Hairy eye of Painted Lady, stereo with CamRanger and 4X obj

[rjlittlefield]

Nothing fancy here, just a crossed-eye stereo showing the remarkably hairy eye of a Painted Lady butterfly (Vanessa cardui):



Discoloration in the eye is due to several years of drying in this old specimen.

The setup for this shot was pretty simple: just a 4X objective (Nikon CFI BE) stuck on the end of a 55-200 mm telephoto lens on a Canon T1i camera. Focus stepping by AF motor control, using the CamRanger. (I couldn't resist assembling a stack of "supporting documentation".)



Here's the entire frame. There is some vignetting because I zoomed out the telephoto in order to slightly widen the field. EXIF data reports 155mm, implying about 3X actual magnification.



Focus stepping is 71 frames at the smallest increment of the CamRanger. The small increment is required despite effective f/15 because the telephoto is front-focusing. The smallest increments of focus are very small when the telephoto is used by itself, but they advance the front of the lens by an amount that is significant compared to the roughly 50 micron DOF of an NA 0.1 objective.

The CamRanger currently only allows 50 frames per stack, so I just ran one stack of 50, then immediately another stack to get the rest of the required frames. The second stack ran into the infinity focus stop of the lens and produced a bunch of identical frames, so I truncated the stack before processing it.

ISO 100, 1/8 second. The CamRanger requires that the camera be in live view mode, so with Canon cameras that automatically implies EFSC.

Zerene Stacker PMax, no retouching. The stereo is +-2%.

BTW, I went looking for an SEM photo of the eye structure, but all I could find was a closeup of a few ommatidia and the bases of some hairs. If somebody can find one comparable to the overview shown here, I would appreciate the link. Thanks!

--Rik

[Saul]

Hi Rik,
How much time takes for CamRanger to take these 50 photos using AF motor ? Very nice setup by the way, worked out perfectly
Thanks,
Saul

[rjlittlefield]

According to timestamps on the image files, it was 170 seconds between frame #1 and frame #50, so about 3.5 seconds per frame.

Yes, this simple setup worked very well with the camera on a stack of books. I tried also with the camera on a flimsy tabletop tripod, but that had too much undamped vibration and did not give sharp pictures unless I shot with flash.

--Rik

[KimP]

Hi Rik,

This is a very nice and relatively simple setup using the AF focus of the relay lens controlled by the Camranger

I have a few questions:

1) Is the microscope objective a finite or an infinite objective (does it matter)?

2) Which focal length would be optimal for the relay lens?

3) Have you refined/changed the method since the original post.

4) I have the old Camranger. Do you know any advantages of the Camranger 2 in addition to the ability to set the start and end points and automatically calculate the step?

5) What would the minimum step length be (probably depending on the actual lens)?

Best wishes,
Kim

[rjlittlefield]

Hi Rik,

This is a very nice and relatively simple setup using the AF focus of the relay lens controlled by the Camranger

I have a few questions:

Kim,

I apologize for late reply -- this one slipped through a crack in my reminder system.

You asked:

1) Is the microscope objective a finite or an infinite objective (does it matter)?

The objective is infinite, but at this low magnification it does not matter much. At higher magnifications such as 10X-50X objectives, infinite becomes progressively more important.

2) Which focal length would be optimal for the relay lens?

For most purposes 200 mm is good. Some objectives and some relays work well with a shorter relay, down to around 100 mm. The sharpest 5X setup in my kit consists of a 10X objective on front of a Canon EF 100mm f/2.8L Macro IS USM focused from infinity going shorter. See AF motor focusing with a microscope objective for some discussion of that setup.

Infinite objectives give their rated magnification at one specific focal length for the relay lens. That is 200 mm for Nikon objectives. Using longer or shorter relay lenses simply increases or decreases the magnification, in proportion to the focal length of the relay. This process is limited on the long end because you get empty magnification, and on the short end because you get vignetting and/or aberrations in the corners of the image. The extent of vignetting depends strongly on unpublished details of the relay lens, so as a matter of practice you have to just try it, or find a report by somebody else who has tested the exact model of lens that you have. Zooms are particularly prone to vignetting.

3) Have you refined/changed the method since the original post.

Not really. The big change since then is that many cameras now include an in-camera facility for shooting a focus stack, so the Camranger is not needed to automate that part of the process. Some people shoot a lot of stacks this way, especially at lower magnification in the field. Most of my own work is done in studio at higher magnifications, with stack lengths and optics for which the AF motor approach does not work well. For me the setup shown here was mostly useful for real-time demo at a conference.

4) I have the old Camranger. Do you know any advantages of the Camranger 2 in addition to the ability to set the start and end points and automatically calculate the step?

I do not have any experience with the Camranger 2, sorry.

5) What would the minimum step length be (probably depending on the actual lens)?

If the relay lens changes physical length with focus, so the objective moves with respect to the subject, then the minimum step size will depend on that, and yes it will vary a lot from one relay lens to another. The particular cheap zoom tele shown here works OK with a 4X objective, but if I recall correctly it gives focus banding with a 10X objective. On the other hand, if the relay lens does not change physical length (internal focus), then the minimum step length is always short enough, even if with a high power objective "short enough" means just a couple of microns. This is because the objective turns a subject that is physically small and close, into one that is virtually large and far, and then the relay lens just steps focus through the large/far virtual subject. Any lens with AF must be capable of focusing well enough, at any position in its focus range, or its users would be unhappy and the lens would be hard to sell. Focus-stepping with the AF motor just takes advantage of this same capability.

Again, see AF motor focusing with a microscope objective for a longer discussion.

--Rik

[KimP]

Rik,

Thank you very much for the detailed and very useful answers.

I tried the technique with a 70-200 lens, that does not change physical length, as a relay lens for an El-Nikkor 50/2.8. It worked quite well, but I found that the maximum stack depth is limited. I can estimate the maximum stack length by turning the focus from closest focus to infinity on the lens. I wonder whether there is any way to calculate this?

I am waiting for a Nikon BE 2 Plan 4x/0.10 infinity objective to arrive and will continue the testing.

My goal is to use the camera's built-in facility for focus stacking in the field, and thus avoid bringing a heavy automated rail system.

Kim

[rjlittlefield]

I tried the technique with a 70-200 lens, that does not change physical length, as a relay lens for an El-Nikkor 50/2.8. It worked quite well, but I found that the maximum stack length is limited. I can estimate the maximum stack length by turning the focus from closest focus to infinity on the lens. I wonder whether there is any way to calculate this?

Yes, limited stack length is a serious problem with this method, especially when using rear lenses that do not focus very close. In the other thread that I linked, the rear lens was a macro lens that of course will focus very close. That allows a longer stack, at the cost of assorted other weirdnesses like big change of scale from start to end of the stack.

Regarding calculation, I tested with some of my own lenses using a standard math model. The calculation did not match what the actual lenses did, and I could not find the source of the discrepancy. I suspect it would require information about the inside of the lenses that is never available. For what it's worth, your Nikon infinity 4X objective will probably act very much like the EL-Nikkor 50/2.8, because both of them are fixed 50 mm focal length.

--Rik