How To Add a Camera to a Microscope

A Few Words About Parfocality

Regardless of the method by which you mount a camera onto your scope, it is imperative that the camera be parfocal with the eyepieces (i.e., the eyepieces and the camera must both be in sharp focus at the same focus setting). If you find yourself having to significantly re-adjust focus when switching between using the eyepieces and the camera, this indicates improper geometry of the camera mount, and this is adversely affecting the quality of your camera’s imaging.

So, what’s actually happening when you refocus a microscope to correct a non-parfocal camera image? Let’s start with the obvious. When you change the focus setting from the position where the eyepieces are sharply focused to favor the camera, you’re moving the stage (either up or down) away from its proper position. In other words, you’re changing the working distance of the objective from where the optics were designed to operate to some other position.

This change in working distance affects the position within the optical tube where the objective projects its intermediate image. By mucking around with the plane of focus of the intermediate image, you are in-turn changing the plane of focus of the final image presented to the camera by the photo eyepiece. When you get the plane of focus of the final image to land on the image sensor within the camera, the image is seen as “in focus”.

You are in effect focusing the camera by forcing the optics to operate at a different optical tube length than that which they were designed to operate, and this introduces spherical aberration to the camera image. Do not degrade the performance of your microscope’s optics by using an improper camera mount! 

How Many Megapixels Do I Need?

“How many mega pixels do I need in a camera for my microscope?” I’m tempted to reply with “How many angels can dance on the head of a pin?”, but the answer to this question is much simpler than that. Let’s first look at the resolving power of a typical light microscope.

The Abbe equation for the diffraction limit (d = wavelength / 2 x NA) defines the minimum separation distance between two objects that can be resolved by a light microscope. If we use standard green light at 533nm wavelength and a numerical aperture of 1.4, Abbe’s equation predicts a limit to the resolving power, due to diffraction, of 0.19um.

This means that a light microscope operating with 533nm illumination at a numerical aperture of 1.40 (which is the practical upper end to the numerical aperture of light microscopes) should be able to resolve two items spaced at least 0.19um apart as separate objects. Objects spaced closer together than this are effectively hidden behind the diffraction barrier and will appear to the observer as a single object.

Now that we have determined the resolving power of a light microscope, all that remains is to find a camera with a suitable pixel density in the image sensor to resolve two objects spaced at the diffraction limit as separate objects, once a magnified image of these objects has been projected onto the image sensor by the microscope optics.

The specific pixel density required to resolve the projected image of these closely spaced objects depends on many factors, including objective magnification, photo eyepiece magnification, and reduction lens magnification (if a reduction lens is present), as well as the physical dimensions of the image sensor within the camera.

It turns out that if you apply the Nyquist sampling criterion to this question (i.e., 2.3 or more pixels in the sensor covering the projected image of two dots spaced at the diffraction limit), you will find that you don’t need very many megapixels at all, at least by modern standards, to resolve the two objects. Believe it or not, in just about every case a 5MP image sensor is more than enough to exceed the resolving power of the best light microscope.

A camera with higher pixel density will add nothing to the resulting images, and in fact, optical theory dictates that the higher the resolution for a given sensor size (i.e., the smaller the pixels), the lower the resulting signal/noise ratio of the sensor, thereby damning higher resolution cameras to produce noisier images than lower resolution cameras of the same sensor size.

If your camera has much higher pixel density than the required 5MP (such as a modern 48MP model, which is tremendous overkill for photomicrography!), you can achieve better imaging performance by setting the camera to utilize “pixel binning”, if this mode is available in your camera. Pixel binning groups multiple pixels together, to function as single, larger pixels with an improved Signal/Noise ratio, thereby capturing images with less noise.

The Importance of Image Sensor Size

[this section will be added soon]

Finding the Total Magnification of Digital Images

Everybody knows that the total magnification of a compound microscope is equal to the magnification of the objective lens multiplied by the magnification of the eyepieces. That’s easy, right? But what is the total magnification when using an objective lens and a photo eyepiece, to shoot images with an attached digital camera? Or how about when using an objective lens and a photo eyepiece with a supplemental adapter with an integral 0.3X reduction? The answer to these questions gets a bit more complicated. After all, the eyepieces have nothing to do with the resulting images in these scenarios.

When looking through the eyepieces, the images you see appear magnified by the mathematical product of the objective and eyepiece magnifications. But the total magnification of a digital image shot through an objective lens and a photo eyepiece depends not only on the magnification of these two bits of optics (objective and photo eyepiece), but also on the size of the image sensor in the camera, as well as the size of the image displayed on the computer monitor.

And this of course depends on the monitor screen size, the window size of the image-viewing software, and on the zoom factor of the image-viewing software, and so on. If you are instead viewing a hard copy of the digital image, the answer depends on the size of the image printed on the page, rather than the specifics of the monitor and image viewing software. I think you get the idea.

Go check out some YouTube microscopy videos. You’ll see plenty of videos identified as “100X” or “400X”, because the microscopist was using the 10x or 40X objective, along with 10x eyepieces when the video was shot! There are indeed a ton of variables to be considered here, but the eyepieces themselves do not affect the total magnification that the camera sees. Not even one tiny little bit!

So, why do we even care about the total magnification? The answer, of course, is so we can determine the size of the specimen and the size of the specimen details within the image. Take my advice and forget about trying to calculate the total magnification of the final image captured by the camera, and just go at it in the following way.

If you know the dimensions of the digital image sensor in your camera, and if you know the relative size of the specimen in the image, as compared with the size of the captured image, then you can easily determine the size of the specimen image that was projected onto on the image sensor via the objective lens, photo eyepiece, and supplemental reductions optics (if applicable).

For example, the APS-C image sensor in a Canon DSLR has dimensions of 22.3mm in width and 14.9mm in height. So, if the specimen fills half the width of the resulting image, then that means that the image of the specimen that was projected onto the sensor was approximately half the width of the sensor, or 11.5mm.

But how do you then go from there to determine the actual size of the specimen? Just divide the size of the projected image on the sensor by the total magnification of the projecting optics (including the objective lens, photo eyepiece, and reduction optics, if applicable), and that will give you the size of the specimen.

For example, if you’re using a 40X objective and a 2.5X photo eyepiece, then the total magnification of the image projected onto the image sensor is 100X (40 X 2.5 = 100). So, just take the calculated size of the image projected onto the sensor and divide it by the total magnification of the projecting optics, and you will have your original specimen size. In this case, that would be 11.5mm divided by 100, yielding a specimen size of 0.115mm, or 115µm.

If you wish to mark your images with magnification information (instead of using scale bars), be sure to not only include the total projection magnification, but also the image sensor type or physical dimensions, as knowledge of one of these is useless without knowledge of the other. The text “25X on APS-C”, for example, provides the viewer with all of the information needed to determine the specimen size, regardless of whether they are viewing the image on a large monitor, a cell phone screen, or on a printed page.

Note however that this method of image marking will not produce accurate size estimates if the photograph has been cropped or if the aspect ratio has been altered from that of the camera sensor. If the image has been cropped or if changes to the aspect ratio have been made, you should use scale bars instead of magnification markings.

Please don’t ever put “400X” on your images, simply because you used the 40X objective and happened to have 10X eyepieces in the scope when you shot the photo. Imagine if you were to shoot an image with 10X eyepieces in the scope, using the 40X objective, then remove the 10X eyepieces and sneak in a pair of 20X eyepieces and quickly shoot another image. Should the first image be marked “400X”, while the second image (which is identical to the first image) gets an “800X” mark?

This example should illustrate how wrong it is to mark that magnification number on your captured images. Keep an eye out as you read articles and watch YouTube videos on microscopy. This error is embarrassingly common.