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

Epi DIC woes

#1 Post by Scarodactyl » Mon Apr 13, 2020 6:21 am

As I mentioned over in the my microscope section I recently acquired an Olympus BH2 with epi DIC capabilities. I have been enjoying it a lot, and the DIC effect is great, except for one persistent issue: I get a weak doubling effect on the image when the DIC prism is inserted. It's not bad for viewing but it really wrecks the quality of my photos.
Here's an example from my most recent shots of the surface of a diamond.
Image
Image
With the DIC prism removed using normal epi brightfield or darkfield I don't get any doubling:
Image
Image
I get this effect with all three prisms as far as I can tell. All of them seem to be in great condition. Does anyone know what might be causing this or how I might go about fixing it?

Microworldofgems
Posts: 54
Joined: Sat Jul 22, 2017 10:13 pm

Re: Epi DIC woes

#2 Post by Microworldofgems » Mon Apr 13, 2020 6:54 am

I’m also eager to see what others think is the problem here.

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

Re: Epi DIC woes

#3 Post by 75RR » Mon Apr 13, 2020 7:33 am

I imagine you have the instruction booklet, this is the link just in case you haven't:

http://earth2geologists.net/Microscopes ... ctions.pdf

Had a quick look through it but nothing caught my eye as yet. As the more eyes the better, perhaps someone else will notice something.
Zeiss Standard WL (somewhat fashion challenged) & Wild M8
Olympus E-P2 (Micro Four Thirds Camera)

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

Re: Epi DIC woes

#4 Post by Scarodactyl » Mon Apr 13, 2020 8:18 am

Thanks--I did have a look through it earlier, but I looked again now and the one thing I have not done is center the halogen bulb (I have a 100w halogen lamp which isn't mentioned in the manual [I did triple check to be sure it wasn't somehow a mercury lamp even though it's labeled hal]). I will try that tomorrow, perhaps it is the source of my issue.

abednego1995
Posts: 137
Joined: Thu Dec 15, 2016 6:32 am

Re: Epi DIC woes

#5 Post by abednego1995 » Mon Apr 13, 2020 3:01 pm

Are the prisms matched to the objective?
Shear amount should be tuned to be a tad under the objective resolution limit.

BR,
John

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

Re: Epi DIC woes

#6 Post by Scarodactyl » Mon Apr 13, 2020 8:54 pm

Yes, these are the exact prisms for these objectives. They also made separate prisms for thr ulwd objectives of this series but they are marked ulwd.
I centered the bulb but it had no effect. I am pretty sure I did it right but the filament was quite large in thr FoV with my phase telescope. Either way, just twiddling the rotational and horizontal position around had no effect.
I tested adding or removing components. With nothing but the DIC prism (no polarizer, waveplate or analyzer) doubling is very strong and obvipus. Putting in just the analyzer seems to make one of the double images a bit fainter. In full DIC with the waveplate in or out the doubling is totally unaffected by the rotational position of the polarizer, though by changing the colors it makes the effect more or less obvious to the eye.
I really hope this isn't just an inherent issue with olympus epi DIC of the era.

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

Re: Epi DIC woes

#7 Post by Scarodactyl » Mon Apr 13, 2020 9:38 pm

I spent some time looking for other examples of Olympus epi DIC images, and found this older thread viewtopic.php?t=4501
Some of his images show the same doubling, particularly the full image of the number 2. I am not sure if other subject don't show it, or if they're just less obvious about it.
The fact that I'm using a 2.5x projective on a crop frame camera is definitely not helping in making this more obvious. But it is a pretty frustrating issue. The prisms look immaculate anyway.

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

Re: Epi DIC woes

#8 Post by PeteM » Tue Apr 14, 2020 1:16 am

I wonder if fiddling with the field and aperture irises might get rid of some stray reflections?

Are the polarizer and analyzer OEM?

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

Re: Epi DIC woes

#9 Post by Scarodactyl » Tue Apr 14, 2020 1:33 am

Everything is OEM. It's in surprisingly immaculate condition, even though it was shipped in somewhat inadequate packaging--all that's wrong is a missing magnetic side cover which doesn't seem to affect the image at all (covering or uncovering does not have any visible effect). I should have mentioned these as well, but I did test the polarizers to see how good extinction was (seems to be in perfect order) and also fiddled with the irises, which also had no effect. The doubling only appears when the prism is pushed in and disappears when it is pushed out. The analyzer makes it slightly less obvious but does not remove the double image. All of my prisms do it.

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

Re: Epi DIC woes

#10 Post by Scarodactyl » Tue Apr 14, 2020 2:46 am

The plot thickens.
One thing I noticed and thought was odd in the manual was the mention that for DIC the polarizer should be inserted so that the 'NIC' label is facing forward, and for polarization it should be flipped so that the "PO" label is facing forward.
I did have it set with the NIC label facing forward, but it is really weird that this would make a difference. This should just be a linear polarizer from a naive perspective, so orientation should not make any difference. But it does--I tried flipping the polarizer and the results are pretty different.
Bear with me on these bad gifs.
With the polarizer in the 'NIC' position:
Image
This is with the prism fully in. Rotating the polarizer has no effect on the doubling, just changin the colors. It is constantly doubled with the same spacing. Every orientation is pretty colorful.
However, when I flip it to "PO":
Image
As I rotate it the two images fade in and out. The PO side has a mark that shows where the polarizer and analyzer are perpendicular (extinct)--the orientations with no doubling are 90 degrees away from that mark on either side. They are also the orientation with essentially no color. I can add some color by sliding the prism in and out, which does not change doubling, but only gives me pastel colors. Adding or removing the quarter wave plate does not seem to do much of anything in this position either (???)


I have no idea what's going on. The polarizer looks to be in absolutely perfect condition. It achieves basically 100% extinction when you take it and the analyzer out and cross them in front of a strong light. Why does the orientation affect anything? Why couldn't the position that works also be one that has nice pretty colors??

There's no question this isn't how it's supposed to be used. The instructions say to have NIC facing forward, and to change coloration by rotating the polarizer (it doesn't say anything about sliding the prism partially in, though it seems to be a normal part of other DIC systems).

edit: I don't know much about DIC, but I am guessing that from a really zoomed out perspective with the polarizer in the "PO" position I'm somehow isolating the two images that should be interfering, which is why there's no color when there's no doubling. What that would suggest is that the prisms are made so the two images that should interfere to make one image aren't aligned with one another. Thus maybe the prisms just weren't made quite right? Maybe this was considered good enough for industrial work by their clients? Or maybe something else it out of alignment, though it's hard to imagine what it could be.

microb
Posts: 729
Joined: Wed Mar 27, 2019 6:39 am

Re: Epi DIC woes

#11 Post by microb » Tue Apr 14, 2020 3:07 am

So the polarizer your have is a linear attached to a circular polarizer (1/4 wave plate). If circular is last, there is a range of EM vectors. When you reverse it, the linear wins and only one vector goes through.

Sorry if I misread and I'm answering something off in left field.

For 3D glasses in the theater, having two film layers gives the view more tilt range that the 3D effect works with. People in the biz call them dating glass, because the guy can lean in closer and still see the movie. Using just linear makes the image go dark when tilting the head.

(https://youtu.be/8YkfEft4p-w?t=1010)

(https://upload.wikimedia.org/wikipedia/ ... x.View.svg)

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

Re: Epi DIC woes

#12 Post by Scarodactyl » Tue Apr 14, 2020 3:20 am

I figured it must have a quarter wave plate attached (pretty much the only way to make this thing directional), but it's unclear to me why it'd be designed that way (the only quarter wave plate in a typical DIC diagram is the separate slider), or why exactly it's interacting with things this way. Of course it's extra confusing because quartz is chiral and also acts as a circular polarizer, and there are two differently-oriented crystal slices in the prism, so I'd probably have to read something a bit more involved than the microscopyU article to get a sense of what's going on.

microb
Posts: 729
Joined: Wed Mar 27, 2019 6:39 am

Re: Epi DIC woes

#13 Post by microb » Tue Apr 14, 2020 3:42 am

If you look through a linear and then circular filter, that order basically makes it a linear. So stick the two together and offer the customer options by flipping it.

So if you do it the other way and get more range of light, I guess I see your question. What does that get a person doing DIC? Looking at the images you gave, it does give more range of colors and as I mentioned gives it over a wider range of angles. I haven't seen any DIC diagrams put a quarter wave plate in there though. Wonder when that was documented as a thing to do.

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

Re: Epi DIC woes

#14 Post by Scarodactyl » Tue Apr 14, 2020 3:53 am

I just had another look, and the weird thing is I am pretty sure I'm getting a real DIC effect with the undoubled image with the polarizer reversed--it's definitely different from crossed polars in brightfield (which with the quarter wave plate basically give you all one color on specular reflections) and is clearly giving more topographic information. I might be able to shift the polarizer a bit away from the ideal position to get a bit more color too, or of course I can just push up saturation in post, but it's all very strange to me.

edit: I tried this out a bit more. It looks like I'll be able to get OK results with this if I tune it very carefully and then edit in post to approximate the more saturated colors of the proper configuration. I tried nudging various components (objective, prism, nosepiece, head) while looking through the eyepieces to see if a fine alignment issue could be the problem, but it appears not. I could gently flex things just a little bit but it did nothing to the doubling in the image.

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

Re: Epi DIC woes

#15 Post by 75RR » Tue Apr 14, 2020 5:18 am

I am pretty sure I'm getting a real DIC effect with the undoubled image with the polarizer reversed
Could this have something to do with the sample itself? Have you tried with a different material, perhaps something plainer might help distinguish what you are seeing.

https://gemgate.com/diamond-info/advanc ... efraction/
Zeiss Standard WL (somewhat fashion challenged) & Wild M8
Olympus E-P2 (Micro Four Thirds Camera)

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

Re: Epi DIC woes

#16 Post by Scarodactyl » Tue Apr 14, 2020 6:49 am

I'm using the diamond because the nice angular crystal faces make the doubling particularly obvious (it is also nice to be constantly reminded why I'm doing this, vs the less diluted frustration of looking at doubled dust specks on glass). Diamond really shouldn't have any properties that would affect anything, and I can unfortunately confirm doubling happens on every sample I tey it on.

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

Re: Epi DIC woes

#17 Post by 75RR » Tue Apr 14, 2020 7:51 am

This drawing would seem to indicate that the angle of the Normaski prism is critical in order to recombine the light waves.

The only other likely culprit would be the half-mirror.
Attachments
NORMASKI PRISM ANGLE.png
NORMASKI PRISM ANGLE.png (72.49 KiB) Viewed 16549 times
Zeiss Standard WL (somewhat fashion challenged) & Wild M8
Olympus E-P2 (Micro Four Thirds Camera)

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

Re: Epi DIC woes

#18 Post by Scarodactyl » Sun May 31, 2020 9:46 pm

Breakthrough!

Here's an image with no analyzer filter in at all (these are all greatly enlarged to exaggerate the effect). image doubling is extreme and both images are of about equal intensity.
Image

Here it is with the Olympus analyzer that came with it slid into position. One image is reduced but not eliminated.
Image

And here it is with the Olympus analyzer removed and my Nikon analyzer off my labophot laid inside the slot.
Image

I think I have found my problem! Though I am absolutely flummoxed as to why that would do it!

Neither analyzer is a strictly linear polarizer, but both give great extinction when oriented correctly. The only major difference seems to be that the Nikon is oriented at 45 degrees while the Olympus is at 90 degrees (or maybe 0 degrees I'm not sure how that's formally described).
But why would the Olympus one be oriented wrong for the prisms? Maybe there is more than one type of analyzer, but I didn't find any reference to that in the manual.
Either way I am super glad that this seems to be the issue, and not something with the prisms.

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

Re: Epi DIC woes

#19 Post by Scarodactyl » Mon Jun 01, 2020 2:30 am

Back to square one it seems. I tried it on a more complex subject and it looks like with the nikon polarizer I'm basically getting epi pol, not epi dic. So that's a bummer.
That said, if I angle the nikon polarizer a bit (it's a lot smaller than the olympus slot) it does start to get a bit of DIC coloration. I assume that if it were rotated to the same orientation as the olympus one it would show full color and full doubling. But I ordered a rotatable one to test further, so we'll see. At least the slightly tilted nikon polarizer is a much easier method than the previous one with the reversed polarizer at a particular orientation and it gets better color too, without any visible doubling. I can probably bump the colors up in post too.

User avatar
c-krebs
Posts: 201
Joined: Sat Jun 18, 2016 7:22 pm

Re: Epi DIC woes

#20 Post by c-krebs » Mon Jun 01, 2020 6:22 am

Question...
Do you have a base light. Can you do a "conventional" brightfield image? If so, do you get a clean image ("no doubling") when all DIC components and other filters are out of the light path?

I doubt this is your problem, but I once had a BH2 vertical illuminator that gave me faint double images. The culprit turned out to be the tube lens itself... in a unit that looked like new.

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

Re: Epi DIC woes

#21 Post by Scarodactyl » Mon Jun 01, 2020 6:54 am

It is a good thought, but unfortunately I have tested it. Doubling only occurs when the dic prisms are in. I can get clean epi brightfield and darkfield.

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

Re: Epi DIC woes

#22 Post by viktor j nilsson » Mon Jun 01, 2020 7:33 am

Really grasping for straws here, but... could someone have flipped the prisms upside down? My Nikon Epi DIC prism holders were sealed very hard with thread locker. It would be obvious if anyone had tampered with them.

microb
Posts: 729
Joined: Wed Mar 27, 2019 6:39 am

Re: Epi DIC woes

#23 Post by microb » Mon Jun 01, 2020 3:12 pm

Scarodactyl wrote:
Mon Jun 01, 2020 6:54 am
It is a good thought, but unfortunately I have tested it. Doubling only occurs when the dic prisms are in. I can get clean epi brightfield and darkfield.
You said you had matched objectives to prisms. But out curiosity: what objectives are you using? I'm assuming these are the flip in-and-out turret prisms.

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

Re: Epi DIC woes

#24 Post by Scarodactyl » Tue Jun 02, 2020 2:49 am

These are Neo Splans with the big (m26) lever-mounted prisms that can be slid in and out, I think they're marked T2 10, T3 20 and T4 50, something like that (edit: it's t4 10, t2 20, t4 50). They also made prisms for their ULWD objectives with the same layout but they're marked ULWD based on the photos I've seen.
Side-note, I tried mixing the objectives around a bit. The 50x gave nice DIC on the 10x prism, the 20x was fine on the 10x and 20x prisms, but the 10x wouldnt' focus on the 20x prism. All of course with doubling, to (as far as I could tell) the same degree.
There's no sign of tampering on the prisms, and I think it would be pretty hard to have messed with the prisms in them. I can't be sure though.
Last edited by Scarodactyl on Sat Jun 20, 2020 5:02 am, edited 1 time in total.

microb
Posts: 729
Joined: Wed Mar 27, 2019 6:39 am

Re: Epi DIC woes

#25 Post by microb » Wed Jun 03, 2020 12:08 am

So here is a objective turret I had tested with DIC. Now looking it up to find your NeoSPlan objectives, I find that the NeoDPlan M26's shown here are not checked marked for DIC like your NeoSPlans M26's. Didn't know that. They seem to work. All these require NFK relay lenses to correct for color, but the double image is of the same color not cyan and green in different places. So it's not chromatic aberration.

The prisms in these holders are held at an angle 5 degrees or so when levered into the light path.
Attachments
NeoDPlanM26_not_for_DIC.JPG
NeoDPlanM26_not_for_DIC.JPG (127.6 KiB) Viewed 15878 times
IMG_4229_cropped_1k.jpg
IMG_4229_cropped_1k.jpg (119 KiB) Viewed 15878 times

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

Re: Epi DIC woes

#26 Post by Scarodactyl » Wed Jun 03, 2020 2:24 am

That's interesting--I feel sure I've seen pictures of Neo Dplans marked 'IC'.
My neo splans do have "NIC" at the end, as well as "IC 10", "IC 20" etc marked below.

I swapped my Nikon head onto it yesterday, but before then I was shooting entirely with an NFK and WK eyepieces, so those corrections were lined up.

Interesting how it describes the msplans as being f-180 for the BH2-UMA or the BH2-MA, the NeoSPlans as being f-180 for the BH2-UMA, and then Neo Splan NICs as just f=180.

How were your results? Any hint of doubling?

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

Re: Epi DIC woes

#27 Post by PeteM » Wed Jun 03, 2020 2:47 am

A couple thoughts as we're grasping at straws:

- You might want to check the polarizer back in the incident light path. It's possibly in a position where one that's degraded (sometimes happens with heat or handling) might introduce a second reflected light source??

- Since the doubling is pretty predictable, I wonder if you could image something like a pin hole and then carefully go back through the entire optical train moving / rotatating / masking etc. components until you find one that moves the doubled images to a different angle?

microb
Posts: 729
Joined: Wed Mar 27, 2019 6:39 am

Re: Epi DIC woes

#28 Post by microb » Wed Jun 03, 2020 4:13 am

Here is a post where I hooked up a B&L DIC filter adapter. (viewtopic.php?f=5&t=7327)

Attached pictures here, shown below, have two not published in the B&L thread that are from the 10x NeoDPlan with and without DIC for comparison. I do see blurring on the Olympus 10x DIC. So I'd have to unbox the frame and set this back up again to see more. But here's what I got from my hard drive for now:

First is 10x Olympus DIC
Second 10x no DIC
Third 10xB&L DIC
Attachments
10xolynosedic_1k.jpg
10xolynosedic_1k.jpg (77.72 KiB) Viewed 15854 times
10xnodic_1_1k.jpg
10xnodic_1_1k.jpg (101.27 KiB) Viewed 15854 times
10xbnldic_1k.jpg
10xbnldic_1k.jpg (99.27 KiB) Viewed 15854 times

microb
Posts: 729
Joined: Wed Mar 27, 2019 6:39 am

Re: Epi DIC woes

#29 Post by microb » Wed Jun 03, 2020 4:26 am

Here are two more 10x NeoDPlan, with and without DIC. With is less satisfying.

Side note:
The purple tint in the middle of these two images is not from the Olympus set up. The blobs are from the USB camera having the Fuji defect where the green pixels on the sensor are "lensed" to get light from a range of angles for autofocusing. The red and blue portions of the sensors get light directly. The result is less green when looking at bright background flares in your artistic photography compositions -- which annoyed those who bought the expensive Fuji DSLRs. In this case the purple is revealed with focused reflected microscope light. Less green means purple and if you look at both images you see the purple blob. It's very faint but it is there.

Basically, if you bought any USB 2.0 4MP microscope cameras, you might have found out where the undesirable but technically working Fuji sensors chips were dropped onto the open market and put into use. So check you cameras on white backgrounds for purple blobs that can sometimes turn into a big plus sign in the middle of your images.
Attachments
OLYNoseDIC10x_1K.jpg
OLYNoseDIC10x_1K.jpg (89.16 KiB) Viewed 15854 times
OLYNoseNoDIC10x_1K.jpg
OLYNoseNoDIC10x_1K.jpg (109.19 KiB) Viewed 15854 times

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

Re: Epi DIC woes

#30 Post by Scarodactyl » Wed Jun 03, 2020 6:15 am

Thank you very much for posting these!
That last image shows definite doubling up close. Maybe this is an inherent issue with Olympus's dic prisms?

Post Reply