Small laptops have come a long way and many phones/tablets have USB-C.
Are there any compact digital microscopes that can be carried in a backpack and powerful enough for fungi spore IDing?
What would be the desired magnification range for spore analysis?
For looking at fungal spores for ID you want at least 400x total magnification, often 1000x and the ability to do size measurements. That 1000x needs to come with actual resolving power too rather than just âemptyâ magnification.
This difference between magnification and resolution is an important one. There are many retailers who will offer handheld digital magnifiers with claimed magnifications of 1000x, even 1600x or 2000x but those will do you exactly as much good as claiming you can get 1000x magnification from a regular camera by zooming in until only a handful of pixels remain.
Actually getting 1000x of usefully resolved magnification is hard enough that you start coming within sight of some quite hard physical limits. For one, getting there requires an oil-immersion objective since otherwise your numerical aperture is limited by the refractive index of air. The amount of light required is great so you will need proper focusing optics for it as well and probably want to use transmitted light. Following from this it will also require making reasonably good slide mounts since the working distance and focal depth of such an objective will by necessity be very short. Any shaking will be magnified enough to render the image unusable so the frame of the device should be rigid and strongly built.
By the time you add up all these criteria the most portable product that satisfies them is essentially a compact version of a classical compound scope with a camera attachment to make it âdigitalâ. At that point I would say you are much, much better off buying a regular compound microscope and transporting the mushroom to the scope rather than trying the reverse.
Edited after posting to clarify things, especially magnification without resolution
Thank you for the detailed write-up. Great info!
I am thinking out loud here.
So the desired magnification range is 400-1000x.
I read that the resolving power is determined by the numerical aperture (NA) of the objective. The higher the NA, the better the resolving power.
To relate to the NA, I looked up the specs of a $5K AmScope IN300. Data sheet shows NA of 0.60 for 40x. For a $2K AmScope T800, NA is 0.65 for 40x. This tells me I donât understand the âresolving powerâ yet. So I stopped this path.
Created a case study for a plausible scenario.
Crimson Waxcap (Hygrocybe punicea) vs. Cherry-Red Waxy Cap (Hygrocybe laetissima)
Assuming these online blurbs are true:
Factoid 1: Spores of Hygrocybe punicea are generally slightly larger and more broadly ellipsoid compared to the slightly smaller, more oblong spores of Hygrocybe laetissima.
Factoid 2: The spores of the Hygrocybe laetissima are about 8.0â10.2 microns by 4.8â6.0 microns.
Factoid 3: Hygrocybe punicea: 8-10.5 x 4-5.5 microns.
So we need a microscope that can show us distinguishable and ellipsoid shaped spores that are ~9” x 5”.
5” x 400= 2,000” (2mm)
5” x 1,000= 5,000” (5mm)
So we need a microscope that can depict tightly packed objects of ~2mm through a view finder.
For giggles, looked up the specs of a $23 AmScope UWT. Claims 50X-500X. Of course no NA spec, so no way to guess how it would show a 5” object. So I stopped this path too.
Letâs say we want to invent the next generation pocket microscope for the amateur scientists of the world.
Goal: Digitally achieve what a 1000x microscope (10X ocular and a 100X oil immersion lens) does.
20x50= 1000
25x40= 1000
Optical glass that can achieve 20X-40X are widely available. Digitally, 25X-50X should be doable. (For my photography gallery exhibits, I used a software called Genuine Fractals to massively enlarge small images. It did marvels. Donât know how it would perform for something like this.)
Will stop blabbing now. Need to do deeper research to define the boundary conditions and see if/how it can be achieved. I think this would be a perfect candidate for crowd funded R&D.
The reason for this difference is twofold. One is that you only need as much NA as the magnification you are working with demands, otherwise you are compromising on other things you donât need to compromise on. All serious microscopes from the cheapest to the most expensive will have a NA of about 0.6-0.65 for the objective meant to be used to deliver 400x magnification.
The second is that a lens is much more than just its numerical aperture. They can differ in quality and usefulness in many other ways such as planarity, dispersion, compatibility with advanced illumination techniques, etc. The same is true of the whole microscope system. The expensive scope youâve looked up has the ability to do epiflourescense- and phase-contrast microscopy, has a much greater working distance at the high magnifications, and is built to look from below the sample for looking at cell cultures; a much more specialised application. These differences are what you are paying for, not the numerical aperture of the lenses.
This suggest to me that you still havenât quite understood the problem of resolution and NA. Lenses that can achieve 20-40x magnification are widely available. Lenses at those magnifications with high enough NA to usefully resolve an image at 1000x total magnification are not. That is not to say they donât exist. You can buy 40x objectives with NA in the 1.15-1.40 range but they are specialty items for specialty applications that will usually set you back an amount of money that could buy several whole microscopes. They also all must be used with an immersion medium, since they have to for physics reasons. Were this distribution of the magnifications an easier way to get to 1000x than the usual 100*10 you would see regular scopes having 40x high-NA objectives and 25x eyepieces, but you donât.
Digital upscalers must rely on either self-similarity in the image (for classical approaches) or the assumption of structural similarity between the image and tendencies in a large reference set of images (for âAIâ approaches). They cannot faithfully add details that do not exist in the first place. In this case the meaningful features of the images are in the fine details around the resolution limit, so they wonât really help you.
Some existing threads that may be relevant:
https://forum.inaturalist.org/t/digital-microscope-for-students-chromebooks/38747
https://forum.inaturalist.org/t/new-years-resolution-low-toying-around-with-a-digital-microscope/47989
https://forum.inaturalist.org/t/the-smartphone-with-a-microscope-camera-arriving-at-last/21197/9
Digital upscaling methods have made phenomenal progress, much of which pushes the boundaries of believable logic. Sign up for the free trial of Genuine Fractals and enlarge a few images. Results are hard to believe. I will leave this here and focus on the rest.
40 x 5 microns= 200 microns (0.2mm)
So if a 40x pure lens based magnification is achieved, the rest âcouldâ be handled digitally. Automated Optical Inspection (AOI) machines used in chip design and PCB/CCA lines already perform this function, albeit for hundreds of thousands of dollars.
I think the crux is figuring out how to utilize a CMOS sensor within a compact and affordable package. It could be a stand-alone device or a screw-on attachment to a camera.
We already have CMOS sensors that have a remarkable resolution of 250 megapixels. The array is 19568x12588 pixels and each pixel size is 1.5 microns. This is monumental, because we already have a pixel size that is a third of a spore. Now add a bit of optical magnification, the scope comes within easy reach.
CMOS sensors with pixel sizes of 3-20 microns are well within budgets to create pocket devices.
I sent out a few inquires to some manufacturers. Will think more about this when I get the technical data and costs.
Like most other innovations, what will likely happen is once a small pocket microscope is developed, companies and universities will buy the first few hundred devices for thousands of dollars each. As soon as volume manufacturing parameters are fine tuned, then the public at large will have access to the same technology at a fraction of the initial cost.
I think such a device would be well received if the initial cost is kept $10K-$15K and finally stabilized around $200-$400.
It seems like you are not listening to what others have to say.
Digital upscaling has itâs place, that place is not in microscopy or in any image that needs to be accurate and trustable. Upscale âmake upâ the rest of the image and thus cannot be used as evidence of something.
âa scaling algorithm based on the use of PIFS When scaling up, Genuine Fractals exploits the self-similarity of an image to increase its size while preserving detailâ
That is the the key point, the program is only meant to be used to increase the resolution of an image and retain a certain quality when enlarged and viewed by human eyes.
Second, to answer your original question, there are no consumer grade âdigitalâ microscopes(most are actually simply webcam image sensors on a basic macro lens) that can view spores. If you want to view spores, get a used optical microscope from ebay, and set up a microscopy corner in your home or car like many mycologists do. Why do you feel a need to not do this?
Please learn about optics, and current microscopy practices, and consider the scope of your original post title.
Digital microscope for field carrying and fungi spore identification? This has been answered, and your theories and proposed solutions in your recent post are no longer on topic.
You are still thinking too much digital and not enough physics here. To utilize an imaging sensor you must project an image onto it, and those projection optics will themselves have a limited resolution below which detail will be blurred. This is caused by the physical nature of light, not the density of your detector. This can be quite unintuitive if your experience of resolution comes mostly from viewing and postprocessing raster images, but there comes a limit at which having a denser detector will not give you a more usefully detailed image, only one where each blur is spread over more pixels. When doing high-magnification microscopy this limit comes at you fast.
The first time I really encountered this was with my Nikon P950 coupled with a Raynox macro adapter at close to maximum zoom. Yes, the magnification is there, but the blurring/blooming from the density of the tiny sensor destroyed most of the detail at that zoom level.
If I take that same adapter and attach it to my Sony A6300 with a good macro lens on it, whole different story. The Sony sensor has four times the physical length of the Nikon and thereâs virtually no detail lost to sensor density.
It absolutely IS physics and a reminder that detail capture is subject to the unforgiving laws of physics, not digital detail density. Light has a physical dimension you just have to learn to live with.
I second this. Unfortunately as @dgwdoesthings pointed out:
There is nothing wrong with collecting the mushroom and then bringing it home to study, keeping it in something like a tackle box while youâre out in the field. Also, Iâd say that at-home microscopes in this case are much more convenient than a portable one. I imagine it getting dirty, and potentially getting damaged while youâre outside.
As for spore measuring, you need software and a computer alongside a microscope and camera. I imagine these extra supplies needing to be lugged around would make the convenience thing even more apparent.
Collecting the spores is also easier at home; especially in the case of spore-prints.
Doing everything after you have returned home with specimens is going to be more streamlined than having a scope out in the field, and buying a reliable optical microscope for the home would be in your best interest. I currently have an AmScope B120 microscope + included camera, and so far it has done everything Iâve needed it to do. Good luck on your microscopy journey :)