Episcopic TTL microscopy compared Bio vs Industrial

A forum to ask questions, post setups, and generally discuss anything having to do with photomacrography and photomicroscopy.

Moderators: Pau, rjlittlefield, ChrisR, Chris S.

JohnyM
Posts: 463
Joined: Tue Dec 24, 2013 7:02 am

Episcopic TTL microscopy compared Bio vs Industrial

Post by JohnyM »

First of all - images are very soft. Magnifications tend to be higher, while NA's tend to be smaller than bio objectives. But working distances are longer
Test subjects are: Bio - semi transparent, semi reflective - Pinus leaf. Industrial - completly opaque - coin.
Now results:
1) Darkfield (Top row) - might be useful in some bio aplications. Images are vibrant and contrasty, but produce lots of specular higlights.
Works briliantly with opaque subject.
This technique fall apart with more complex 3D subjects as light path is obscured by subject itself. Very light inefficient.

Brightfield (Bottom row) - Bio images are hazy, contrast is medicore, they absolutely lack color, highlights are terrible. Unusable.
Works briliantly with opaque subject.
This technique works well with any kind of subject structure, best light efficiency of them all (50% of light that stuck the subject).
Image

2) Polarizer in (Top row) - very slight improvement over brightfield in both aplications at cost of 1 stop of ight.
Crossed polars (Bottom row) - That's the way to go with Bio
Vibrant, contrasty, very light inefficient.
Image

3) Tint plate... (Top row) tints everything to user selected color... with some parts of the subject being slightly different color. Special technique, might work well with some subjects. Exposure is actually faster than crossed polars.

DIC (bottom row) - absolute disaster with bio subjects.
Works well with opaque subjects, but lack that "wow" effect seen with bio dic.
Image

4) Fluorescence - usually used to detect contamination in industrial subjects. This time, it produced no image with coin (was cleaned before imaging) while bio subject shines nicely as supposed. Used cube was B-2A.

DIC + tint plate - AFAIK in this configuration it's used to detect how subject vary in height by color change.

Image



TL DR: if interested in biology only, stick with brightfield illuminator with polarizers. Rest of techniques are costly and next to unusable.

abednego1995
Posts: 72
Joined: Tue Nov 01, 2016 11:53 pm

Post by abednego1995 »

Splendid comparisons!
What are the magnifications you worked with here? I tend to see low mags in brightfield with lower contrast (probably internal reflections of the objective masking the sample reflections), but that improves with higher mags (higher NA).

Cheers,
John

ChrisR
Site Admin
Posts: 8563
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Interesting:smt023 . Impressed by Crossed polar bio. I daresay epi DIC will appeal to coin folk.
Thanks.
Chris R

JohnyM
Posts: 463
Joined: Tue Dec 24, 2013 7:02 am

Post by JohnyM »

Thanks for input. Thought nobody is interested with it.
First comprarision was done with Nikon BD Plan DIC 5x/0.1 through 1,5x teleconverter on APS-C A6500 sensor.

Indeed, things look better with higher magnifications, but i have a different theory behind why it's like that.

Here's comparision with Nikon BD Plan ELWD 50x/0.5 through 1,5x teleconverter on APS-C A6500 sensor. Subjects are:
Bio: fern leaf, bottom side.
Industrial: precision m5 shim

Darfkfield and brightfield
Image

Polarization and DIC
Image

DIC with tint plate and Fluorescence*
This time industrial subject shines... with bio contamination. AFAIK that's main use of Fluo in wafer industry, to identify if it's and error or contamination.
Image

My observations are identical as with 5x, but indeed everything does looks slightly better imo.

JohnyM
Posts: 463
Joined: Tue Dec 24, 2013 7:02 am

Post by JohnyM »

Subjects respond VERY well to sharpening and "dehaze" feature of CameraRAW. Here are 100% crops from bio parts, that were optimally (usually heavy) processed with those techniques.

Reason why bio is problematic with epi TTL is because it's partially reflective and partially transmitting. Why we usually get 2 competing images. One of surface and one of inside of cells. "Inside" usually provides the color, and "outside" texture. Sadly, those are usually mutually exclusive.

Of course, this isn't strict rule. I suppose arthropods, with their reflective armor will respond very well to this technique and will be next follow-up.

Darkfield
Image
Here we can see, there is some nice diffused light that create image of the internal parts, while ignoring cell surface. But "proper" darfkield light, as indended, creates enormous higlights in every surface dent. And those are everywhere in bio subjects. Should be usefull in some subjects. I imagine insect eyes or hairs should work well. To be tested.

Brightfield
Image
Here we can see, that we got very little of "inside", and very much of "outside". In this case outside is rather flat and uninteresting.

Crossed polars
Image
Basically a brightfield with reflections killed. Very nice image of the "inside" with "outside" practically invisible.

DIC
Image
Now, in contrary to test with 5x, DIC does seem like an improvement over plain brightfield. I need to take a closer look at it in the future.
It seems like it's reversed cross-polar (which is quite funny actually...)
Killed all the "inside", while making "outside" more pronounced. I can imagine, it might be usefull with some subjects. Could be a "killer" with some insects and as Chris pointed, coins.

Tint plate DIC
Image
Only if one loves colors.

abednego1995
Posts: 72
Joined: Tue Nov 01, 2016 11:53 pm

Post by abednego1995 »

Since you already have a nice setup, what happens if you add a small cylindrical diffuser just outside of the image illuminated with the BD light path?
(Perhaps a miniscule semi-transparent sphere like a miniture ping-pong ball would work better?)

I sometimes see star shaped points of light in BD images that probably come from the internal lenses in the outer light path. Thought that it might be a nice way to squeeze in diffused light in otherwise near impossible WDs.

Cheers,
John

JohnyM
Posts: 463
Joined: Tue Dec 24, 2013 7:02 am

Post by JohnyM »

Im not sure what you mean.
You want to futher diffuse BD light path? Or transform it into typical macro-rig side lighting?

If 2nd is the case, i can freely use a ping-pong / diffusion sheet with some Jansjo lamps. WD's are quite manageable, not a mitutoyo, but not too shabby. But i think i also have normal 20x BD with short WD so i could use it for test purpose.


Actually set i've purchased is quite interesting / rare sight in old finite era.
DIC prisms are for (210mm CF Nikon, WD is recalled from memory):
-5x WD ~9mm
-20x ELWD WD ~10mm
-40x ELWD WD ~10mm
-60x ELWD WD ~6mm
-40x PlanApo WD ~0,5mm

So i understand that it was assembled with long working distances in mind as it lacks BD 10x objective (which, for some reason wasnt produced as LWD version and have only ~3mm WD), but still have high-res option of 40x PlanApo 0,8

I've scored a lucky shot here, as it's close to the version i would assembled myself, but replacing 60x with 10x.

abednego1995
Posts: 72
Joined: Tue Nov 01, 2016 11:53 pm

Post by abednego1995 »

Gosh, I'm so bad at verbally describing things!
Well, I'm thinking of this kind of setup.

Image

Those ELWD objectives, even with the 60x have ample WDs. I've enjoyed using the non-BD versions in the past.
But I'm currently mulling the possibility using higher NA alternatives (like the CFI60 LU/TU BD Plan Fluors or BD Mittys) and in those cases normal ping-pong diffusion begins to become overly inefficient.

Would illuminating a diffusing cylinder near the object with the BD light path be feasible?

Regards,
John

ChrisR
Site Admin
Posts: 8563
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

I have tried tiny cylinders of paper in that position. It "worked" but I found it too fiddly to persist with!
Chris R

abednego1995
Posts: 72
Joined: Tue Nov 01, 2016 11:53 pm

Post by abednego1995 »

Thanks Chris!
Working with flat large specimens, yes I think those cylinders will be a huge pain... perhaps some silicone tubing might work. Must experiment.

I'm currently thinking of photographing things smaller than the front lens diameter so just making a fairly long cylinder tacked on the objective, and inserting the specimen in it by a thin rod on a micropositioning stage probably would do the job.

Thanks for the input!

Cheers,
John

ChrisR
Site Admin
Posts: 8563
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

The surgical silicone rubber tubing I use as elastic to take up slack in mechanisms is translucent.
Next time doc puts a drip in your arm for something, demand the entrails...

Ebay - "clear silicone O rings" yields: from 4mm x 2mm
Image
Chris R

Post Reply Previous topicNext topic