Question about a photograph

[Chris Mower]

Hello everyone, just (re)joined the group after a rather lengthy hiatus. Looking forward to many happy hours at the 'scope after a long absence.

I have a question which the very knowledgable members of this group might be able to answer. I think I know the answer but it would be good to have confirmation, and if wrong it would just as good since I know that I have a lot to learn here.

So, to the picture. It is of an LCD screen from a defunct Mino HD camera and shows the LED's that make up the colour display. It was taken using reflected DIC using a Nikon Optiphot which has previously given scintillatingly good results. Pin sharp details with a not unreasonable DOF. What I found with this was that I just couldn't get a really sharp focus anywhere on the silicon and I was struggling to understand why when it occurred to me that it might have something to do with the glass that covers the silicon. Since the lens is a Metaloplan and is designed for use without a coverslip, could this be causing the objective to behave in a way that prevents obtaining a sharp focus?

I have tried uploading the picture to the site but it just says that the file size is too big even though I keep reducing it to the point where it is practically worthless, so I have put a reduced copy in my DropBox and am putting the links here. If anyone knows what I am doing wrong when trying to upload I would be grateful for the help.

Small version
https://dl.dropboxusercontent.com/u/341 ... SC4306.jpg

Large version (18MB)
https://dl.dropboxusercontent.com/u/341 ... 201%29.jpg

Thanks,
Chris[/url]

[rjlittlefield]

... it occurred to me that it might have something to do with the glass that covers the silicon. Since the lens is a Metaloplan and is designed for use without a coverslip, could this be causing the objective to behave in a way that prevents obtaining a sharp focus?

Yes, absolutely. Even if the glass is perfectly smooth and planar, shooting through it will introduce spherical aberration at image center and other aberrations away from center. What you have here looks as good as similar shots that I recall seeing from other people. For my own efforts and some discussion, see HERE. Note my comment that image quality got worse at 20X NA 0.40 than it was at 10X NA 0.25 . That's consistent with degradation caused by aberrations, which get a lot worse with increasing NA.

I have tried uploading the picture to the site but it just says that the file size is too big even though I keep reducing it to the point where it is practically worthless, so I have put a reduced copy in my DropBox and am putting the links here. If anyone knows what I am doing wrong when trying to upload I would be grateful for the help.

There are two independent limits: 1) image size no more than 1024 pixels on either side, and 2) file size no longer than 300 KB. If you exceed the pixel count limit, then the forum software will automatically shrink the image and then accept the shrunken version. If you exceed the file length limit, the image will simply be rejected. Your symptoms are consistent with exceeding the file length limit.

The best approach is to shrink the image to 1024 pixels maximum, then sharpen to taste, then adjust the JPEG compression quality while saving so as stay just under 300 KB. If you're using Photoshop, it helps to Save For Web & Devices rather than Save As. That's because Save As preserves all the metadata which consumes space but cannot be seen, while Save For Web & Devices strips the metadata and uses all available bytes for image detail.

It's generally a bad idea to let the forum upload software shrink the image for you. At best you cannot predict exactly what it's going to do, and almost always what it does is not as good as what you can do yourself with some thoughtful fiddling.

--Rik

[Chris Mower]

Rik,

thanks for the reply. It is good to have confirmation of my interpretation of the problem. I had been fiddling around for about 20 minutes trying to figure out why I couldn't get it to focus properly when I remembered about coverslips being essential for transmitted light work and it occurred to me that I was looking at the opposite effect so to speak.

As for the file upload problem I had seen the 1024x1024 limit but missed the file size limit (smacks forehead).

Thanks once again.

Chris

[phil m]

If you are able to find some feature of the coating glass to focus on, a scratch maybe, you will probably find perfect focus.
If anyone reading this , knows where to find them; I am looking for thin, flint glass coverslips, in order to attempt compensation for the effect described here. Some company near to Baltimore used to make them , around the year 2000 but I have been unable to find them.
Not all Met objectives have this degree of problem though. I bought a Bausch & Lomb Balplan 80x .90 flat field fluorite Met objective a number of years ago for almost nothing. It has a 25mm thread , so it was looking for a home and it was a few years before I addressed the issue of putting an adapter on it for R.M.S. When I did and tried it out on a biological Balplan( hoping it might be usefull for dry high magnification dark field),with a coverslip,I was surprised that the image it provided was almost as good as the Balplan oil immersion planachro, which are excellent almost completely chroma free planachromats. It is certainly as good and better, than any other 60-80x dry objective I have ever used, which usually do not perform as well as similar magnification immersion objectives.

[Pau]

Chris,
If this specific kind of subjet (glass covered sensor) is really important for you you could find adequate the use of an objective for biological inverted microscope: they are designed to see through a thick glass (microscope slide or botom of a Petri dish) and someones have a compensating collar (for exemple for 1.2 to 0.6mm glass)
Phil,
I find your experience with the use of a high NA M objective with coverglass really extrange... maybe outside of the optics general knowledge, could you explain it further or provide images? 0.90 is very high NA for a dry objective and really very sensitive to coverglass thickness!

[phil m]

I did too, Pau. AO had a no cover 100x .90 planachro cat.#1867 made back in the 80's, by Reichert and I tried it under similar conditions to the B&L 80x .90 and it's image was about the same as using a 100x oil immersion planachro without oil.
I will post some pictures -----comparative they will have to be because I am still limited to afocal photography and a fairly basic camera..

[Pau]

... and it's image was about the same as using a 100x oil immersion planachro without oil.

The image given by an oil corrected objective used dry is very poor due to strong spherical aberration (someones even are unable to deliver any image at all)

[phil m]

There should therefore be increasing loss of resolution the more off axis one looks. Also rays scattering beyond the objective's angle of acceptance , reducing the size of the cone of light being captured, cause a drop in N.A. below 1 . The effect is similar to closing the iris down on an iris equipped immersed objective in bright field.

[phil m]

Phil,
I find your experience with the use of a high NA M objective with coverglass really extrange... maybe outside of the optics general knowledge, could you explain it further or provide images? 0.90 is very high NA for a dry objective and really very sensitive to coverglass thickness![/quote]

Here are some unaltered pictures taken through a Balplan with 10x W.F. oculars. Afocal, using a Kodak C195, point and shoot.
The rectangular frame is just about as close to the edge of the field as you can get and not have vignetting.
The subject is an unstained slide I use as a resolution and chromatic aberration tester. It is a difficult slide to resolve andas well ,pushes the limits of objectives as far as their correction for chromatism goes.

1) 100x 1.25 Flat Field achromat oil--- without oil
2) 100x 1.25 planachromat oil--- without oil. You can clearly see the loss of N.A.,when compared to 3 & 4, without the immersion, moreso with the planachromat, than the Flat Field. As well , in the next two pictures, there was a remarkable improvement in B&L's achromats over about the 10 year period that separated the mfg. of the two Balplan objectives.
3) 100x 1.25 Flat Field achromat oiled
4) 100x 1.25 Planachromat oiled
5) 100x 1.32 Planapo oiled( This is an American Optical/Reichert on a different stand, for comparison.
6) 80x .90 Flat Field Fluorite Met 211/0
Normally , the image in picture 6 , should have been degraded, in a similar fashion to those in 1 & 2, due to the fact that the objective is a no cover objective and is corrected for 211mm as well but the image is very good, as well as having a very low level of chroma.
The only other objective I have that is of a similar specification is a Hacker ( Heurtel & Reuss) 80x .90 achromat. It is coverslip corrected. This is an N.O.S. objective and I believe is corrected for 160mm, so it is not usable on a Balplan stand. I have tried it on both B&L and Spencer 160mm stands but the image is slightly less good than the B&L Flat Field fluorite. I don't know whether that is because it lacks resolution due to design or because the telon lens systems of those two stands are not correcting properly but the chroma does not stand out and the image is quite flat, so it seems the correction is o.k. It only goes to show that the B&L Met objective is little affected by the coverslip.

[rjlittlefield]

6) 80x .90 Flat Field Fluorite Met 211/0
Normally , the image in picture 6 , should have been degraded, in a similar fashion to those in 1 & 2, due to the fact that the objective is a no cover objective and is corrected for 211mm as well

I'm unclear on exactly what you've done, but if you've both omitted the cover slip and significantly shortened the tube length, then be aware that these two perturbations will compensate for each other to some extent. See this microscopy-uk article.

--Rik

[phil m]

The 3 100x biological objectives are simply there as a reference, to show the same specimen with lowered N.A. and with N.A. approaching the specification of the objective.
The loss of N.A. is due to oblique deviation of the rays in air, reducing the illumination cone, resulting in empty magnification.

The Met objective is the one under test. It is a no cover objective being used with a coverslip, not the other way around.Normally in such a scenario, the image is degraded in a similar fashion, to those from the unoiled , immersion objectives but due to somewhat different causes.
I am just responding to Pau's question, with pictures. The 80x Met objective , rather than having a degraded image due to the use of a coverslip, actually has a rather good , well corrected image...as good as any high magnification dry objective , I have ever used.
I do not know what the tube length of a Balplan is. What I do know is that, if you try 160mm objectives or infinity objectives on a Balplan, you can seldom get focus and when you do, it is usually a very poorly corrected image for chromatic aberration , sometimes fuzzy and often at a ridiculous working distance. I've never been able to get a Balplan objective to focus on another stand . I have been told, by someone who sold the microscopes and continues to service them, that there is negative lens in the system but I am unsure what that implies.

Curiously, I also have a Balplan Met 5x objective( from the same microscope as the 80x) , which I have tried in a standard biological Balplan stand, too. It yields a very distorted , curved image, with about 70% of the image increasingly riddled with chroma towards the perimeter, so it does not seem that tube length has anything to do with the performance of the 80x, otherwise the 5x should perform similarly well,because it too is a 211mm objective.
There clearly is a tube length reduction to some degree because based on the value of 100x as the baseline for the other objectives , and a measurement of the images, the image provided by the 80x , gives around 72x, which suggests a tube length of around 190mm.
I cannot explain why the 80x works so well,it just as I originally stated, does, and is a very useful tool because I hate immersion oil.

[rjlittlefield]

if you've both omitted the cover slip and significantly shortened the tube length, then be aware that these two perturbations will compensate for each other to some extent.

It is a no cover objective being used with a coverslip, not the other way around.

Sorry, I had the vision correct but completely scrambled my writing.

Let me ask a question that may give some clarity.

Background: You currently have a setup in which this 80x met objective gives an excellent image despite the presence of a coverslip that it's not expecting.

Question: in that same setup, if you use the objective without a coverslip, does it still give an excellent image?

--Rik

[phil m]

The image is about the same. If you scan away from the covered specimen to the balsam ridge at the edge of the coverslip or do an epithelium smear on another part of the slide or focus on the inking of a label with transmitted light,in other words not change anything in the settup but just choose another uncovered available specimen, the image i spretty much similar.
Perhaps , if it is o.k. with Chris Mower, I could do just that and send some pictures. I used some reference slides to begin with that would give a baseline because the auto focus on my current camera is a bit difficult to control and those images are actually sharper to the eye than the camera captured.

[Pau]

Phil, thank you for posting the test images. In fact your 80x .90 works clearly better than the oil objectives without oil (not really surprising) but still better than I could expect with covered specimens, and very well about lack of CA.
But your test slide has not fine detail to see actual resolution differences, I would recommend you diatom test slides for this purpose (none of the posted images have really fine detalils resolved).
About the camera, autofocus is really a bad thing in a microscope camera, can't you disable it?

[phil m]

yes, I can do some diatoms, or chromosomes.