3d printed Micro Photo/Video setup/supports

Have questions about the equipment used for macro- or micro- photography? Post those questions in this forum.

Moderators: rjlittlefield, ChrisR, Chris S., Pau

econopotamus
Posts: 13
Joined: Sun Aug 27, 2017 8:31 pm

3d printed Micro Photo/Video setup/supports

Post by econopotamus »

Short Version: Lookee Here!

I've been lurking here about 1.5 years, reading and learning. For example I learned a ton from this post and set off to build my own setup. I'm an engineer and even had a project that needs high speed micro-video to watch what's going on in a (micro) fluid flow. I built a setup with a Canon DSLR and microscope objectives but found that with micron-level resolution the long lens stack was picking up vibration from some of the equipment that made the video hard to analyze.

Thus I designed and built a 3d printed macro-rig support using standard 15mm camera tube. Photos and links. The new setup travels and mounts much easier (tripod connection is now under the center of mass) and the vibration is gone. I can now take microscopic video so stable I can do frame-to-frame image subtraction to track motion.

I wouldn't have been able to build this whole setup without reading some of the great material on this board about coupling microscope objectives to DSLR and I wanted to thank you all for having such a great board and discussions!! I made sure to link back here when I posted the 3d designs of the project.

Pau
Site Admin
Posts: 6051
Joined: Wed Jan 20, 2010 8:57 am
Location: Valencia, Spain

Post by Pau »

Hi econopotamus, welcome aboard!

This is a very nice project. Would be great to see your results if doable.

Optically I find two issues at your setup:

- I see a long extension tube between the camera and the lens. This focuses the lens much closer than infinite while infinite corrected objectives are designed to work with a lens (converging lens or "tube lens", in your case the Canon zoom) focused to infinite. Because low power objectives are more forgiving about the tube length it could work reasonably well with your 5X but could be terrible with high magnification objectives. Just remove the tube and focus the zoom to infinite in manual mode.

- You use an old Leitz objective. Good as they are, they are designed to work with a compensating eyepiece (Leitz Periplan) that completes the optical correction of the objective aberrations. With your setup you can expect chromatic aberration most visible at the image periphery.
Pau

econopotamus
Posts: 13
Joined: Sun Aug 27, 2017 8:31 pm

Post by econopotamus »

Pau wrote:Hi econopotamus, welcome aboard!

This is a very nice project. Would be great to see your results if doable.

Optically I find two issues at your setup:

- I see a long extension tube between the camera and the lens. This focuses the lens much closer than infinite while infinite corrected objectives are designed to work with a lens (converging lens or "tube lens", in your case the Canon zoom) focused to infinite. Because low power objectives are more forgiving about the tube length it could work reasonably well with your 5X but could be terrible with high magnification objectives. Just remove the tube and focus the zoom to infinite in manual mode.

- You use an old Leitz objective. Good as they are, they are designed to work with a compensating eyepiece (Leitz Periplan) that completes the optical correction of the objective aberrations. With your setup you can expect chromatic aberration most visible at the image periphery.
Wow, it's like you're in my lab looking over my shoulder. I do have multiple higher objectives but the results were not great in quick tests. I will get rid of that tube!

As for the Leitz I haven't noticed the chromatic aberration. I'm doing extremely high contrast work (super bright light shining through a thin transparent construct and tracking moving shapes that block the light) but I'll go back and look for that. I have a bunch of objectives I can try, fortunately for me, because it turns out the probe station I have uses infinite focus objectives.

I need to carve out some time to take some pictures that are non-proprietary and interesting! I will look for some time :)

Thanks so much for your observations, that's why this forum is so great!

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Interesting stuff! :D :D
I expect a number of us would have jumped on you for the same reasons Pau mentioned :).
Are you aware of CA on the higher magnification lenses? If your bright objects have colored edges....
What objectives are you using at higher Magnifications. There are options and pitfalls...
I assume you're using proper glass thickness "tanks" for the fluids, ...?

Do you do it monochrome anyway perhaps?
How are you visualizing the flow - suspensions, or some micro-Schlieren technique :?

I'm glad you came back. Bugs are OK, but if you're allowed to show us more, that would be great.
Chris R

econopotamus
Posts: 13
Joined: Sun Aug 27, 2017 8:31 pm

Post by econopotamus »

I went back and looked at the videos and the Leitz is indeed giving visible R/G/B banding. I hadn't taken note because it didn't effect the data I needed to take.

I'll get some info on the other lenses for ChrisR

The fluid flow paths are in custom made tiny flow channels within Acrylic or Lexan. So if you were worried about pressure bursting or anything there isn't really any chance of that. Flow channels get much stronger as they get smaller and these are through single pieces of strong materials.

Pau
Site Admin
Posts: 6051
Joined: Wed Jan 20, 2010 8:57 am
Location: Valencia, Spain

Post by Pau »

...and the Leitz is indeed giving visible R/G/B banding
The fluid flow paths are in custom made tiny flow channels within Acrylic or Lexan...
...and these are through single pieces of strong materials.
Maybe the major part of that R/G/B banding could be due to the plastic channels more than to the Leitz lack of compensation, even more if they are curved or thick or both, the lateral chromatic aberration due to the objective will show as red/blue halos symmetrical respect to the center of the image and increasing towards the periphery and mostly shown at high contrast borders.
Pau

econopotamus
Posts: 13
Joined: Sun Aug 27, 2017 8:31 pm

Post by econopotamus »

Thanks Pau, I'll get more details and look more closely at what is going on!

Please keep the comments coming, I do not take them as criticism and instead love that I'm learning stuff after posting that will make my work better. I also apologize in advance for any delays in getting more information, I need at least a day or two of coding and CAD design work to get some long lead time items moving before I can put a lot more into the microvideo setup :roll:

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Ah, no - I was thinking that the liquids were aqueous and contained in little glass walled tanks. The Objectives will work best with the intended glass thickness. So not what you're doing at all.
Some objectives are designed with an adjuster for a range of thicknesses, like 0-3mm, so may suit your subject.

RGB sounds like dispersion - as in a prism. . More reason to try single wavelengths , maybe.

How does the flow "show up"?

Using any polarizers? We were mulling slightly over optically active (chiral) fluids recently.
Chris R

econopotamus
Posts: 13
Joined: Sun Aug 27, 2017 8:31 pm

Post by econopotamus »

I see, yes that is something different. I'm a little uncertain as to why the thickness would be an issue beyond refractive bending, which I've already calculated. I can remove than in my data (and could correct the images with a bit of python if I wanted). Is there something else going on that you are thinking of?

Single wavelength does sound like a good idea. I've got a pretty big HeNe laser I can put on the same table, any thoughts on how I figure out what the maximum light is I should allow to hit my sensor? Maybe I should put in a neutral density filter and just slowly adjust and stay away from visual saturation.

The flow is visible in bubbles or in dark particles, both of which I can add. I recently learned how to make neutral buoyancy brightly colored particles, although getting those just right would require temperature controlling the whole system which I probably don't want to do.

I haven't tried polarizers. In theory Schlieren works for turbulent flow (which is what we have) but I've never seen someone use it successfully at the micro scale on anything less than a sonic shockwave (which we certainly don't have).

ChrisR
Site Admin
Posts: 8668
Joined: Sat Mar 14, 2009 3:58 am
Location: Near London, UK

Post by ChrisR »

Thickness- Spherical aberration in the "glass"
Last edited by ChrisR on Thu Aug 31, 2017 9:47 am, edited 1 time in total.
Chris R

mjkzz
Posts: 1681
Joined: Wed Jul 01, 2015 3:38 pm
Location: California/Shenzhen
Contact:

Post by mjkzz »

great stuff, some day I am gonna get a 3D printer.

One side note, you can buy off the shelf components for the 15mm rod system, just search on eBay for "lens support for 15mm rod" and here is a pic of my setup. Of course, this is going to ruin the fun of making it.

Image

Chris S.
Site Admin
Posts: 4042
Joined: Sun Apr 05, 2009 9:55 pm
Location: Ohio, USA

Post by Chris S. »

Econopotamus,

(A different Chris, here.)

Welcome to the community! Great post--well-thought-out details are notable throughout your setup. I particularly appreciate your highlighting the usefulness of 3D printing in creating your setup. From private communication, I know that at least a couple of forum members use their 3D printers heavily. 3D printing has come up in prior threads, but this technology likely deserves much more discussion than it's received, as a potentially ubiquitous problem-solver for the macro/micro photographer. As I have a fabricator who works in metal, does outstanding work, and charges modestly, I've little need of a 3D printer. But for others who don't have such a fabricator, 3D printing often seems attractive.

Two minor questions: Is your Canon zoom lens focused at infinity? If not, it should be. I ask because the original implementation that included an extension tube would logically have also extended the focusing helicoid to its closest focus. But as Pau explained, with infinite objectives, we want the converging lens to be focused at infinity.

Also, is image stabilization turned on for your Canon zoom, as it appears in your images? I know little of Canon lenses, but if this were a Nikon zoom, this setting would need to be in the "off" position. (Granted that given the original inclusion of the extension tubes--likely without electrical contacts--this point would be moot. And you probably know very well to turn this setting off after removing the tubes--just being sure.)

As ChrisR said, a big issue, when shooting through any medium whose composition and thickness were not designed into the lens, is spherical aberration. The impact of this spherical aberration is--counterintuitively--a step function, because spherical aberration scales to the fourth power of numerical aperture (NA). At NA's of 0.30 and below, it doesn't matter; at higher NA's, it matters immensely.

If I were in your shoes--apparently, making research-supporting images of microfluidic assemblies--I would either use microscope objectives whose NA is 0.30 or less (where spherical aberration has little effect--effectively, 10x and lower magnifications), or at higher magnifications, use microscope objectives with correction collars that facilitate dialing out the thickness of intervening material.

--Chris S.

rjlittlefield
Site Admin
Posts: 23561
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

econopotamus, let me add my own welcome! :)

Other people have given you good information, but I see still some gaps in the discussion.
econopotamus wrote:I'm a little uncertain as to why the thickness would be an issue beyond refractive bending, which I've already calculated. I can remove than in my data (and could correct the images with a bit of python if I wanted). Is there something else going on that you are thinking of?
When looking through dense medium, there are several related but different issues to be considered.

The discussion so far has concentrated on spherical aberration, which will soften or even ruin your images with thick glass and/or high NA objectives. This aberration causes loss of contrast. It varies in proportion to glass thickness to the first power and NA to the fourth power, so sufficiently thick glass will cause a problem even with low NA objectives such as 10X NA 0.25. For critical work, there is no substitute for an objective that is specifically designed to look through thick glass, preferably with an adjustable correction collar as mentioned earlier.

A second issue, not yet discussed, is that looking through thick glass may (or may not!) give you a geometrically distorted view of the subject that you're looking at. I notice that you mention "refractive bending, which I've already calculated", and I wonder if this means that you've tried to correct the geometry by taking into account the bending of slanted light rays away from the center of field. If so, then be aware that this issue is much more subtle than it appears at first, and all of the obvious "corrections" are very likely to give the wrong answer. This is because the required correction depends on the entrance pupil location of the objective, and that location is very rarely at one focal length away from the focus point as implied by the too-simple diagrams in most optics discussions. With a "telecentric" objective, the entrance pupil location is effectively located at infinity and no correction at all is required. Most objectives are not telecentric but lean in that direction. Those would require a correction that is more than zero but less than the one-focal-length-away case. Unfortunately the information required to compute the correction is never published, so for critical work you would have to measure a known object under the same viewing conditions, or determine the entrance pupil location yourself by some careful measurement that is too complicated to discuss right now.

Also,
Single wavelength does sound like a good idea. I've got a pretty big HeNe laser I can put on the same table, any thoughts on how I figure out what the maximum light is I should allow to hit my sensor? Maybe I should put in a neutral density filter and just slowly adjust and stay away from visual saturation.
With laser illumination, I expect that you'll get horrible results no matter what you do. This is because the long coherence length of a laser will cause your entire field to be covered by intense speckle. Moving particles within that field may be visible, but their appearance is not likely to be anything close to what your analysis software is expecting. See http://www.photomacrography.net/forum/v ... hp?t=16847 for a quick look. For the sort of work that you're doing, "monochrome" really means using narrow bandwidth but incoherent illumination. Green-filtered white light is a standard approach. Green LED should be OK also.

I hope this helps!

--Rik

econopotamus
Posts: 13
Joined: Sun Aug 27, 2017 8:31 pm

Post by econopotamus »

So much to reply to! I'll post picture materials first then the wall of text :)

First of all, I also have 20x and 50x dark field capable objective lenses (by Leitz):
Image

Here is a photo of the system with the 20x objective on and without the focusing tube, to show I can indeed take advice :D :

Image

For comparison and to let people see the optical results I figured I would take pictures of some pixels on my Samsung Galaxy S8 with the 5x and 20x objectives (I don't have a bug collection yet).

5x objective:
Image

20x objective:
Image

Thanks for the advice everyone! Next post will be replies to specific questions...

econopotamus
Posts: 13
Joined: Sun Aug 27, 2017 8:31 pm

Post by econopotamus »

ChrisR:
Thickness- Spherical aberration in nte "glass"
Okay, I think I get what you're talking about now. Right now I address that computationally. Great point though! And per the comments from rjlittlefield it looks like it may be more complicated than I have been calculating.

mjkzz:
you can buy off the shelf components for the 15mm rod system, just search on eBay for "lens support for 15mm rod" and here is a pic of my setup. Of course, this is going to ruin the fun of making it.
Yep, I'm probably guilty of wanting to build it myself although I also made the parts because I have this crazy idea that in the long term I'm going to add servos and do automated microphotographic focus stacking, so I wanted to add some bolt holes and clearances as part of that future idea. No idea if I'll actually find an excuse to really build all that. It was also cheaper to 3d print (by a LOT) although of course one must count time spent as well. Since I also managed to learn a lot during the engineering I'm happy with DIY for now. Definitely could imagine buying the components in the long run.

Also, that's a nice looking lens on yours!

Chris S.:
<3d printing comments>
Yeah, once you have a 3d printer the math changes on a lot of things. Printing this build cost about $8 of input plastic. I've been printing tons of stuff. Fortunately my engineering work justified buying the 3d printer.

Is your Canon zoom lens focused at infinity?
Now it is :D . The macro tube was leftover from my path of development of the system, but now is removed. Also image stabilization is off.
I would either use microscope objectives whose NA is 0.30 or less (where spherical aberration has little effect--effectively, 10x and lower magnifications), or at higher magnifications, use microscope objectives with correction collars that facilitate dialing out the thickness of intervening material.
This is a great comment. I'm going to read up more on these collars and try to understand the implications.

rjlittlefield:
<discussion of correction and telecentric lenses>
That's complicated, I think I'll take your advice and get some sort of known pattern under the lens so I can correct things better.
<discussion of laser and speckle>
Ah, yes I have encountered speckle before. I suppose I can either try speckle reduction (e.g. spinning ground glass) or the LED approach. I don't have any high intensity LEDs handy but I can put something together. I looked at your laser results - yup, speckle. Highly educational!

Post Reply Previous topicNext topic