What am I doing wrong?

[Sally]

I started off using a Photar 25mm, but after reading on the forum that much better results could be obtained with a microscope objective - I got one on eBay - it's a Nikon PlanApo 10/0.45 160/0.17 I use it on bellows - so I know I needed a finite objective, and ideally I should've got one that didn't have a cover glass thickness on it - but I also read that at low magnification this wouldn't make much difference.
Using the objective and the photar in the same holder (so difference in internal reflections should be discounted) - why is the photar giving me sharper results, and why can't I get a decent image with the objective - is it the cover glass - or did I get a duff objective?
I've attached a couple of (edit) links to pictures to show the difference:
http://www.sallyswebsite.com/Leitz%20Photar%2025mm.jpg
http://www.sallyswebsite.com/10x%20plan%20APO.jpg

[Planapo]

Sally,

Though the PlanApos are highly corrected (and hence very expensive) objectives and are excellent for use on the microscope in conjuction with their matching (photo) eyepieces, for direct projection on the sensor the ones for use with a cover glass seem not to be ideal. That is because in contrast to the not so highly corrected objectives most of us here use for direct projection, those PlanApos like yours that were designed to be used with a cover glass, are more sensitive to cover class thickness and usually have a very short working distance which makes lighting tricky.

Moreover, it would be helpful to know what Nikon PlanApo you are referring to (you could post a photo of your objective) as there are older ones that needed a compensating eyepiece for best results. And apart from that, it may be advisable to check your PlanApo on a microscope with a matching eyepiece to see if it then works properly, to make sure that it's not damaged.

--Betty

[Sally]

Thanks for the quick response Betty - I was afraid that would be the case with the cover glass
I come from a photographic background (and love my APO lenses) - I have no knowledge of microscopes - I can't understand why a thin flat piece of glass in front of something would alter how the objective performs.

Many thanks,
Sally

[Sally]

Here's a couple of quick photos of the objective:
http://www.sallyswebsite.com/Objective1.jpg
http://www.sallyswebsite.com/Objective2.jpg

[Craig Gerard]

Sally,

Thanks for uploading the images of the objective as Betty suggested; that definitely removes one obstacle in addressing your concerns.

Very nice objective! That is a CF if I am not mistaken. Corrects for CA within the objective without requiring compensating eyepieces.

You've mentioned you are using the Apo and the Photar in the same 'holder'; could you expand upon this; are you referring to a microscope turret or a microscope objective (RMS thread) to camera adapter?

I can see via the EXIF you are using a Nikon D700. Could you explain your setup. Bellows, extension tubes, and the amount of extension (if any) etc? Also, what f stop or aperture setting are you using on the Photar? Working distance with the Photar will be approximately 25mm. What is the working distance with the Apo objective?

To my eyes, comparing the two sample images, the Apo objective is delivering a superior result. Focus stacking will unveil its true potential.

Craig

[rjlittlefield]

Sally, if your demo image is most of a frame, then the degradation seems so severe that I wonder if the whole problem is due to lack of cover glass.

However, if the problem is just missing cover glass, then there is a simple "fix" that may solve your problem.

Just obtain a cover glass of the proper thickness (0.17 mm) and tape it flat across the end of the objective.(!)

I can't understand why a thin flat piece of glass in front of something would alter how the objective performs.

The answer to this question hints at why the fix works.

When you look through a piece of glass at an angle, each light ray gets shifted sideways by an amount that depends on its own angle with respect to the glass. For each point on the subject, a microscope objective such as yours accepts light rays from a wide range of angles, from on-axis to almost 27 degrees off axis. To form a proper image, all of those incoming rays must be redirected to a single point on the sensor. The lens is designed to do this when the incoming rays have been shifted by an amount corresponding to passing through 0.17 mm of glass. When the glass is not present, the incoming rays are not shifted as expected, so they hit portions of the objective that have the wrong slant. As a result, rays from a single point on the subject get directed to multiple points on the sensor. Similarly, rays from many points on the subject get directed to a single point on the sensor. Either way you think about it, the result is blur.

Now it turns out that the amount of shift does not depend on the position of the glass. The usual case is that the cover glass is in contact with the specimen, but in theory exactly the same shifts occur if the cover glass is in contact with the objective or anywhere in between. I say "in theory" because the analysis applies only to cover glass that is perfectly planar and uniform. In practice, imperfections in the cover glass will have more effect when the glass is farther away from the subject. However, even an imperfect cover glass at the objective should be much better than no cover glass at all.

The reason I have some niggling doubts that your problem is entirely due to cover glass is that I have experimented with this effect using even higher power and wider aperture objectives, 40X at NA 0.65. The effects of cover glass were obvious in that case, but they were not as extreme as what I think I see in your image. (See HERE for illustration.)

Still, the experiment is cheap and easy to do if you can lay hands on some cover glass, and the results should be useful no matter which way it comes out.

Oh, there is another aspect that is important and I don't see mentioned in your post: the amount of extension. Your 160/0.17 objective is designed to be used at exactly 150 mm distance between the sensor and the shoulder of the objective's mounting threads. More or less than this, and the image will degrade. The degradation is worse for larger NA, and at higher magnification, and seems to be worse for some objectives than others. But in any case, the best image should occur at or near the design point. Differences of a few mm will make no significant difference, but running the objective at say 100 or 200 mm instead of 150 would be risky -- maybe OK, maybe not.

I hope this is helpful. Please let us know any of your results.

--Rik

[Sally]

Thanks Rik and Craig!

When I said 'holder' I meant the RMS to M39 (or is it M42) mount adapter that came with the Photar, and then an f-mount adapter to Nikon PB4 bellows.
The Photar seems to give the best results at f/4 ish, and I've been trying both at different extension amounts - I'll try 150mm for the APO and order a cover glass to tape over it - fingers crossed.
The working distance for the APO is a lot less - about half a centimetre.
Thank you for the explanation regarding the cover glass Rik.
I shall most definitely let you know how I get on.
Thank you all so much,
Sally

[ChrisLilley]

I started off using a Photar 25mm,

Is that the Leitz Photar 25mm f/2.5?
http://macrolenses.de/ml_detail.php?ObjektiveNr=103

but after reading on the forum that much better results could be obtained with a microscope objective - I got one on eBay - it's a Nikon PlanApo 10/0.45 160/0.17 I use it on bellows - so I know I needed a finite objective, and ideally I should've got one that didn't have a cover glass thickness on it - but I also read that at low magnification this wouldn't make much difference.

Very nice objective from what I read, highly corrected, but the large aperture means the depth of field is very shallow. Depending on how you are focusing, that might be a problem.

Using the objective and the photar in the same holder (so difference in internal reflections should be discounted)

Reflections from the holder, yes. Reflections from the shiny metal ends of the objective, no.

- why is the photar giving me sharper results, and why can't I get a decent image with the objective

Leaving aside the issue of the cover glass, to which others have responded, another difference is the depth of field with the two systems.

Using your D700 pixel pitch to get a circle of confusion of 30 μm, and seeing that you are using your D700 on a PB-4 which has a maximum extension of 180mm, the total maximum sensor to objective distance is 46.5mm+180mm = 226.5mm. (More if you are using a spacer ring between the D700 and the PB-4).

With the Photar, that will give an effective aperture of f/22.5 and an optical magnification of 8.0 at maximum extension, with a depth of field of 21 μm. Stopped down to f/4 on the lens, gives an effective aperture of f/36 and a depth of field of 34 μm.

With a 10x 0.45NA objective, at a sensor to objective flange distance of 150mm (160mm tube length - 10mm eyepiece offset) i.e 103.5mm extension on the bellows, yields an optical magnification of the rated 10x, an effective aperture of f/12.2 and a depth of field of only 9.4 μm. Which is barely touched by diffraction so should give a very sharp and well corrected image - on the tiny sliver which is within the depth of field.

Depending on what focusing rail you are using and its resolution, you might be missing the plane of sharp focus entirely.

Depending on what extension you are using, the objective might be out of its design range for magnification.

[ChrisLilley]

According to this chart, variations from 0 to 0.15mm coverglass thickness have minimal effect on contrast at NA 0.4 so the effect at 0.45 should be modest. At higher NAs the effect is clearly significant. This chart is shows a thickness error up to 0.15, so 0.17 (i.e. none) should be similar


image credit: Olympus microscopy resource center

The graph does not show other effects though, just contrast.

[rjlittlefield]

Depending on what focusing rail you are using and its resolution, you might be missing the plane of sharp focus entirely.

True in general, but I'm presuming the subject shown in Sally's test image athttp://www.sallyswebsite.com/10x%20plan%20APO.jpg has an overall spherical shape, in which case the focus seems clearly indicated as a ring of high contrast between center and edge.

I've studied that image at the highest resolution posted (1,151×1,242 pixels), and I can't find anything sharp in it.

On the other hand, Craig wrote that "To my eyes, comparing the two sample images, the Apo objective is delivering a superior result" so maybe I'm missing something.

Craig, can you give me coordinates of some sharp feature in the APO image that I should be looking at?

--Rik

[ChrisLilley]

The working distance for the APO is a lot less - about half a centimetre.

Sorry for multiple posting. I keep coming back and thinking of something else.

In a Nikon CF objectives brochure, the working distance for the CF Plan Apochromat 10x NA 0.45 is given as 2.75mm.

[ChrisLilley]

Depending on what focusing rail you are using and its resolution, you might be missing the plane of sharp focus entirely.

True in general, but I'm presuming the subject shown in Sally's test image a thttp://www.sallyswebsite.com/10x%20plan%20APO.jpg has an overall spherical shape, in which case the focus seems clearly indicated as a ring of high contrast between center and edge.

Certainly on the other image posted, there is a clearly visible, high contrast ring on the eye indicating the plane of focus.

Then I wondered if that ring was a reflection of some part of the camera system, and just happened to be in focus with the Photar. In that case it would show up in all images.

I've studied that image at the highest resolution posted (1,151×1,242 pixels), and I can't find anything sharp in it.

I don't either. That is why I wonder if the subject is simply too far away for the extremely small working distance on this objective.

[rjlittlefield]

Certainly on the other image posted, there is a clearly visible, high contrast ring on the eye indicating the plane of focus.

Then I wondered if that ring was a reflection of some part of the camera system, and just happened to be in focus with the Photar. In that case it would show up in all images.

Good point, that the ring of high contrast could be a feature of the subject & illumination and not the focus. I don't recall considering that possibility.

On the other hand, what I really use those high contrast regions for is to guide where to go looking for detail. In the APO image, the sharpest things I found are for example at x=180,y=706, near what I think is the front edge of the high contrast zone. Stuff there seems much sharper than anywhere else, for example the edge and the center of the eye are completely fuzzed out. So even after another look, I'm inclined to think the image does include best focus. It's just not very sharp.

--Rik

[Sally]

1. I've acquired a cover glass and taped it on (not an easy feat in itself!).

2. I've coloured all the shiny bits on the front and back with a black marker pen - it's not matt, but it's not as reflective.

3.I've lined up the circle with a horizontal line through it on the camera (the focal plane - is that where the sensor is?), with 0 on the bellows, and the end of the thread/beginning of the body of the objective 150mm away - does that sound right?

4. I've changed the lighting too (needs more work) - I think the original set up was giving me flare / glare.

Chris the Photar is the 25mm f/2, not the f/2.5 - it looks a bit more modern, and I've picked up a linear stage - I think if my maths is right, the micrometer on it will do 0.01mm graduations - at a pinch, I could do half a graduation - but I've not tried stacking with this objective yet - I've just been trying to get something in sharp focus!

Anyway, I think I'm getting somewhere - I have bits in focus:
http://www.sallyswebsite.com/SAL_0677.jpg

This is a 100% crop of the bits in focus.

I'll report back when I get the lighting right and manage a stack.
Thanks for all the help so far
Sally

[ChrisLilley]

3.I've lined up the circle with a horizontal line through it on the camera (the focal plane - is that where the sensor is?), with 0 on the bellows, and the end of the thread/beginning of the body of the objective 150mm away - does that sound right?

Yes.

Chris the Photar is the 25mm f/2, not the f/2.5 - it looks a bit more modern,

Ah, okay.

and I've picked up a linear stage - I think if my maths is right, the micrometer on it will do 0.01mm graduations

That sounds more like it, for the DOF of that lens.

Anyway, I think I'm getting somewhere - I have bits in focus:
http://www.sallyswebsite.com/SAL_0677.jpg

This is a 100% crop of the bits in focus.
Sally

That is looking a lot better, to me.