sensor size / aspect-ratio / fov

Here you can discuss everything related to taking light micrographs and videos.
Post Reply
Message
Author
mete
Posts: 114
Joined: Thu Aug 26, 2021 8:31 am
Location: Switzerland

sensor size / aspect-ratio / fov

#1 Post by mete » Tue Sep 14, 2021 7:32 pm

I am trying to decide on my optimum solution for camera (and I plan to not use any optics for camera connection, so it will be 1x). I wonder if there is a best practice on aspect-ratio and/or FoV coverage (FoV I am targetting is 23mm, % coverage below is based on that). It seems there are a few options:

- square image (e.g. APS-C or MFT sensors at 1:1) all inside FoV, the best is just touching to FoV, covering ~60% of FoV. but this will look small on displays, since all displays are 16:9.
- 16:9 image (e.g. MFT 16:9) all inside FoV. this is best for displays, but its coverage is less than 50% of FoV I think.
- a little bigger 16:9 image (e.g. APS-C 16:9), with some small vignetting, but covering ~75% of FoV.
- full frame 16:9 image with lots of vignetting but covering ~100% of FoV.

(all these sensors have pixel pitch between 3um-5um, so I guess there is no difference in that respect)

Assuming the image quality is the best in the center, the optimum might be the maximum 1:1 inside FoV. But for practicality, maximum 16:9 inside FoV might be chosen because of displays. On the other hand, to not lose too much of FoV and still have 16:9, an APS-C 16:9 sounds better. If the image quality is not that different in the center and on the corners, then why not to lose any area and use a full frame. I started to find it very odd that all optics are circular, but all sensors are rectangular (and not 16:9) and all displays are 16:9. What do you think ? What do you use or what do you think it is the best to use ?

I should add I am using camera not only for taking pictures but also for display connection so display connection part is important for me. However, I can sacrifice 16:9 if needed. At the moment, I am using APS-C at native or 16:9 on my stereo and I am waiting for my compound, so my question is not only for stereo.

Mete

User avatar
micro
Posts: 203
Joined: Mon Jul 22, 2019 6:11 am

Re: sensor size / aspect-ratio / fov

#2 Post by micro » Wed Sep 15, 2021 12:14 am

I don't know about the mathematical fov stuff but for taking pictures and videos of microbes I prefer micro 4/3 cameras because of the zoom effect they have. People say full frame is technically more detailed but the zoom factor with micro 4/3 and CMOS sensors are useful. The Canadian nature photographer has articles on it. What are you planning to look at with the microscopes?

mete
Posts: 114
Joined: Thu Aug 26, 2021 8:31 am
Location: Switzerland

Re: sensor size / aspect-ratio / fov

#3 Post by mete » Wed Sep 15, 2021 5:01 am

I am using the stereo microscope with practically anything up to 10x. I am planning to look at plant cross sections and protozoa on the compound microscope at first.

apochronaut
Posts: 6272
Joined: Fri May 15, 2015 12:15 am

Re: sensor size / aspect-ratio / fov

#4 Post by apochronaut » Thu Sep 16, 2021 1:34 am

I guess you have done some research into this. I'm wondering though, why you assume that because you do not have a relay optic that the magnification will be 1X.
That factor will vary depending on whether the system is fixed tube or infinity and what distance you place the camera sensor at. You also may not be factoring in that most microscope objectives benefit from corrective optics downstream and without the benefit of such, it is often necessary to crop a considerable portion of the image out in order to remove poorly corrected peripheral portions of the image.

mete
Posts: 114
Joined: Thu Aug 26, 2021 8:31 am
Location: Switzerland

Re: sensor size / aspect-ratio / fov

#5 Post by mete » Thu Sep 16, 2021 6:39 am

Actually I said 1x only because such optics always have magnification numbers on them naturally, and the ones without optics, so only adapting the microscope thread to T2 are 1x (At least it is Zeiss terminology I think, I mean particularly the 60N-T2 1x adapter, I dont know if 60N-C 1x adapter has any optics). So yes you are right there can be sth inside the tube but what I mean is additional element that is put on the camera is 1x. I am not concerned with this magnification factor actually, I am just not happy with additional optics if not 100% needed and also associated costs of them.

I am only talking about infinity system and still not sure about if camera connection optics have benefits (in Zeiss), I am inclined to believe that any explicitly designed correction is done inside the microscope (without eyepiece and camera optics, so within the objective and tube lens), but not sure how this can be verified.

It is very different but I didnt observe a benefit and actually I didnt like the outcome on stereo with 2.5x optics for full frame. However I didnt do a quantitative test and compound microscope is different so maybe I had wrong impression and I should wait and see first.

When you say considerable portion of the image, what exactly do you mean ? I ask because if I am not wrong, even using for example Zeiss camera on Zeiss scope (as in the official system overview pages) has cropping I guess to eliminate vignetting but I wonder if this is also done on purpose to eliminate very edges of the image because of limited correction there. If that is so, to answer my question, using an MFT sensor makes sense.

apochronaut
Posts: 6272
Joined: Fri May 15, 2015 12:15 am

Re: sensor size / aspect-ratio / fov

#6 Post by apochronaut » Thu Sep 16, 2021 12:10 pm

With plan optics it is possible to obtain a camera image that recreates the entire view of the microscope field faithfully. This can be either a circular image inside the camera frame, capturing the entire viewed field or various aspect ratios of a rectangular image, which reach the extremeties of the field in the corners only. Uncorrected distortions will occur near to the periphery. Microscope optical systems vary in their design principals depending on the manufacturer, such as the degree of correction of the objective and thus the degree of further correction taking place in the telon lens or eyepieces, since all manufacturers of note end up with very similar results coming out of the eyepiece.
My understanding is that Zeiss engineers have opted for further corrections to be incorporated into the telan lens and eyepiece as part of their total optical system, thus it seems a likely necessity to replicate those corrections in the photo tube as well. This is outside of a question of magnification but occurs within a magnifying optic but usually one of much lower magnification than a viewing eyepiece.
Most modern infinity plan objectives are quite well corrected internally for various aberrations and distortions with chromatic aberration and flatness of field corrections often being completed outside the objective. I would say , one of those at least, always. If the eyepiece plays no role in these further corrections it is what would be called a neutral eyepiece. Since neutral eyepieces confer only a magnification factor upon the objective/telan lens combination, they show no ca near to their aperture border. Correcting eyepieces will show blue and compensating( the inverse of correcting) eyepieces will show yellow to orange. Your photo relay optic needs to be able to replicate this condition. Where a neutral or near neutral condition is required, some very large low power relay optics, those from photo copy systems for instance will satisfy this condition due to their very large aperture and freedom of aberration/distortion.
You can determine the ca correcting condition of your optical system, i.e. the eyepiece, merely by looking through it. In some systems, an equivalent optic, even an eyepiece is fully adequate as a relay optic, especially with APS-C sensors.
Many modern eyepieces used in infinity systems are also plan compensating. Two eyepieces from very similar systems may in all ways be the same, except when it comes to curvature of field. Where one will provide a perfectly flat field the other will provide a similar image except have a small degree of either barrel or pincushion distortion. It may be slight and perfectly acceptable for visual use ; the brain can usually accomodate or overlook such stuff but the sensor cannot, so if field flattening is in the cards for your downstream optics, so must it be in the relay optic.

mete
Posts: 114
Joined: Thu Aug 26, 2021 8:31 am
Location: Switzerland

Re: sensor size / aspect-ratio / fov

#7 Post by mete » Sat Sep 18, 2021 2:16 pm

I think that (correction in tube and eyepiece) is correct for finite designs, but in infinite ones it is mostly/always? done in tube lens only, at least that is what I get from all what I read. I am still waiting for my compound microscope but it has the same eyepieces if I am not wrong with the stereo, and I dont see any CA. I am not sure about field curvature correction but I will test this in near future.
For the issues in extremities, I think it makes sense to use MFT sensor. It also makes it similar to actual system recommendations of the vendor.

Post Reply