Microscope - camera setup

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

Moderators: rjlittlefield, ChrisR, Chris S., Pau

FTL
Posts: 11
Joined: Fri May 15, 2020 1:29 pm

Microscope - camera setup

Post by FTL »

Hi,

My goal is to take 4K videos of spores at the size of 30 - 100 microns through a microscope.

My question is - what is the most quality setup for this goal? Is it a microscopic camera attached to the eyepiece? Or is it a mirrorless / DSLR camera connected to a trinocular binocular / stereo microscope?

What do I need to know about image crop, if I go for large sensor mirrorless / DSLR cameras (not larger than APS - C)?

Also, I once inserted my mirrorless Fuji XT-20 into my telescope with an adapter, and the images turned out blurry, after achieving maximum focus. Even my smartphone took sharper images. Do anyone know what the problem is? I don't want this to happen at the microscope setup aswell.

And what about image magnification? Can I know what the magnification and field of view would look like?

Is there anything else I should know?

Thank you!

Ichthyophthirius
Posts: 1152
Joined: Thu Mar 07, 2013 5:24 am

Post by Ichthyophthirius »

Hello,

Welcome to the Forum. You questions are valid :wink: but there are too many variables for a satisfying answer. Especially for the "most quality" bit.

There is much more skill involved in preparing the sample and setting up the microscope, than in taking a video of what you see. The microscopy is the limiting factor! Once you have mastered the microscopy (takes a few weeks to months), taking the video will be easy (a smartphone will be enough for a good microscopist).

Have you done microscopy of the spores before?

How you attach the camera will depend mainly on the microscope and microscope optics you have available, to some degree on camera you use, and finally on your budget.

You can use anything from a $20 smartpone adapter to a copy stand for a mounted camera (a near-ideal solution for almost any older microscope): https://www.photomacrography.net/forum/ ... mall_1.jpg

Let us know which microscope and camera you want to use and I'm sure we can find an answer.

Regards, Ichty

FTL
Posts: 11
Joined: Fri May 15, 2020 1:29 pm

Post by FTL »

Thank you very much for the reply!

Yes, I've done microscopy of spores. I just want to upgrade my current equipment, because my setup now is not giving me satisfying results.

As for which microscope and camera I want to use - this is exactly what I'm trying to find out :)

Currently my setup includes a 720p video microscope camera alone, without any additional optics - https://www.aliexpress.com/item/32630914138.html

I also tried using different cameras, and attaching them to different microscopes, without good results. The 720p alone gave me the best results, but they are really far from what I want.

As I said, I thought about some options:

1. Staying with the current setup, with just upgrading to a 4K video camera (do you know of a good 4K microscope camera that has really good quality?)

2. Attaching a mirrorless or DSLR to a trinocular stereo microscope or to a compound microscope.

Do you know from experience if a mirrorless / DSLR will give better results than a microscope camera?

Thank you!

JKT
Posts: 420
Joined: Fri Oct 28, 2011 9:29 am
Location: Finland
Contact:

Post by JKT »

I suspect what you are asking can't be done.

Let's assume object width of .35 mm, so your largest spore would be about a third of the image width. That would take 100x objective on APSC-sensor.

If you had NA=1 (which you can't have without immersion), your horizontal resolution would be roughly 1000 px. Any APS-C camera will have much more than that, so the camera will not limit your final image quality.

At the same time your dof will be about 0.67 um. So only a very narrow slice of your spore will be sharp ... and that is sharp in respect to the low resolution. Usually this would be handled by stacking, but with video that can't be done. So I suspect the resulting image would not be what you want.

Dropping NA to 0.75 would double your dof, but at the same time the horizontal resolution would drop to about 740px.

So it would seem conventional microscopy doesn't really give what you likely want ... at least as long as my assumptions about your intensions and requirements hold.

P.S. The numbers I've given are from my spreadsheet, but they should be in the ballpark.

FTL
Posts: 11
Joined: Fri May 15, 2020 1:29 pm

Post by FTL »

Thanks for the reply JKT. But I'll be happier if you could explain this in words that I will understand :)

I'll try to see if I understood it correctly:

APS-C sensor will give a very narrow image in a compound microscope. Is that right?
What about slightly smaller sensor size? How will this change the image?

And what about attaching the APS-C to a stereo trinocular microscope? which gives much lesser magnification than a compound. Will I get the same problems?

(Stereo trinocular microscope like this one https://www.telescopeadapters.com/vmag2-diagram.jpg)

JKT
Posts: 420
Joined: Fri Oct 28, 2011 9:29 am
Location: Finland
Contact:

Post by JKT »

The problem is in the resolution you can reach with the microscope objective. For that theoretical 100x NA=1.00 objective, the minimum detail size you can possibly see is about 1/3 of um. There simply is no more information available. Whatever you do on the sensor side won't change that for better.

IIRC the best actual values for non-immersion objectives are around 0.9. In order to improve the available resolution you'd have to use water or oil immersion. Unfortunately that would then decrease your dof even more.


What you really need is considerably shorter wavelength than light. But then the required sensor is no longer a camera.

rjlittlefield
Site Admin
Posts: 23562
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

FTL, let's go clear back to your original post and make sure we understand what you want.

You wrote "My goal is to take 4K videos of spores at the size of 30 - 100 microns through a microscope."

How large do you want each of those spores to appear in your video? That is, how many pixels high and wide for each spore?

--Rik

Scarodactyl
Posts: 1617
Joined: Sat Apr 14, 2018 10:26 am

Post by Scarodactyl »

What field of view do you want?
With a (currently very cheap!) Nikon planapo 20x/0.76 coupled to a 200mm tube lens onto aps-c you can get a very nice 2mmish FoV with perfect color correction and good resolution for the FoV (assuming you're using slides with cover slips.)
A stereo is definitely not the right choice here. Outside of some specialty setups their NAs tend to cap off at around 0.2 with a 2x objective or auxiliary lens.

FTL
Posts: 11
Joined: Fri May 15, 2020 1:29 pm

Post by FTL »

rjlittlefield, you are right, I didn't note all the details.

I want a field of view of approximately 1mm high and 1mm wide, which lets me to see atleast 30 spores at a time (answering to your question aswell, Scarodactyl). I don't know how to translate it to pixels.


Scarodactyl - "2mmish FOV", can you please describe this so I can understand the scale?

JKT
Posts: 420
Joined: Fri Oct 28, 2011 9:29 am
Location: Finland
Contact:

Post by JKT »

I see. If you only want roughly 1 mm x 1 mm FoV, the mentioned 20x objective gives 1.1 mm x 0.75mm on APS-C sensor when used with nominal length (f = 200 mm) tube lens. The horizontal resolution you could get would be around 2250 px.

The problem of DoF remains, though. It is about 1.3 um. The way to increase that would be to use objective with lower NA, but that also drops the resolution. With NA=0.4 your DoF would be about 4 um, but the resolution only 1290 px.


You can slightly modify the parameters by using different tube lens, but the the basic issue between DoF and resolution remain. You didn't explicitly say what kind of DoF you wanted - a slice are full depth of the spores. The latter won't work ... unless they are extremely flat.

Edit: "2mmish FOV" = roughly 2 mm FoV. You would get that by using around 100 mm tube lens (or FF sensor with 150mm tube), but the corner resolution might suffer with that objective.

metebalci
Posts: 29
Joined: Fri May 01, 2020 1:32 pm
Location: Switzerland
Contact:

Post by metebalci »

FTL wrote: And what about attaching the APS-C to a stereo trinocular microscope? which gives much lesser magnification than a compound. Will I get the same problems?

(Stereo trinocular microscope like this one https://www.telescopeadapters.com/vmag2-diagram.jpg)
Not an expert here, but I have -I guess- a mid-level stereo microscope (0.63x-5x and up to 10x with front optics). After some try-errors, I think the best camera connection is using the photo tube with trinocular and using an MFT sensor without a camera adapter (no optics) (and probably using pixel shifting). However, I almost never get a nice image on the stereo microscope at higher magnifications (higher in its range) because of dof, and since the focus block is not automated or has very fine adjustments, it is impossible to do proper fine focus stacking. So if you are around 20x, stereo microscope is a no-go for various reasons if you don't probably get a very expensive setup.

FTL
Posts: 11
Joined: Fri May 15, 2020 1:29 pm

Post by FTL »

Do you mean that no matter how good the camera is, the resulting image with using NA=0.4 objective lens will be worse than NA=0.7 objective?
Just trying to understand the basics first :)

About the DoF (Depth of Field?) - I prefer a zoomed out video which will make the spores be small but still visible, and by zooming out - I get more DoF, right? This is how I want it to be looked like:

Image

I once used a nice telescope simulator where you can insert the telescope's specs and it shows you the resulting images you will get with these specs. Is there something similar with microscopes?
I just want to see what I'll get more or less before I buy my next setup.

rjlittlefield
Site Admin
Posts: 23562
Joined: Tue Aug 01, 2006 8:34 am
Location: Richland, Washington State, USA
Contact:

Post by rjlittlefield »

JKT's numbers look way too pessimistic to me.

There's a handy calculator at https://www.microscopyu.com/tutorials/m ... resolution , titled "Matching Camera to Microscope Resolution".

When adjusted for a field width of 1.07 mm at 20X NA 0.40, the calculator returns a number of 3120 pixels wide. Here is a screen grab of that setup.

Image

The calculator does not support sensors as large as APS-C, so I've had to compensate for that limitation by specifying a 0.60X video coupler and a sensor size of 12.8 mm width. The net magnification on sensor is then 20*0.60 = 12.0X, and field width is 12.8 mm / 12.0 ~= 1.07 mm.

The same result can be obtained solely by calculation, independent of sensor size, by noting that the cutoff frequency (cycles/mm), on subject, for any objective, can be estimated as (2*NA)/lambda [REF]. Minimum sampling is Nyquist at 2 pixels per cycle, so at NA 0.40 and green light lambda = 0.00055mm, the required number of pixels is 3103/mm on subject. Multiply by 1.07 mm field width to get the number reported by Nikon's calculator. (There's a slight difference due to rounding.)

If we also note that FTL said he wants "1mm high and 1mm wide", but shows us an image that is long and skinny, then we might guess that an area of 1 square mm would be fine. Noting that 4K video has a 16:9 aspect ratio, that would make the desired field size be 1.33 x 0.75 mm, raising the horizontal pixel count for NA 0.40 to be about 3870.

So, it looks to me that there's really a good match between 20X NA 0.40 and the 4K video that FTL has said he wants.

I agree that the nominal DOF for NA 0.40 is only 3.4 microns (=lambda/NA^2), but I wonder if there's actually any need to focus both the perimeter and the front face of each spore at the same time.

--Rik

Edit: to correct the 16:9 calculation.

Scarodactyl
Posts: 1617
Joined: Sat Apr 14, 2018 10:26 am

Post by Scarodactyl »

FTL wrote: Scarodactyl - "2mmish FOV", can you please describe this so I can understand the scale?
Sorry that was a typo, I did mean 1mmish. Basically a view that's about 1mm across (as JKT correctly pointed out). I wouldn't try for a wider field of view, it's already pushing it with that field of view if I remember right. You won't find a better objective for less though, they're an insane deal.

Note stacking is possible with a stereo, even without fine focus. It's just more of a pain to do. I used to do it on my sz7 at full zoom (sometomes eith my 2x aux on), focusing with one hand and taking photos through thr eyepiece with my phone in the other. That was.......not the best way to do that but it is possible. Not for this sort of work though--the highest resolutions I know of in stereos are around na of 0.35 (not counting attaching compound objectives to one) if I remember right, and you'd have to really shell out for one of those.

metebalci
Posts: 29
Joined: Fri May 01, 2020 1:32 pm
Location: Switzerland
Contact:

Post by metebalci »

Rik, thanks for the detailed calculations, I was to ask where the px number is coming from. Since this px number is assumed to be equal to nyquist freq., it should be taken as a lower bound, is that right ?

Mete

Post Reply Previous topicNext topic