Epi DIC woes

Here you can discuss different microscopic techniques and illumination methods, such as Brightfield, Darkfield, Phase Contrast, DIC, Oblique illumination, etc.
Message
Author
microb
Posts: 729
Joined: Wed Mar 27, 2019 6:39 am

Re: Epi DIC woes

#61 Post by microb » Sun Nov 29, 2020 7:08 pm

So I’m wondering if moving the UMA up and down improves the BHM2 reflectance DIC images.

The UMA’s BF cube has a lens (marked in green) that I assume does focus to the back focal plane of the objectives. If the stand, which I used was a Silicon Valley Scientific BH2 stand, is not the right thickness offering the right distance from the UMA to turret, or maybe even the turret itself is for a frame’s metal casting, the end result might be blur (?).
Attachments
BF_lens_marked.jpg
BF_lens_marked.jpg (108.27 KiB) Viewed 14721 times

viktor j nilsson
Posts: 760
Joined: Thu Aug 23, 2018 10:12 pm
Location: Lund, Sweden

Re: Epi DIC woes

#62 Post by viktor j nilsson » Thu Dec 03, 2020 10:07 pm

I've kept thinking about this, and have started to wonder if this type of image doubling isn't a problem that is inherent to finite epi DIC systems. As you know, in a finite system, it's really important to only place optical accessories such as beam splitters and dichroic mirrors within an infinity space, as they would otherwise produce severe aberrations and ghost images:

Image

That's why these components are always placed in an intermediate tube between two telan lenses.

But in finite EPI DIC, we are basically putting a prism with two tilted wedges in a converging light path (just behind the objective). And in some cases, like the Olympus system, this prism is even inserted at an angle. It seems to me like this is exactly the situation which we expect will result in ghost images?

I guess that whether the doubling is perceived as noticable or not would very depending on the shear angle, tilt prism and magnification.

Scarodactyl
Posts: 2775
Joined: Sat Mar 03, 2018 9:09 pm

Re: Epi DIC woes

#63 Post by Scarodactyl » Fri Dec 04, 2020 3:26 am

The bh2 epi DIC is an infinity system. In contrast Nathan/Microworldofgems (@microworldofgems on instagram) currently uses a finite zeiss epi dic system which does not show doubling from what he can tell, certainly none visible in his pictures. I do think it's probably inherent to the system though--it was probably good enough for the subjects it was meant for, since it is only objectionable with sharply angled features and little specks. On a wafer it looks really nice.

viktor j nilsson
Posts: 760
Joined: Thu Aug 23, 2018 10:12 pm
Location: Lund, Sweden

Re: Epi DIC woes

#64 Post by viktor j nilsson » Fri Dec 04, 2020 10:33 am

Scarodactyl wrote:
Fri Dec 04, 2020 3:26 am
The bh2 epi DIC is an infinity system. In contrast Nathan/Microworldofgems (@microworldofgems on instagram) currently uses a finite zeiss epi dic system which does not show doubling from what he can tell, certainly none visible in his pictures. I do think it's probably inherent to the system though--it was probably good enough for the subjects it was meant for, since it is only objectionable with sharply angled features and little specks. On a wafer it looks really nice.
Right. I forgot about that. Sorry. I continue to find this very strange.

Scarodactyl
Posts: 2775
Joined: Sat Mar 03, 2018 9:09 pm

Re: Epi DIC woes

#65 Post by Scarodactyl » Fri Dec 04, 2020 9:48 pm

Yup, it's all very odd. Every time I think I have a good idea it just doesn't pan out. I've almost got my second nikon dic system set up (later version of the optiphot system) so i'll have a comparison test in the next couple weeks if all goes well.

Scarodactyl
Posts: 2775
Joined: Sat Mar 03, 2018 9:09 pm

Re: Epi DIC woes

#66 Post by Scarodactyl » Sun Apr 25, 2021 3:44 am

Well, my DIC woes are finally (w)over!
The second Nikon DIC setup I mentioned in the last post came with seriously gummed up prism sliding mechanisms. I don't know how they came up with the design for the sliders--it looks like they copied the general form of the Olympus BH2 prisms with their lever action but somehow overcomplicated a mechanism that boils down to "thing that moves in or out". I note that for their next system they just used a slot and slider above each objective. The results (when I managed to slide a prism in) were basically identical to the other Nikon prisms I'd tried--only partial compatibility with the mitutoyos and enough doubling to bother me. I later had my local microscope shop (Microscope Solutions) have a look, and they managed to get the prisms moving smoothly so I sold it on.
Then I spotted a unicorn on eBay--a Nikon L-DIC slider, the one used in their newer Eclipse LV industrial microscopes, at a very reasonable price. In a departure from the optiphot era this is a single prism slider for all their objectives, with one knob to move the prism in and out and another to move it up and down, very fancy and potentially good for tuning compatibility.

Actually getting it put onto my microscope was a bit of a challenge--skip these two paragraphs if you're not into tedious mechanical adapting stuff. Basically, to use this DIC slider I needed a nosepiece with a slot to accept it. Those nosepieces are all part of the Eclispe series, but my stand is a UM-2 measurescope which is in the Optiphot system. Fortunately I found an Eclipse series head/nosepiece holder mislabeled on eBay, and I was able to design an adapter and have it machined so I could mount the whole thing on the UM-2. Then I had to 3D print a sleeve to get my Optiphot illuminator to fit in the larger Eclipse dovetail (infinity optiphot illuminators have a labophot/optiphot sized male dovetail on the bottom and a larger eclipse sized female dovetail on top).
The relatively inexpensive NBD5 nosepiece has an external covered slot which looks like it should accept a slider, so I ordered one along with some m32 adapters from Thorlabs. When I got it in I found out that that slot is a complete lie--behind the cover is the side of the cylindrical darkfield stop, which extends through the whole length of the nosepiece. I spent a while waiting to see if a DIC-compatible nosepiece would show up cheap on eBay but it didn't. I got to thinking that they probably used some of the same parts between them, or the cheaper nosepiece wouldn't have the fake prism slot, so I got a second one even cheaper and removed the darkfield stop. This reveals space for a slider, except the space tapers from 23.8mm at hte top down to 23.2mm at the bottom, while the slider is exactly 24mm wide. I assume that they cast them like this and then do more machining for the DIC-compatible nosepieces. Fortunately I had files, a dremel and some time and was able to widen the slot out enough to get the slider to fit (though it isn't quite as smooth as I'd like, so I'll probably need to take just a bit more off).

Long story short, though, I got it to fit!
Image
After all that I was pretty concerned it wouldn't work well with my mitutoyo objectives, but fortunately it works great with them, even the 2x! Here are a few pictures of surface features on a Herkimer "diamond" quartz crystal:
Image
Image
Image
Image
The colors are really nice, just as good as the Olympus BH2 prisms, but the doubling is much better controlled--it is still present to some extent (I think it's kind of unavoidable) and visible on tiny specs when zoomed all the way in, but isn't enough to cause problems with photos. I'm finally getting just what I want from this!
So far the only adjustment that seems to make a big difference in compatibility is the vertical one--on the 20x objective the coloration is more even when it's set to the "a" position.
Anyway, I am super pleased with this and looking forward to finally spending some quality time with various crystals' surface geometry.
Last edited by Scarodactyl on Sun Apr 25, 2021 4:45 am, edited 1 time in total.

PeteM
Posts: 2982
Joined: Tue Dec 06, 2016 6:22 am
Location: N. California

Re: Epi DIC woes

#67 Post by PeteM » Sun Apr 25, 2021 3:53 am

The images look great. Persistence rewarded!

User avatar
75RR
Posts: 8207
Joined: Sun Oct 12, 2014 2:34 am
Location: Estepona, Spain

Re: Epi DIC woes

#68 Post by 75RR » Sun Apr 25, 2021 8:09 am

Congratulations.

Read the 'tedious mechanical adapting stuff' must say, you certainly earned those images!
Zeiss Standard WL (somewhat fashion challenged) & Wild M8
Olympus E-P2 (Micro Four Thirds Camera)

Antartica
Posts: 226
Joined: Wed Mar 04, 2020 4:52 am

Re: Epi DIC woes

#69 Post by Antartica » Thu Jun 24, 2021 3:14 am

Curious if there have been any updates since last time? I would also like to explore epidic and gems in the future so I was bummed to find out Olympus bh2 has issues with epidic.

Have you had any new breakthroughs? I guess I just find it hard to believe that the Olympus is incapable of decent images given that the Zeiss Universal that Nathan uses is a few decades older. Surely it must be possible to obtain great sharp images. Is there something else we're missing?

Scarodactyl
Posts: 2775
Joined: Sat Mar 03, 2018 9:09 pm

Re: Epi DIC woes

#70 Post by Scarodactyl » Thu Jun 24, 2021 3:15 pm

I'd say my final conclusion is some degree of doubling is unavoidable because of the nature of DIC. For some subjects this doesn't matter, but if there are very sharp angles or tiny specks it's more likely to be objectionable. The Zeiss system, much like the newer Nikon one I now have set up, apparently has a smaller offset between the images. I do have one of Nathan's prints on my wall, and if you get right up to it you can spy a hint of doubling same as with the new Nikon system, it's just fairly minor.
I wouldn't say the Olympus system or the finite nikon one is unusable by any means, some of this was me overreacting to what I thought was a defect rather than an inherent drawback. It would be pretty easy to stitch some images together and make the doubling much less significant while still getting exceptional color.

Antartica
Posts: 226
Joined: Wed Mar 04, 2020 4:52 am

Re: Epi DIC woes

#71 Post by Antartica » Thu Jun 24, 2021 7:43 pm

Thanks for the answer! I don’t know anything about dic or Epi dic . I just use my Olympus bh2 for bright field and simple organisms.

Is it not possible to use dic for metallurgy? Im a bit confused as to what epi dic is for?

PeteM
Posts: 2982
Joined: Tue Dec 06, 2016 6:22 am
Location: N. California

Re: Epi DIC woes

#72 Post by PeteM » Thu Jun 24, 2021 10:45 pm

There's sort of a trade off in the choice of shear angles for DIC; between contrast and resolution. On the transmitted side, Olylmpus offered three choices for each magnification to better match the subject. Seems the reflected side, as Stephen says, was optimized for wafer inspection, tiny surface defects, etc. So, yes, epi DIC is good for viewing things like grain structure in a polished metallurgical specimen, deposition layers on a silicon wafer, etc.

Scarodactyl
Posts: 2775
Joined: Sat Mar 03, 2018 9:09 pm

Re: Epi DIC woes

#73 Post by Scarodactyl » Sat Jul 24, 2021 6:20 am

Just a general note. I recently bought a Zeiss universal with epi DIC to restore/try out/sell on. This is not new info, but I can confirm the Zeiss system felt much better to me than the contemporary Olympus, Nikon and Bausch and Lomb options I have tried--it is at least as good as my modern Nikon system. This is the one you want to seek out among those options. There are disadvantages: short working distance, limited nosepiece options, and the inability to remove the prisms from the light path for normal epi (in addition to the other issues with old zeiss scopes) But the optical quality is great, there's almost no visible doubling and it has great color and resolution. I am waiting on a replacement stage and a better camera hookup, but pictures will be incoming.
Last edited by Scarodactyl on Sat Aug 28, 2021 4:08 am, edited 1 time in total.

eKretz2
Posts: 62
Joined: Sat Jul 31, 2021 9:07 am

Re: Epi DIC woes

#74 Post by eKretz2 » Sat Aug 21, 2021 2:29 am

I recently picked up a Vanox AHMT (this uses a similar f=180/infinity setup and same Neo SPlan objectives as the BH2) and will hopefully soon have some prisms to try. It will be interesting to see if it has the same issues with doubling. I wonder if the mystery markings on the flip-out prisms (T-2, T-4, etc.) are similar to the newer versions' HC (high contrast) or HR (high resolution)? Maybe one of those is more prone to the doubling.

Also I've read that any birefringent materials are not good candidates for DIC. Some diamonds (apparently more than do not - what was the source of your particular diamond?) do display birefringence. Did you try imaging any solid opaque objects?

Scarodactyl
Posts: 2775
Joined: Sat Mar 03, 2018 9:09 pm

Re: Epi DIC woes

#75 Post by Scarodactyl » Sat Aug 21, 2021 5:07 am

Birefringence has no effect on epi DIC, at least not for surface imaging.
It's quite possible they had an hc vs hr split or something like that. There is an inherent tradeoff between resolution and contrast because of the splitting and recombination of the image (my current Nikon system has an HC equivalent so the default ine must lean more to resolution which would line up). I have not found any information to that effect about the Olympus splan system but there are some big knowledge gaps here. Even if not, the doubling could just indicate they were aiming for higher contrast by default. The color produced by this system is excellent.
That all said, if it ends up being an issue you can always use a higher power objective and make a panorama. Doubling becomes less significant but the excellent rendering remains.

eKretz2
Posts: 62
Joined: Sat Jul 31, 2021 9:07 am

Re: Epi DIC woes

#76 Post by eKretz2 » Sat Aug 21, 2021 5:36 am

How about stage rotation? Did you try that too? One of the pages I read somewhere mentioned that major artifacting can be introduced at certain angles to the prism. There were some sample images that showed a specimen in sharp detail and then others with the stage slightly rotated in which the specimen was pretty much unrecognizable.

Scarodactyl
Posts: 2775
Joined: Sat Mar 03, 2018 9:09 pm

Re: Epi DIC woes

#77 Post by Scarodactyl » Sat Aug 21, 2021 5:58 am

The shear is in one particular direction, so rotating the subject could certainly help minimize the appearance of artifacts in some cases. It won't affect the degree of doubling though.

eKretz2
Posts: 62
Joined: Sat Jul 31, 2021 9:07 am

Re: Epi DIC woes

#78 Post by eKretz2 » Tue Aug 24, 2021 4:09 am

Got some prisms on loan, first crack with them tonight. Tried a razor edge for these images. Don't have a good camera set up yet, these were just through the eyepiece. The bottom looks like it may have some of the doubling, but that was taken with the prism shifted off center to get some color. With the prisms set in their locked, centered position everything looked pretty clean.
20210823_230501.jpg
20210823_230501.jpg (113.59 KiB) Viewed 12887 times
20210823_230516.jpg
20210823_230516.jpg (158.83 KiB) Viewed 12887 times
20210823_230819.jpg
20210823_230819.jpg (173.17 KiB) Viewed 12887 times

PeteM
Posts: 2982
Joined: Tue Dec 06, 2016 6:22 am
Location: N. California

Re: Epi DIC woes

#79 Post by PeteM » Tue Aug 24, 2021 4:16 am

Uh, oh. Now you're hooked . . .

The 500x image is good. Add focus stacking and it would be spectacular.

eKretz2
Posts: 62
Joined: Sat Jul 31, 2021 9:07 am

Re: Epi DIC woes

#80 Post by eKretz2 » Tue Aug 24, 2021 6:18 am

Yeah that is no joke. DIC is pretty darn cool. I am absolutely planning on getting a better imaging setup so I can do some stacking. I noticed that one of the prisms is just a smidge off from where I'd like it to be for the best contrast, do any of you guys know of any documentation as far as setup and adjustment of the prisms? I noticed there's a stop screw for setting where the prisms rest when they flip in, but not sure if you are just supposed to adjust it to where you like or? Is there a centering procedure maybe?

Scarodactyl
Posts: 2775
Joined: Sat Mar 03, 2018 9:09 pm

Re: Epi DIC woes

#81 Post by Scarodactyl » Tue Aug 24, 2021 6:29 am

Doubling if oresent will only really be evident on small specks or sharp acute angles. Things like scratches on textured backgrounds, features with lots of right angles (so basically wafers), smoothly varying features etc will usially not make it very evident.
I do like the look of your pictures there!

PeteM
Posts: 2982
Joined: Tue Dec 06, 2016 6:22 am
Location: N. California

Re: Epi DIC woes

#82 Post by PeteM » Tue Aug 24, 2021 6:44 am

Eric, some of the microscopy sites (Olympus, Nikon, etc.) have guides to DIC alignment.

A useful tool is a "phase telescope" aka "centering telescope" and sometimes incorporated in a scope as a "Bertrand lens." Might find a nice used phase/centering telescope on Ebay for $50 or so and cheap new ones aren't much more.

Pop it in instead of one of the eyepieces and it can focus to various planes in the optical train. You can more readily see that you have perfectly crossed polarizers - and then see that the interference bands are properly centered. It's also useful for seeing where some annoying bit of dust etc. might reside - and of course for making sure phase contrast rings and annuli properly overlap.

eKretz2
Posts: 62
Joined: Sat Jul 31, 2021 9:07 am

Re: Epi DIC woes

#83 Post by eKretz2 » Tue Aug 24, 2021 7:18 am

Thanks Pete, I will dig around a little. I had some contact with an Olympus rep, maybe I will try rattling his cage too. It would be nice to find out once and for all what the "T-2" "T-4" etc. mean in reality.

Scarodactyl were you using a tint plate in your images? I mostly don't get any color at all really unless I slightly move the prisms out of their resting position. But as you pointed out I have only imaged a few fairly flat items so far.

Antartica
Posts: 226
Joined: Wed Mar 04, 2020 4:52 am

Re: Epi DIC woes

#84 Post by Antartica » Wed Aug 25, 2021 12:05 am

So the Olympus bhm is capable of sharp Epi dic :D that’s good to know. I’m slowly piecing mine together.

Post more images! I’d love to see more with different angles like scaro mentioned to see how it performs.

eKretz2
Posts: 62
Joined: Sat Jul 31, 2021 9:07 am

Re: Epi DIC woes

#85 Post by eKretz2 » Wed Aug 25, 2021 6:28 am

So this one was pretty cool. Edge damage on a razor blade. The width of the "crater" is about 230μ. This resized version isn't as sharp as the larger file, that looks a lot better.
20210825_012744.jpg
20210825_012744.jpg (109.8 KiB) Viewed 12767 times

eKretz2
Posts: 62
Joined: Sat Jul 31, 2021 9:07 am

Re: Epi DIC woes

#86 Post by eKretz2 » Wed Aug 25, 2021 2:21 pm

I'm not sure what subjects would make good "sharp angle" candidates. I don't really have any diamonds floating around. Scaro, what magnification were you using? And did you check the fringe patterns of your prisms to be sure they were centered? I've found that it is actually pretty easy to see if they're centered by just placing a tiny dot of liquid on a mirror and then focusing on it until you see the interference band. If it's not in the center of the sphere of liquid it's not centered. I noticed my 20x prism doesn't look centered, going to try to adjust it. The adjustment screws are pretty solidly locked with what I assume is threadlocker, so I'm going to disassemble to keep the prism safe and try some heat. Last night I took the prism housings apart and removed the spring mechanism that forces the prism into the locked position so that I can slightly translate the prisms to get some colors until I get a tint/waveplate set up.

All of my DIC with the setup I've got seems to be grayscale. If I translate the prism just a bit I can get a range of colors, which is cool, but I see a little of that double image/blur sometimes with that, so the tint plate is the better option I think. Here's the difference a very slight translation of the prism can make. Oh, and focus was at the top of the central blob, so the rest is a little out of the depth of field.
20210825_092738.jpg
20210825_092738.jpg (194.92 KiB) Viewed 12452 times
20210825_092728.jpg
20210825_092728.jpg (124.98 KiB) Viewed 12452 times

Scarodactyl
Posts: 2775
Joined: Sat Mar 03, 2018 9:09 pm

Re: Epi DIC woes

#87 Post by Scarodactyl » Wed Aug 25, 2021 3:21 pm

I had 10x, 20x and 50x objectives on it. I've since sold the wcope so I can't really do more with it translating the prism doesn't affect the degree of doubling (you can overlay theze two images and see they'll line up) but can adjust how apparent it is.

eKretz2
Posts: 62
Joined: Sat Jul 31, 2021 9:07 am

Re: Epi DIC woes

#88 Post by eKretz2 » Wed Aug 25, 2021 4:09 pm

Hmm, it does on my setup, but only after they've moved a bit. I can watch things move and shift while the prism is translated. Translating the prism on my setup shifts which side of the transitions are light or dark, so the images don't exactly line up. As that fringe leaves center it shifts things pretty considerably.

eKretz2
Posts: 62
Joined: Sat Jul 31, 2021 9:07 am

Re: Epi DIC woes

#89 Post by eKretz2 » Fri Aug 27, 2021 6:10 pm

So after some more playing around I think that the prisms with T-2 seem to be akin to the modern "HR," or high resolution, and the T-4 seem more akin to "HC," or high contrast. The T-2 prisms reflected off a mirror change the color of the entire field of view at once when the prism is translated, where the T-4 prism shows multiple color bands at once. In use I've noticed that gives a sharper view with more of a light/dark contrast in a single color band with the T-2 but a more highly color contrasting image with differing colors with small elevation differences on the T-4.

That's just observations from my own use though. I asked an Olympus rep about this and he just said that scopes from this era are antiques, everybody who knew anything about them has retired and they have no documentation for them. :?

eKretz2
Posts: 62
Joined: Sat Jul 31, 2021 9:07 am

Re: Epi DIC woes

#90 Post by eKretz2 » Fri Aug 27, 2021 7:30 pm

I happened to have a silicon chip lying around that I was looking at the other day and noticed Scaro's image from page 2 of the same subject. Here are the two next to each other at about the same scale. This is a T-2 prism at 50x for my image and a T-4 at 50x for his image.

Image

Image

I think you can see the little bit of color banding from the T-4 prism where the T-2 prism has a pretty uniform coloring across the field. I wonder if that makes the doubling a little worse on those maybe?

Also not sure why the surface texture on the lower image doesn't seem consistent. Scaro was that surface texture actually concave in the lower right section? Or did that change as you moved the chip like it was switching in that part of the field from convex to concave? I think that would indicate some sort of prism misalignment if it happened.

Post Reply