call "acceptable," but perhaps my standards are just too high... or I have already spent too much time comparing myself to the incredible images people on this forum produce.

Scarodactyl - thanks for linking that Ebay Chinese clone. It's no return, which is risky, but for $150 instead of thousands I may have to take a leap of faith.

Adalbert - I appreciate the comparison. That certainly is convincing for the fact that I should either go with a 50x/.75 or .8 if I really want a noticeable difference.

I think the common thread through all of this is that I just do not have enough experience or knowledge to even judge for myself what "acceptable" sharpness and results are, which is why all your input is invaluable. Sorry for asking so many questions but in the past few weeks, having finally immersed myself in microscopy and photomicrography, I have realized that there is more information out there than I ever dreamed I would even want to know! I appreciate the input from each and every one of you in helping me to learn so much

[Scarodactyl]

My test of the similar Chinese olympus clone long working distance objective is here btw viewtopic.php?f=25&t=43193. I unfortunately didn't have a mitutoyo 50x to do a head to head with, and I've since sold the Chinese lmplanfl objective with the scope it came on.
Incidentally I did get a Motic elwd apo 50x/0.55 (mitutoyo clone) a couple months back, but the increase in resolution over my mitutoyo 20x/0.42 was somewhat counterbalanced by axial purple/green CA, so I ended up selling it. I would not recommend that model unless you find it really cheap.
The newer Chinese 50x/0.65 mitutoyo clone linked above does seem promising, though it will probably be a while before we have enough of them in the community to have an idea of quality consistency and to get some head to head tests.

[Lou Jost]

When using a microscope objective, sharpness is seldom affected very much by sensor resolution.

However, if (as is the custom here) people are comparing images at 100%, then comparing across sensors will give "sharper" results with the sensor with the WORST resolution. If comparing across sensors, one should compare crops at a fixed proportion of the field of view, not at a fixed Photoshop zoom level.