Lighting and DOF ?s

[jmarkus]

Rik,
The lights are for a Tomlov scope, and are incredibly bright (6 watt approx 5000k). There is a rheostat to adjust intensity in about 3 steps for one or two lights. I set the WeMicro to 0.9um steps and a total of 222 steps. I had to chimp the first step to see the beginning focus point as the live view would not let me go below 1/30th of a second, and I was shooting 1/5th of a second. I ran it - then checked the last frames focus point, and it wasn't deep enough - so I ran it a total of four more times which fixed some of the field curvature focus and depth. I did the last one on my base mac m2 mini - pulling the files (11.4gb) from a Sandisk 4tb G40 external drive (very fast) and I could watch the preview move down then rotate towards the end like a movie file, because it compiled the stack so fast - just a minute or two.

In Photoshop I would done auto align, then auto blend. It would probably still be working on the whole stack. However, this kind of movement would have been corrected by increasing the canvas size and adding each image aligned before the blend.

I have some small 45 watt halogen lights that get hot as hell - that what you talking about?
Jim

[rjlittlefield]

I set the WeMicro to 0.9um steps

0.9um would be appropriate for your 40X NA 0.65 objective.

But for your 10X NA 0.25 objective, 5 um would be more appropriate.

Which objective were you using?

I have some small 45 watt halogen lights that get hot as hell - that what you talking about?

That will work. The key thing is that the light gets produced by filament that is hot enough to glow. An LED that gets very hot still won't work.

--Rik

[jmarkus]

Rik,
I was using the Leica Achro 10/.25 160/.17 objective with the 0.9um increments.

I tried reshooting with the 665nm converted 5DmkII using the 5um increments, and it was even less sharp.

When I first started shoot what I though was IR - I got frustrated by the focus never being close to the lens barrel engravings. The Canon 5D does not have live view so I had to chimp shots to find the proper focus on anything less than infinity, and even infinity was never the infinity marking on the barrel. I use to find it for a particular lens and put a strip of monster tape on it so it wouldn't drift as I shot sequences for stitching larger resolution landscapes. I shot only with manual focus Nikkors.

With the 5DmkII I am dependent on the live view to see critical focus, and not being able to use that makes it feel like a blind approximation to getting a proper starting point. So, I think it came out worse, because of my error.

Though I know pinhole cameras can cover a large area - I have been surprised that such a small microscope objectives can cover not only a crop censor, but even my FX sensors. Since IR isn't currently working out - I went back to visible light with a third camera body - the Nikon D800. Same subject (though 50% smaller lacking the crop factor) However, I imaged it differently. Shot raw > imaged in Lr > export to jpg > slight topaz sharpen ("standard" 1,2,1) > then stacked the 243 images three ways - shot at 5um increments - using Rik's Zerene Stacker.

I have a used Dolan Jenner fiber optic illuminator with a ring and goose necks coming, better objectives on a broken scope, and I am building a stage that I hope will give me a little of the angle I would get with just mounting an objective directly to the camera. Lighting seems to be the biggest challenge in micro photography.

[rjlittlefield]

This last image looks better than your earlier efforts. But it still looks like not enough diffusion yet.

I suggest trying to add "too much" diffusion. For this sort of subject it's very hard to actually get too much, and in the unlikely event that you do, it's always easy to back off.

When I first started shoot what I though was IR - I got frustrated by the focus never being close to the lens barrel engravings. The Canon 5D does not have live view so I had to chimp shots to find the proper focus on anything less than infinity, and even infinity was never the infinity marking on the barrel. I use to find it for a particular lens and put a strip of monster tape on it so it wouldn't drift as I shot sequences for stitching larger resolution landscapes.

For landscapes, a filter labeled "665 nm" probably did get you mostly IR. What forms the image is the product of source emission, filter transmission, and camera sensitivity. Those three numbers get multiplied together at each wavelength separately, then the products summed over all wavelengths. For your studio work, the LED is not good for IR photography because it emits almost no IR in the first place. But the sun emits lots of IR -- it's basically just a very hot black body with minor alterations caused by absorption in the atmospheres. In that case, most of what gets through the 665 nm filter really will be IR.

I have been surprised that such a small microscope objectives can cover not only a crop censor, but even my FX sensors.

The size of hole in a lens seldom has much effect on coverage. What limits coverage is (1) loss of image quality away from center due to aberrations, and (2) vignetting caused by two holes in the lens and/or couplers that conflict with each other to block off-axis rays.

Lighting seems to be the biggest challenge in micro photography.

For stable subjects, I agree completely.

--Rik

[jmarkus]

The "broken" Leica CME arrived today, and I got everything working, fashioned a camera adapter, temporary lighting, and took some iso 2500 images with a fourth camera - the Nikon D850. Everything about this instrument is better than the BF200, and finding information other than a overly simple 6 page manual, marketing brochures, or instructions on the condenser seems impossible to find. I think the BF200 is from 1996, and this CME from about 2000, but there appears to be two generations of improvements. I suspect the CME was replaced by the DME in very short order, but I have no clue as to why. I would have loved to have found a service manual with exploded views and diagrams. Never-the-less, I think it is a great microscope. I found a cool diaphanous paper to diffuse the light and tone down the highlights - "Nu-Pore" (it's for getting grease off your nose and face).

Subject = Pentium III core tilted by laying it on the cover glass of a slide. Once I get the goose neck fiber
optic lights ISO can drop.

[jmarkus]

Tried 100x on the edge of a milkweed seed. Furnace kicked on and half of them blew off my desk so I trapped one between slides.

[jmarkus]

Still waiting on the fiber lights, but got to use the CME some more. The Leica E2 Plan objectives are noticeably better than the pre E Leica Achro objectives I have on the BF200. The focusing is a coaxial arrangement with an almost resistant free fine adjustment. It doesn't change position - so it works better with the WeMicro. The top vertical tube has a focusing helicoid that is a little too loose to hold the camera as still as I would like - A set screw would have been a nice addition. The stage is amazing, and I don't understand the light & condenser as well as I should - looks like there is room for filters to be added. The "broken" part wasn't broken at all - the light wouldn't come on, and I assumed the rheostat had a switch built in - it didn't. I looked multiple times for a separate switch, and could not find it. Checked fuses and bulb, and they looked fine. One handy feature was a place to coil the cord on the back of the scope. It looked like it had never been uncoiled. In fact, the scope looked it had never been used. I uncoiled the power cord, and viola - there was the switch buried deeply next to and under the bracket.

Subject = the transition on a Money Plant seed pod that turns into a double pointed spire. Leica E2 Plan 10X about 125 .5um increments.

[jmarkus]

Been reading about DIC, and I'll admit to thinking about putting a subject between two polarizing filters I have. But in DIC there are 45 degree Wollaston prisms that I am not understanding their purpose. The light already gets directional selection by the first polarizing filter - wouldn't the prism just remove more light getting through, because it's a straight path from illuminator to eye piece. Additionally, the second prism cuts more light before you can rotate the second polarizer to restrict the light for the angle of light you want. It just seems superfluous to even have the Wollaston prisms in the mix?

[jmarkus]

Couple more with the new to me CME

[rjlittlefield]

Been reading about DIC, and I'll admit to thinking about putting a subject between two polarizing filters I have. But in DIC there are 45 degree Wollaston prisms that I am not understanding their purpose. The light already gets directional selection by the first polarizing filter - wouldn't the prism just remove more light getting through, because it's a straight path from illuminator to eye piece. Additionally, the second prism cuts more light before you can rotate the second polarizer to restrict the light for the angle of light you want. It just seems superfluous to even have the Wollaston prisms in the mix?

You need to read a lot more about DIC. I suggest the explanation at https://en.wikipedia.org/wiki/Different ... microscopy .

--Rik

[jmarkus]

Rik, That is the first place I went. It was this paragraph that threw me:

Adding an adjustable offset phase determining the interference at zero optical path difference in the sample, the contrast is proportional to the path length gradient along the shear direction, giving the appearance of a three-dimensional physical relief corresponding to the variation of optical density of the sample, emphasising lines and edges though not providing a topographically accurate image.

What that says to me is the prisms' function is to split the polarized beam into two identical, yet offset (adjustable) beams. That through that adjustment - you can create the illusion of depth. Hence the "not accurate" statement. Something that looks very similar and is derived mathematically occurs in Photoshops' "Emboss" filter. So my question is: Is DIC the analog version of Ps Emboss?
Jim

EDIT to add example of the last photo I uploaded with an emboss filter applied

[rjlittlefield]

Is DIC the analog version of Ps Emboss?

No. Emboss takes visible patterns and changes them into other visible patterns. DIC takes invisible patterns of phase differences, and turns them into visible patterns of intensity.

In the extreme case, you could have a subject which appeared almost uniform gray in simple brightfield, but in DIC would be seen to have vivid patterns. Physically, those patterns would be gradients in refractive index, say on the boundaries between pure and slightly saline areas of water. In practice the patterns are often somewhat visible in brightfield also, but are way more obvious and easier to capture in DIC.

As illustration, see https://zeiss-campus.magnet.fsu.edu/art ... iques.html and look at Figure 2:


I am annoyed that the above panel of images uses different subjects for each imaging technique. I would much prefer that it was the very same slide imaged each way, but in quick search I could not find an image like that.

Something close is at https://imb.uq.edu.au/research/faciliti ... techniques :


Careful use of oblique brightfield can produce much the same effect as DIC, but the oblique brightfield is far more difficult to set up and get good images with. See viewtopic.php?t=15142 for one study.

--Rik

[jmarkus]

Thank you Rik,

Those links took a while to digest, and then think about what my initial focus has always been when I began this journey. From my first post:

Is there a way to increase DOF, yet maintain high magnification? Do people shoot through the equivalent of a smaller aperture to create greater DOF? The next disappointment, for me, is nothing looks like real life because of back lighting. Colors and dimension (depth) are lost, and getting the light on top of the subject is really a challenge.

All the lighting methods in those links still do not look real life except some polarized light samples. Not only that, but they aren't even accurate representations of real life. Creating structures that don't exist out of phase variances with highly manipulated light isn't what I want to pursue. It is valuable for clinical studies, and finding shapes that elude the eye and traditional imaging, but I'm just a photographer that likes to record light reflections.

I have been having my best results -IMO- with combining darkfield and lightfield lighting. Maybe polarized light could enhance that further towards my "looks like real life" goal?

I really appreciate you and the others that responded taking all the time and effort to explain the lay of the land to this neophyte. I have my expectations under control, and the direction narrowed which I want to pursue in microscope photography because of your condensing it in understandable terms.
Jim

[rjlittlefield]

Maybe polarized light could enhance that further towards my "looks like real life" goal?

Maybe, but I'm not optimistic. In ordinary photography, polarizers are mostly used for reflection control, for increasing color saturation (as a result of controlling surface reflections), and for darkening blue skies. In microscopy, polarizers are usually used to split and recombine light beams that have different phases, so as to take advantage of interference effects. DIC is one example of that; another example is using cross-polarization to reveal birefringence of strongly ordered molecules like crystals, stressed plastics, and muscle fibers. Polarizers can be used for reflection control in microscopy, similar to ordinary photography, but it's more difficult because of the wider ranges of angles that are in play.

The best examples that I know of "looks like real life" microscopy all come from using heavily diffused illumination, with just enough directionality to provide 3D modeling. One technique I think we have not discussed is to remove the condenser of your microscope, and shine light up from a diffused source that is hidden from direct view of the objective. This method makes it relatively easy to get high mag shots that have much the feel of oblique backlighting of larger subjects. See for example the third image at viewtopic.php?t=1076 . But most examples are with heavily diffused front illumination as we have discussed earlier. That technique, implemented with fiber optic bundles shining through a pingpong ball, was what Charles Krebs used for his groundbreaking fly head that won the Nikon Small World competition in 2005. Now 20 years later, that particular shot would be completely routine, but conceptually similar methods are being used every day at all magnifications as high as a microscope will go.

--Rik

[jmarkus]

That's the ticket! What Charles Krebs did with the fly head is close. I would have had a light ratio like right side stay the same, but a shadow gradient going over the left side of the fly's face to add drama. I just purchased my intended optical fiber light modifiers which I will have to modify on my drill press. Have to wait on a custom milled bushing for the goose-neck lights anyway, and the fiber optic light source needs a new plug, and rivet. Ping pong balls - I love it. Thanks Rik.