Maybe a way to get the perfect image with a DSLR?

[hip17702@gmail.com]

Dear members,

Most microscopy pictures I saw by a DSLR camera are ...let's say.... below expectation. But pictures with a small sensor (5 MPX) are quite OK. Some very expensive dedicated camera (16.25MP for more than $11000) are stunning. How comes that DSLRs are so bad at the job, even with a quality sensor up to 60 MP? I may have a clue, and a solution that needs some testings.

First thing first, aren't you amazed that the tube you need to use a DSLR on your favorite microscope is so long, so bulky? When you see videos from our guest, Oliver -all hail to him - don't you have the feeling he has recycle a submarine periscope and throw his Canon DSLR on top of it? And it seems to be a common annoyance for all type of microscopes. More, if you don't have the exact length with the needed optical adapter you have a lot of difficulties to find the focal plane. Oliver videos show a purple fringe on one side of the object, and a yellow one on the other side, same for pictures taken by the Canon DSLR. Some members posted pictures where the center of the image is somewhat correct, but the borders are blurry and distorted (pincushion deformation).

There are two main factors creating these deformations, IMHO... The lengthy tube, even manufactured with care, cannot ensure the top lenses are exactly on the optical axis. Any small misalignment, even with a very small angle creates the purple-yellow fringes. I got this effect on my microscope, and slightly tilting the tube worsen the fringes. If you try to shorten the tube, to minimize the misalignment effect, you need to set the microscope objective above its nominal working distance, creating a severe pincushion deformation.

I find a solution to these problems on my Optika IM-3 (inverted microscope) using a SONY Alpha 7 R4 (61 MP) to get sharp and crisp images even at the border. The trick I use is quite simple: a 2x tele-extender! The image formed by the objective is 22mm, but to frame it properly in a full frame sensor you need 44mm... hence the 2x extender "trick". The distance between the "trinocular port" and the extender is only 15mm when the microscope objective is at its nominal working distance. I mean I set the image on focus looking through the eyepieces, lock the focus knob and then put the camera with the extender above the video tube until getting a sharp image. Distance is around 15mm in my configuration.

If you are interested to test this you need a quality 2x extender for a full frame size sensor, and 1.4x extender for an APS-C size sensor. Using a 2x extender on APS-C should work, but you will get a somewhat enlarged image. If it works... because I have only tested this on an inverted microscope with infinity objective. It may be different for upright microscope with 160mm objectives or other optical configurations. So if you try this, please let me know the results you have.

[Hobbyst46]

There are two main factors creating these deformations, IMHO... The lengthy tube, even manufactured with care, cannot ensure the top lenses are exactly on the optical axis. Any small misalignment, even with a very small angle creates the purple-yellow fringes. I got this effect on my microscope, and slightly tilting the tube worsen the fringes...

It would help to know some details.

Is the lengthy tube on your microscope manufactured with care - and you can nevertheless tilt it ?

Is the "periscope tube" firmly attached to the frame ?
Does "tilting" move the tube - or the whole microscope ?
Is the slide shifted on the stage upon tilting ?
Is the camera firmly attached directly to the tube, or are there some adapters in the way ?

[hip17702@gmail.com]

Sorry if I wasn't clear...

Is the lengthy tube on your microscope manufactured with care - and you can nevertheless tilt it ?

. Yes, because the tube is a part that locks inside "the trino port", and it is possible to tilt it a bit when the lock screws are loose. It was a test to see if tilting has an effect on image fringing.

Is the "periscope tube" firmly attached to the frame ?

When screws are locked and the tube secured in place, on my system a small amount of tilt remains, creating some color fringes. This fringes cannot be seen on eyepieces.

Is the camera firmly attached directly to the tube, or are there some adapters in the way ?

For DSLR a T2 ring is used to attach the camera on the tube.

[daruosha]

Can you show us a picture of the actual setup and a sample picture?

[Hobbyst46]

Thanks for the detailed answers, and for the idea of extender.

[hip17702@gmail.com]

Please find here a picture of the actual setup and a sample picture: https://we.tl/t-qQrZ2Vm6mO

The files will be available for one week. If you have any question please do not hesitate.

[75RR]

Zip file!

Be easier to include them in a post. Max size is 1024 x 1024 and 512KB Just drag and drop

[Malpi]

a 286MB zip file even !! ( at 550KB/s adsl ,,,wake me up in a while )

[hip17702@gmail.com]

I'm sorry... problem is to share 61MP RAW pictures (9504x6336) without compression to see what the quality really is with the system.

[Malpi]

A little while later :-

Ah yes, I see what you did now, most of it was the megga dng (which I had not expected ) which I will view with interest later

The little jpg is interesting thank you.

[Malpi]

The trick I use is quite simple: a 2x tele-extender!

thanks for the tip, I remembered I had a 2x/3x tele-extender somewhere still here from my filmSLR days in the '60s !
It only took me 3hrs to find it ,,,, and it works
much less precarious than the stack of tubes and the helicoid I was using.

[Scarodactyl]

Most systems will direct project onto aps-c with no trouble, no telecoverter needed unless you want to cover full frame. In systems where this is possible (no additional corrections needed, bit enough image circle--ie nikons, almost all stereo systems, probably most modern infinity systems) it is probably the best possible approach--big field, no extra glass and most importabtly no extra dust to show up in your stacks. Teleconverters have been explored by the folks over at photomacrography, and i believe the consensus was that they need to be of pretty high quality (and perhaps recent vintage) to not add troublesome abberations, but are a really good option if you use a good one.
I disagree a bit with the premise though, as it does not match my experience. The best photomicroscopists I follow all use dslrs or mirrorless cameras (outside of specialty stuff like fluorescence images).

[Malpi]

To avoid any misunderstanding :
When I said "it works" in respect of my vintage extender I was not referring to any claims of "perfect image" as in the topic title !
I was referring to the reduction of precariousness wrt to the periscope effect.

[hip17702@gmail.com]

I understand now that my topic's title is not the wisest.... "A way to get a better image" rather than "a way to get the perfect image" should have been more appropriate. My apologies... Hopefully Malpi gets me right on that.

To be more explicit, my microscope is an Optika IM-3: https://www.optikamicroscopes.com/optik ... duct/im-3/ and in order to take pictures from a DSLR the following accessories are available: an extension tube https://www.optikamicroscopes.com/optik ... uct/m-699/ and an optical adapter https://www.optikamicroscopes.com/optik ... uct/m-173/. With an inner diameter of 28mm and the optical adapter taking place inside this 28mm tube, the results I got were below my expectations. The base of the tube is "secure" by only one 5mm screw, the optical adapter slips inside the tube and is "secure" with another screw, and on top of that you have to place your valuable DSLR weighting and costing more than a webcam. After one day of fiddling with this system, singing to myself "Jingle lenses, jingle lenses..." I didn't get any useful results. I got the idea to use an extender and the results were encouraging, I suppose due to a shorter tube and the sturdy coupling system (compared to two screw, long tubes and a "slip-in" optical device).

When Scarodactyl said "Most systems will direct project onto aps-c with no trouble, no telecoverter needed unless you want to cover full frame." he is certainly and definitively right. Problem, I have only access to one system that provides a 22mm field image. Even on an APS-C sensor the image is vignetting, and on a full frame sensor you get a nice 22mm circle inside a dark frame, providing you have a lot of distance between the DSLR and the microscope. Also, my microscope not a top of the range instrument, I have to find some "poor man" solutions in order to get some usable results.

Last but not least, this idea may work for other people with affordable "not top of the range microscope" and maybe, by using a photographic extender they already have, they may spare some money. If only I had known I will have spare $300 in accessories for nothing Well, if it works on other systems... it is why some testing are useful.

[Hobbyst46]

To be more explicit, my microscope is an Optika IM-3: https://www.optikamicroscopes.com/optik ... duct/im-3/ and in order to take pictures from a DSLR the following accessories are available: an extension tube https://www.optikamicroscopes.com/optik ... uct/m-699/ and an optical adapter https://www.optikamicroscopes.com/optik ... uct/m-173/. With an inner diameter of 28mm and the optical adapter taking place inside this 28mm tube, the results I got were below my expectations. The base of the tube is "secure" by only one 5mm screw, the optical adapter slips inside the tube and is "secure" with another screw, and on top of that you have to place your valuable DSLR weighting and costing more than a webcam....

Thank you for the information! disclosure of these details is valuable service to the community, especially to folks who are in doubt about buying an old quality microscopes vs a new of-the-shelf affordable microscope. The mechanics are at least as important as the optics, and should get the same attention from potential buyers...

[75RR]

Please find here a picture of the actual setup and a sample picture: https://we.tl/t-qQrZ2Vm6mO

The files will be available for one week. If you have any question please do not hesitate.

So can you post a photo of your setup in the forum so members do not have to download that zip file?

[hip17702@gmail.com]

Here you go:

Configuration.jpg (134.53 KiB) Viewed 9043 times

[Hobbyst46]

Thanks for the photo. Does the 28mm bore socket in the gray-colored stand receive a dovetail, a thread, or is it just smooth ? so the 5mm locking screw presses onto a round surface ?

[hip17702@gmail.com]

It is just smooth and the locking screw presses onto the inner ring. Depth = 8mm. With the object centered on the eyepiece I was never able to center it on the DSLR. There are currently 3 screws, I drilled two more holes which I tapped with M5 thread. Now when the object is centered on the eyepiece it is also centered on the DSLR.

[Scarodactyl]

To be more explicit, my microscope is an Optika IM-3: https://www.optikamicroscopes.com/optik ... duct/im-3/ and in order to take pictures from a DSLR the following accessories are available: an extension tube https://www.optikamicroscopes.com/optik ... uct/m-699/ and an optical adapter https://www.optikamicroscopes.com/optik ... uct/m-173/. With an inner diameter of 28mm and the optical adapter taking place inside this 28mm tube, the results I got were below my expectations.

Thanks for the further info! Interesting that they won't cover aps-c. I know some of the amscope objectives will, but maybe the head i troduces some vignetting here. Either way the center is probably the best.
This is a great solution for this kind of microscope, where the seller will at most offer these generic 2x dslr adapters. Those can really wreck your image if they aren't well matched to the system (and they almost never are--I would guess the issue here has nothing to do with tilt on the adapter, and everything to do with the quality of these adapters). I imagine we'll be linking this thread for years to come whenever someone wants to attach a dslr to a chinese-made scope without an in-system dslr solution.

[hip17702@gmail.com]

Reading your post I just realize you are absolutely, definitively and positively right on target my dear Scarodactyl! After a quick search I found the exact same adapter with AmScope branding! And there is more: if you look at this adapter designed for my microscope https://www.optikamicroscopes.com/optik ... t/m-620-1/ it is obvious it is technically far more advanced than the DSLR adapter. The coupling part is conical (instead of cylindrical on the DSLR tube) to insure proper centering and locking, this adapter is focusable and the lenses are obviously more than 28mm in diameter. Maybe using a DSLR on this type of microscope is rarely a user request, and a good C-mount for scientific camera is what they sell most? If it is the case, they just have to propose a generic, off the shelf adapter not really tailored for the microscope. Also, the AmScope adapter is to be used with an upright microscope having 160mm objectives, and the one I use is inverted with infinity objectives. If both adapters came out the same shelf labelled "generic DSLR adapter", no wonder I cannot get any usable results using it. Well, it's just a guess... but some evidences point a straight finger at it.

Now the interesting part of your message:

I imagine we'll be linking this thread for years to come whenever someone wants to attach a dslr to a chinese-made scope without an in-system dslr solution.

Well, I'm blushing . On a more serious note we currently know using an extender is working for Malpi and me... and "just the two of us" is not enough to be sure it can be a solution for most users. So if you can test this on your microscope and let us know what's the results, it will be appreciated.

[daruosha]

On a more serious note we currently know using an extender is working for Malpi and me... and "just the two of us" is not enough to be sure it can be a solution for most users. So if you can test this on your microscope and let us know what's the results, it will be appreciated.

Mate, I tried Canon EOS 5D II with EF 2x II combination and and held it over the camera port by hand and it works just fine, much better than the crappy adaptor with builtin 2x optics I have. However I couldn't mount the EF 2x tele-converter on the microscope head, I need a hollow/helical tube to connect the T2 ring to the trinocular head and I'm pretty sure it's not available in my country. Need to talk to a machinist for a custom build

I added a EF 1.5x teleconverter and a custom build focus-able tube to my future purchase list.

[hip17702@gmail.com]

Thanks Daruosha. Can you tell us what type of microscope you use (upright, inverted...) and the objective's type (160mm, infinity...) in order to see if the solution can be used on a wide range of microscopes?

The current connection between the T2 ring M42x0.75 side to the camera port is an ABS 3D printed part. Works just fine and may be an alternative solution if you have access to a 3D printer?

[daruosha]

I tried it on both 160mm (finite) and 180mm (infinity) tube length microscopes and surprisingly it was very sharp corner to corner. I don't have 3D printer at home, but I think I can design the tube in a CAD software and use the commercial 3D printing services (it's quite cheap). I have zero experience with 3D printing (I know there's lots of tricks in design and printing the models) and I need some help I guess.

BTW all my microscopes are upright.

[Scarodactyl]

If you don't have access to 3d printing or local machining RAFcamera provides a great service, both with unusual off-the-shelf adapters and custom made ones for quite reasonable prices. Shipping isn't even that slow given things are coming from Russia.

[hip17702@gmail.com]

If you can tell me how the extension tube is attached to your microscope I may design the adapter in a 3D CAD software for you...

[daruosha]

If you can tell me how the extension tube is attached to your microscope I may design the adapter in a 3D CAD software for you...

Thanks for your generous offer, i'll take some pictures and measurements of the mount and send you a PM.