Poor man's Repro-Nikkor (f/1.0, m=1x)

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: ChrisR, Chris S., Pau, rjlittlefield

Lou Jost
Posts: 4423
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Curt, I had no idea that lens had such a wide field. How is the image quality? I hear lots of things about blue or purple fringing on non-Apo Nikon objectives.

I just tested my 4x Plan Apo. Wow!! Amazing image quality but small image circle. Even with a 90m tube lens the corners are mushy.

Lou Jost
Posts: 4423
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

My "poor man's Repro-Nikkor" is an answer to a challenge Rik made to me during our discussion of the advantages of coupled lenses and the capabilities of small sensors.

I calculated that the sweet spot for the Oly hi-res mode would be at EA=4; this would minimize aberrations and should be just on the edge of showing diffraction. Now at last I am ready to test this combination lens at EA=4. I shot through three filters this time (between the rear lens element and the subject), as I now suspect that the Oly 50mm f/2.0 macro lens for Four Thirds sensors is designed to expect the same 4mm thick sensor filter stack as a Micro Four Thirds sensor. I cut a paper iris that lowered the exposure exactly two stops below the no-iris exposure with the lenses wide open; I think this should be EA=4.0.

With this smaller iris there was virtually no CA anywhere.

Unlike the high-res shots I posted this morning, which were made from 80 Mp RAW files, this time I am using high-res mode jpgs direct from the camera with no sharpening. Olympus resizes these to 8160 x 6120 pixels, figuring that there is some redundancy in the 80 Mp files. I stacked them directly in DMap and retouched them from PMax. I did no sharpening before or after stacking, and no correcting CA or exposure. The results would be much better if I had used the RAW files, but that also complicates things and adds variables.

Here is the resulting image, resized for the forum (this is a very small Riodinid butterfly wing at life size on a 17mm sensor):
Image

Here is a 100% crop (the color of the scales is iridescence from grazing light, not CA):
Image

Here is a 200% crop:
Image

Again, this is with no manipulation or sharpening on my part. There is no degradation in the edges and almost no degradation in the corners. I am very pleased with these results. This is a good demonstration of the big advantages of stacked lenses versus lenses on extension when m is in the range of 1-3.

ChrisR
Site Admin
Posts: 8546
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

I have a cfi 4x 0.1 I think, which doesn't have a big image circle. Removing the front cover reveals a longish lens construction underneath.

There's a picture of it here somewhere - I think..
Found it - 6 years ago:
Image

The "BE" 4x infinites DO have a huge IC, though, especially when you take the front off.
Rik's:
Image
Chris R

Lou Jost
Posts: 4423
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Chris, thanks. I'm unfamiliar with the BE objective and searching for it isn't turning up anything ('BE' being too common a word, I suppose). Is it a Nikon objective?

rjlittlefield
Site Admin
Posts: 20855
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

Lou Jost wrote:My "poor man's Repro-Nikkor" is an answer to a challenge Rik made to me during our discussion of the advantages of coupled lenses and the capabilities of small sensors.
And it's looking like a very good answer.

The essence of my challenge was to demonstrate high resolution, all across the sensor, using lenses that are common enough that the setup can be reproduced routinely.

If I understand correctly what you've done, then it looks to me like you've satisfied all those criteria. The darkened corners still need some work, but if they're sharp, then I'm pretty comfortable the darkening could be removed with some simple correction.

Nice work!
I'm unfamiliar with the BE objective and searching for it isn't turning up anything
That's a Nikon MRN70040. See http://www.photomacrography.net/forum/v ... 380#101380. My picture that ChrisR references comes from http://www.photomacrography.net/forum/v ... 444#101444 .

--Rik

curt0909
Posts: 607
Joined: Wed Oct 26, 2011 10:06 am
Location: Pittsburgh, PA

Post by curt0909 »

Superb results, Lou. Yes, the Nikon achromats do have some blue fringing. I honestly don't see much of a reason to change anything. Your current setup is working beautifully aside from some darkened corners.

ChrisR, that objective is not a 4x Plan. It is a strain free 4x Pol. The Pol versions have a different optical formula(non-Plan, shorter WD, apparently smaller FOV, etc)Also, when I tested the 4x Plan, it was important have it mounted as close to the 50mm lens as possible or vignetting would occur. I see your setup with turret mounts the objective a bit further away from the lens than is ideal. I had forgot about the BE 4x. That's a cheaper option yet.

ChrisR
Site Admin
Posts: 8546
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Thanks Curt, I was wondering. I have a couple of other Nikon 4x infinites, which are also not the CFI, :)
The lash-up pictured was in answer to a question, the extension is superflous of course.


Lou the BE is this one
http://photomacrography.net/forum/viewtopic.php?t=27205

I've posted half a dozen of them to countries where there seem to be local difficulties.

There's a silly pic with the BE, here somewhere...
Chris R

ray_parkhurst
Posts: 2655
Joined: Sat Nov 20, 2010 10:40 am
Location: Santa Clara, CA, USA
Contact:

Post by ray_parkhurst »

I question the use of a macro lens in this application, similar to my questioning their use as tube lenses for infinite objectives. There are tradeoffs in the design of a macro lens that compromise infinity focus performance in favor of close-up. Certainly the design of macro lenses is more complex than it would need to be if they were not designed for close-up work, with FL shortening and flat-field correction. If you are using the same lens on both sides of the stacked composite, then I'd expect these corrections (and the associated compromises) might be unnecessary and counter-productive for this application.

This may be obvious to all, but if you stack two 50mm lenses, don't you end up with a 25mm lens? The "tube" lens is focused at infinity, so has 50mm total extension (assuming thin lens). The "objective" is also focused at infinity, so has 50mm working distance. The composite is at 1:1, so extension and working distance are 2*FL, thus FL is 25mm. Is this correct?

If I think about the 105PN, which is also a 1:1 optimized lens, perhaps it was designed this way. Each half of the lens would be infinity-optimized and have 210mm FL.

Lou Jost
Posts: 4423
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Chris and Curt, thanks for the info and the links. I will look into those, for my friends in Ecuador as well as for me, since they are so economical.

Rik, thanks for providing the challenge! It was just what I needed to focus on this particular problem.

Ray, I agree with you in part. Since the lenses have to couple with each other front-to front, though, I thought that there may be a benefit to flat-field lenses; I can imagine that the corners of the image will be sharper if both lenses are designed to have perfectly flat object planes (though at infinity I guess all lenses are perfectly flat)...Also macro lenses are designed to be very sharp across the field while most normal lenses make sacrifices in the corners (not just field curvature but other sacrifices as well) in order to allow higher speed.

Nevertheless I chose this macro lens because it was sharper at infinity than most lenses. Published resolution tests, even of macro lenses, are usually done on distant targets, and these showed that this lens beats most other lenses. It is a really good lens.

The other lens I am experimenting with, the Sigma 70mm macro, is also better at infinity than most other lenses.

Anyway the results seem quite good. I look forward to trying out a 105PN on the same target.

Lou Jost
Posts: 4423
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

I just made a careful test of the lens shooting through filters vs without filters. There is some slight purple fringing in the corners and edges in the individual images of both stacks, but this mostly goes away when stacking except at the most contrasty points. There is more purple fringing in the corners when no filter is used. The difference is small though.

There is also slight red/green fringing in the corners on out-of-focus edges, and this is also reduced by the filters, but not eliminated. Maybe there is a correct thickness that would eliminate most of it?

ChrisR
Site Admin
Posts: 8546
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Question - if we tried this with a couple of 4x objectives, na 0.1 or 0.2 finite "160's" say, nose to nose, do we get a decent f/1 lens operating at a WD of ~150mm? The ray bundles are coming at your sensor more parallel than from the camera lens, so perhaps the thihck slab of glass would matter less.

I have no idea how many 4x objectives I could dig out, but I do have 3 at NA 0.2.

I could make pairs of 10x NA 0.45, 20x NA 0.42, 40xNA 0.5.
Also 20x NA 0.75 but arranging the limiting aperture with very short WD of the original objective would be hard.

There must be a reason why this wouldn't work...
Chris R

Lou Jost
Posts: 4423
Joined: Fri Sep 04, 2015 7:03 am
Location: Ecuador
Contact:

Post by Lou Jost »

Chris, interesting idea! The NA would have to be 0.25 to get f/!.0, but I have to stop down my "F/1.0" two stops to get rid of aberrations, and you probably don't need to do that with a microscope objective, so a 4x with NA=0.2 would still be good enough.

I think the problem will be sensor coverage. The size of the good-coverage circle on the subject with my 4x iis only about 5mm , and that would translate to 5mm on the sensor if it is reversed.

I should emphasize that for the lenses I've tested, after stopping down, the effect of the glass slab is very subtle, almost undetectable at EA of 4.0.

Post Reply Previous topicNext topic