Trials with Mitutoyo 10X and Canon MP-E on Nikon D800E body

[papilio]

Very interesting, getting to know the new D800E. I have the feeling that it's too much camera for me, but I'll get used to it. Be warned, batch converting, stacking and even post take forEVER. I'm not yet anywhere close to pixel level detail on it ... always open to hints. I don't believe it's anything as simple as vibration, inadequate lenses, etc. Likely a matter of gaining more experience.



Here are quite a few test images, some with the Canon MP-E fitted to the Nikon, some with the Mitutoyo Apo 10X/0.28 @ 10X and 6X. Thanks to Rik and Chris for cluing me in to reversing the Raynox lenses ... Wow, quite a transformation. Rik, you said that the corners which you were getting were much better than what I had posted, but you were as usual understated. The corners are quite satisfactory whereas before they were atrocious.


The two Mitty/tube lenses, giving 10X at ~f/18 and 6X at ~f/11


Woops ... when putting the lens together for this photo I put the iris behind the Raynox rather than behind the Mitty. :/


MP-E on Nikon body




As I had bought my previous Canon 6D largely in order to be able to use the MP-E 65mm, I'm happy to say that adapting the lens to fit the Nikon appears to be working wonderfully. Canon lens/Nikon body adapters are non-existent as far as I can tell. I bought a Fotodiox Canon EOS set of inexpensive extension tubes in order to use its lens mount ring. On the reverse side of this ring I super glued the thin metal ring which fits into the Nikon body, unscrewed from a Nikon adapter which I happened to have on hand. I used Gorilla Glue, a gel-type super glue which gives just a bit more working time in order to center the ring as accurately as possible. In my case the adapter ended up being 13mm thick.





The drawback of the arrangement is the I have no control over aperture of the MP-E which I'd prefer to have at about f/3.5 rather than wide open at f/2.8 in order to eliminate some of the false color. Joaquim mentioned to me that he'd heard about fitting the MP-E onto a Canon body, setting the desired aperture, pushing the DOF preview button and immediately removing the lens -- it's supposed to then stay at the set aperture. Anybody know whether this is the case?


Since my stacks tend to lose much of the fidelity of the source images (need to learn how to improve this), they were sharpened and de-noised during batch conversion prior to stacking.
Apologies for the variations in illumination.
Unless otherwise stated all stacks are DMap.

Click through to flickr on the FF images to view 100% pixels.

Mitutoyo 10X, 320-grit emery sandpaper ... beautiful.


Mitutoyo 6X, 320-grit emery sandpaper



100% pixels:

Mitutoyo @10X
Center, corner
Mitutoyo @6X
Center, corner



MP-E 5X, 320-grit emery sandpaper


Stacking this showed interesting effects. I find that, especially with the lens wide open, 5X has quite a bit of false color, both longitudinal and lateral.
It's not too bad when stacked with PMax, but really shows up with DMap.

PMax, DMap -- just off-center


MP-E 1X, 320-grit emery sandpaper



Mitutoyo 10X, 120-grit silicon carbide


Same field with shallow DOF, 23-image stack


Mitutoyo 6X, 120-grit silicon carbide



MP-E 1X, effective f/5.6, 120-grit silicon carbide ... outstanding crispness.


Nikkor Micro 105mm VR 1:1, effective f/5.6, 120-grit silicon carbide



MP-E vs. Nikkor


As can be seen the Nikkor has a slightly smaller scale. Likely this is due to the MP-E adapter working as an extension tube.

I've sold the Nikkor, but plan to rent one for a day and re-do some test shots ... I'd never been pleased with mine, but seeing the results here I simply can't believe that all 105mmVR lenses are really this bad, I must have just ended up with a bad copy.

[EDIT 8.20.14] After selling the Nikkor I bought again the first macro lens I'd had, a Tokina 100mm f/2.8 AT-X PRO. It had always been a very sharp lens, and indeed I again find it outperforming the Nikkor which I had, but testing now I see that it's pretty bad with false color. So I sent it back, did more research and ended up getting a Sigma 150mm Apo OS. So far my impressions are excellent, and its color correction is indeed outstanding ... it's the first 1:1 macro I've used so far on the D800E which appears capable of allowing the camera to reach its resolution potential. I'll post a follow-up review and comparison soon.


I do love the D800E ... the enormous image files are a lot of work, and the camera is extremely demanding, revealing troubles not so easily visible with lower resolution.

[rjlittlefield]

As I had bought my previous Canon 6D largely in order to be able to use the MP-E 65mm, I'm happy to say that hacking the lens to fit the Nikon appears to be working wonderfully.

The results look quite good. I would not describe this as "hacking the lens", however, because if I understand correctly the lens itself is untouched. I see what you've created as being a DIY adapter.

The drawback of the arrangement is the I have no control over aperture of the MP-E which I'd prefer to have at about f/3.5 rather than wide open at f/2.8 in order to eliminate some of the false color. Joaquim mentioned to me that he'd heard about fitting the MP-E onto a Canon body, setting the desired aperture, pushing the DOF preview button and immediately removing the lens -- it's supposed to then stay at the set aperture. Anybody know whether this is the case?

Yes, this works. The trick is to have the lens stopped down when it is twisted out to remove. So the sequence is: camera on, aperture set, DOF preview pushed and held, lens lock pressed and held, lens twisted and removed. No one has ever found a warning about this damaging lens or camera, and every report that I've read says it works fine. Certainly it works with my MP-E and T1i.

Your observation about false color with the MP-E agrees with what I see. In comparison the Mitutoyo objectives are almost completely free of false color.

--Rik

[papilio]

The results look quite good. I would not describe this as "hacking the lens", however, because if I understand correctly the lens itself is untouched. I see what you've created as being a DIY adapter.

Thanks Rik, I'm pretty excited by all of the new equipment and am having a blast. So happy about being able to continue using the MP-E.

[Edit] You're right of course, I removed "hacked".

Yes, this works. The trick is to have the lens stopped down when it is twisted out to remove. So the sequence is: camera on, aperture set, DOF preview pushed and held, lens lock pressed and held, lens twisted and removed. No one has ever found a warning about this damaging lens or camera, and every report that I've read says it works fine. Certainly it works with my MP-E and T1i.

Your observation about false color with the MP-E agrees with what I see. In comparison the Mitutoyo objectives are almost completely free of false color.

--Rik

Excellent, thanks again! The MP-E is quite good at lower magnifications, making the Nikkor look pretty pathetic, but closing down the effective aperture should help the higher end. Close inspection shows rather serious color error though even at 1X (I suppose I was quickly spoiled by the amazing Mitutoyo).

[papilio]

Joaquim pointed out that I was being unfair to the 105mm Nikkor by shooting the stack at f/5.6, and indeed this f/16 stack is vastly improved. Not quite as crisp as the MP-E is at f/5.6, due likely to the increased diffraction.

But the fact that the MP-E can perform so well at effective f/5.6 would seem to say quite a lot about the quality of the lens, despite its greater problems with false color.

[rjlittlefield]

Just checking... I know that you've labeled both images in your first post as "effective f/5.6". But did that really mean that you were running the MP-E at its wide-open "f/2.8", thus becoming f/5.6 at 1X? And you were setting the D800E on "f/5.6" because it works directly in effective aperture? Or is the Nikkor one of the older lenses that has to be corrected for magnification too?

--Rik

[papilio]

Just checking... I know that you've labeled both images in your first post as "effective f/5.6". But did that really mean that you were running the MP-E at its wide-open "f/2.8", thus becoming f/5.6 at 1X? And you were setting the D800E on "f/5.6" because it works directly in effective aperture? Or is the Nikkor one of the older lenses that has to be corrected for magnification too?

--Rik

The Nikkor shows correctly on the camera display as Nikons do with modern lenses. The MP-E though is a 'non-cpu' lens, for the time being wide open at f/2.8 and so at 1X was also f/5.6.

Did I get that right?

[rjlittlefield]

Just checking... I know that you've labeled both images in your first post as "effective f/5.6". But did that really mean that you were running the MP-E at its wide-open "f/2.8", thus becoming f/5.6 at 1X? And you were setting the D800E on "f/5.6" because it works directly in effective aperture? Or is the Nikkor one of the older lenses that has to be corrected for magnification too?

--Rik

The Nikkor shows correctly on the camera display as Nikons do with modern lenses. The MP-E though is a 'non-cpu' lens, for the time being wide open at f/2.8 and so at 1X was also f/5.6.

Did I get that right?

That sounds right to me. As a cross-check, did you get the same brightness at same exposure time / flash power for both lenses?

--Rik

[papilio]

Just checking... I know that you've labeled both images in your first post as "effective f/5.6". But did that really mean that you were running the MP-E at its wide-open "f/2.8", thus becoming f/5.6 at 1X? And you were setting the D800E on "f/5.6" because it works directly in effective aperture? Or is the Nikkor one of the older lenses that has to be corrected for magnification too?

--Rik

The Nikkor shows correctly on the camera display as Nikons do with modern lenses. The MP-E though is a 'non-cpu' lens, for the time being wide open at f/2.8 and so at 1X was also f/5.6.

Did I get that right?

That sounds right to me. As a cross-check, did you get the same brightness at same exposure time / flash power for both lenses?

--Rik

Yep. I used the studio strobe, all manual -- no changes between lenses.

[Chris S.]

Michael, what a fascinating post! You covered a lot of ground, here—if you had used three posts to do the work of this one, they would all be worth reading.

Finally—finally—the world gets to see a trial of the Canon MP-E 65 on a short tube, with mechanical adapter, on a Nikon body. I’ve long wondered about this possibility. Perhaps I’ve missed it, but have done online searches several times over a few years, looking for anyone who has tried and documented this simple approach. So to my knowledge, you are the first. Bravo! Lots of Nikon users have lamented being unable to use this lens, and I’ve wondered about the possibility of adapting a short extension tube with a mechanical Nikon mount, just as you’ve done here. With most lenses, we would expect such an adapter to work close-in, but deprive the lens of the ability to focus at infinity. But of course, with a dedicated macro lens, infinity focus is a non-issue. On the other hand, the MP-E 65 is an optical oddball, so I’ve wondered if it might prove intolerant of even a short added extension tube. As you work with your adaptation further, I hope you’ll continue to share results.

(While I’d enjoy experimenting with the approach, I already have solid options in the 1-5x range, and could not justify the cost of an MP-E 65 lens just to satisfy my curiosity about creating a simple Nikon adapter for it. So I’m very happy to see the question addressed by you. )

As to the inability to adjust the MP-E 65’s aperture on a Nikon body, this issue has long struck me as tractable to modest reverse engineering. Surely somebody who has Canon equipment and a little electronics knowledge could determine how a Canon body actuates the iris. And with this information, it should not be difficult to make an inexpensive controller that would give an adapter like yours the capability to adjust the MP-E 65’s iris.

A while back, I enjoyed a very productive conversation with the founder and CEO of Fotodiox, discussing at length a quite different product his company had introduced. He impressed me as the sort entrepreneur the world needs more of: A fellow looking for problems photographers have that he can solve with innovative, high-quality products. I’ve often considered asking him if he would consider making an adapter that would permit the Canon MP-E 65 to be used with other camera brands, with full aperture control. I’ve hesitated to pick up the phone because of two questions I think he’d ask, whose answers disturb me. One: “Would you use and test this product?” I’d have to answer “no”—I’m not the target market, don’t own an MP-E65 lens, and don’t wish to buy one. Two: “How many of units do you think we’d sell, and what price do you think photographers would be willing to pay?” To this, I have no idea. (Lest I give the wrong impression, this gentleman did not strike me as greedy, but a fellow who delights in delivering solid value—but no company can make and sell a product without generating an honest profit sufficient to remain in business.) If anyone can shed light on these questions, I’m all ears.

Very interesting, getting to know the new D800E. I have the feeling that it's too much camera for me, but I'll get used to it. Be warned, batch converting, stacking and even post take forEVER.

You probably need a new computer—or at least a thorough servicing of the computer you have. A modern computer in clean, efficient condition should quickly process D800e files.

Your images of the 320-grit emery sandpaper also contain a skin flake, and what looks to me like an attached textile fiber. These are very common components of house dust. To my eyes, they add information, and I’m glad they are there.

To my eye, your demonstration of 10x Mitty at 6x shows an optical disaster. I suspect that the 5x Mitty, used at 5x, would look much better. Do I presume correctly that you used the Raynox DCR-250 (125mm focal length) for this?

Your demonstration center vs. corner performance for the Mitty 10x, shot at 10x, tends to confirm my current preference to use Mitutoyo objectives on DX (APS-C) sensors, rather than full frame. This approach strikes me as using the highest quality portion of the objectives’ image circle. It also (with current Nikon camera bodies) packs the highest number of pixels within this circle.

Again, thanks for a compelling and information-rich post!

Cheers,

--Chris

[liuto]

Which thread diameter do the Fotodiox tubes have?
If it is 52mm it would be possible to pair them with the bajonett part of the Nikon K-rings. This way you have a more reliable connection than with super glue.

Regards
Hermann

[Yousef Alhabshi]

Finally!! Someone succeed using the MP-E65 on a Nikon body! Well done Michael

I did some tests in order to see if my D800E would adapt the MP-E65 with no success. Although I sold my Canon body, I kept the MP-E65 so I can use it with my Sony Nex camera -Haven't bought Nex-Canon adapter with aperture control yet!-


Chris
You can check with those who would like to be beta testers of the adapter & have no issue paying for the beta solution. I would love to be one of those since I have both the lens & a Nikon body

[ChrisR]

Congrats indeed Michael!
Much as Chris_S, I have other means to cover the range on a my Nikon, and having a Canon body appealed to my laziness..

I got as far as finding extension tubes with exposed connections, and scouring around finding info on how the data interchage works. It's moderately well explained, but has evolved so is unlikely to stay fixed. It's quite a complex serial bitstream which would need a programmable box to talk to the camera as well as the lens, so not very appealing, and not without risk of breaking one part or the other during development. Though something to just set the aperture electronically would be a lot simpler.

Which Nikkor 105mm are you using here? There's the AF-D and the VR, which are quite different designs. I have read that the former is better, (but you have to be careful about what you read..).

If Fotodiox were to make an adapter, an obvious addition would be a manual aperture. It would be in the "wrong" place, but worth a test.
If the D810's reported Electronic Shutter opening turns out to be really vibration free, the combination will have more appeal.

--

The other only slightly oddball thing to do is turn the MP-E around. The desined-for distances are workable for some of the magnificatin range, and would doubtless stretch a little.

--
Hermann - no. I'm pretty sure Fotodiox tubes aren't 52mm .

[papilio]

Thanks everyone for the kind comments.


I'll reply at length in a bit, I'm currently re-doing a number of the stacks beginning with RAW conversion so as to correct several stacking and processing mistakes I made, the new results will hopefully give more accurate impressions of the camera and lenses.


Just a quick response to Hermann,

Which thread diameter do the Fotodiox tubes have?
If it is 52mm it would be possible to pair them with the bajonett part of the Nikon K-rings. This way you have a more reliable connection than with super glue.

Regards
Hermann

That would certainly be ideal, sadly neither the Nikon nor the Canon Fotodiox ext tubes (each of which is of a different diameter) were found to have standard step-ring threads.


Also, and this may tie in with some posts made on the thread http://www.photomacrography.net/forum/v ... 947#147947

... Rik, do you recall my surprise at finding to be correct your claim that the Canon JPEG and CR2 files were nearly identical? It now appears that Nikon has quit going nuts with the in-camera processing of their JPEGs, in fact I'm seeing the reverse to a fairly subtle extent. Without opening the NEF file with CaptureNX2 but rather by taking screen captures of the two files opened with irfanview, the NEF appears to be the sharper.

JPEG vs. NEF, 100% pixels, Mitutoyo @ 10X, 120-grit silicon carbide


The superiority of the NEF image is even evident on thumbnails. btw I use a custom version of the 'Neutral' Picture Control with the sharpening turned off.


I recall reading the comment made in DPReview's discussion of the D800E's files on the 'RAW Performance' page ...

"At base ISO, raw output from the D800 and D800E offers even more cleanly resolved detail than the camera's default JPEG processing. Resolution advantages of the D800E over the D800 are less perceptible here, implying that Nikon is applying more in-camera sharpening to D800E JPEGs."

[Chris S.]

Chris
You can check with those who would like to be beta testers of the adapter & have no issue paying for the beta solution. I would love to be one of those since I have both the lens & a Nikon body

Yousef,

You are, so far, the one and only person who has said he would purchase an adapter that permits the Canon MP-E 65 lens to be mounted on a Nikon body with aperture control. As I'm sure you understand, I can't make a good-conscience business case to Fotodiox, or any other manufacturer, on the basis of just one prospective buyer. But thanks for your response! If others feel as Yousef does, please PM me. I'll be happy to make an appeal to this manufacturer, if I receive enough affirmative responses to honestly suggest to him that development and manufacture of such an adapter might be a business risk with a positive expectation.

Cheers,

--Chris

[Steve S]

While I certainly don't think it is going happen, there might be some modest demand for a shallow adapter providing aperture control of Canon EF lenses on F-mount cameras. It could be used with many EF-mount macro lenses, not just the MP-E. AF and IS function might be nice, but not required.