PROJECT: ADDING A MODERN DSLR TO AN POLYVAR MICROSCOPE

As I do more microscopy, I am becoming increasingly motivated to replace my Coolpix 990 - 4500 series cameras with a more modern digital SLR.  Specifically, I'd like to add a camera that meets the following criteria:

1) Capable of sending live images directly to a computer for preview purposes.

2) Capable of capturing live images (shutter open), thereby eliminating the detrimental effects of shutter vibration.

3) Free of the "ring artifact" that plagues the Coolpix 900 - 4500 series cameras (the ring artifact seems to be amplified by epi-illumination of metals).

After a lot of internet searching, I learned that relatively few consumer-grade cameras meet these requirements.  Helpful guidance has come from Bobby Martin of Martin Microscope Company.  He recommended Canon EOS cameras that have "LiveView" capability and can be tethered to a computer and controlled using the Canon EOS utility.  This camera-software combination looks very impressive (see a Canon EOS Utility demonstration).

Bobby Martin favors the Canon EOS Ti3 (18 MP and articulating screen).  I happened to find the Ti3 kit (including the standard lens) listed on eBay's "Daily Deal" for $650 with free shipping.  I decided to buy it with the expectation of eventually selling my Coolpix cameras (three of them).  In spite of their age, Coolpix cameras have a strong following among the microscopy and digi-scoping communities.

Options for Mounting

I seem to have four options for attaching the Canon T3 to my Reichert Polyvar SC microscope:

Option 1: Attach the camera directly to one of my oculars.  This is the set-up I currently use with my Coolpix cameras (see the photo below), but this approach has the obvious disadvantage of limiting direct observing to only one ocular.

Another limitation of the Coolpix is that my field of view is very limited compared to the FOV I enjoy when looking through the binocular eyepieces.  The diagram below illustrates the field of view for my current setup, as seen through a 20x objective and 10x widefield eyepiece.  My FOV looking through the ocular is 1200 microns across (represented by the circle).  In contrast, my FOV seen through my Coolpix 4500 mechanically attached to the eyepiece is 420 x 280 microns (red box).  The camera was zoomed in just far enough (zoom level = 4.0) to eliminate vignetting.

According to Bobby Martin's website, a Canon EOS camera coupled to his MM-SLR yields a field of view of ~ 420 x 280 microns (for a 20x objective).  Therefore, it seems that converting to a DSLR is unlikely to reduce my field of view compared to my present Coolpix setup.

Option 2: Attach the camera directly to a horizontal photoport.  My microscope has two identical horizontal photoports: one on the left side of the microscope and one on the right side.  One of these photoports can be seen at the top of the microscope in the photo above, and the photoport is shown in more detail below.  The diameter is approximately 39 mm, and the depth is approximately 45 mm.  I can see a focused image when I look directly down the empty photoport.  In contrast to the round field of view I see through the binocular eyepieces, the phototubes yield rectangular field of views.  The FOV is greatly reduced compared to the FOV visible through the binocular eyepieces.

My microscope came with a "TV Adapter" lens with C-mount (shown below) that inserts nicely into the side photoport (Part A = 33mm; Part B = 38mm; Part C = 40mm).

The TV adapter contains multiple lenses, and the left side has a 25mm C-mount.  When I place this tube in the side photoports and place my microscope eyepiece (10x widefield) on top of the C-mount, I can see a sharp image that is parfocal with the binocular eyepiece.  The blue box (above) shows the field of view (380 x 250 microns) obtained with this configuration.  In this case, I needed to zoom the Coolpix to zoom level = 4.5 to avoid vignetting.

Option 3:  My last option is to attach the camera to a vertical photoport on top of the microscope. 

This hole was originally used for a Polaroid camera.  It is 39 mm in diameter (like the side photoports), but is more shallow (15 mm deep).  From a weight distribution point of view, this would be a good place to attach the camera.  I can obtain a image here using the same method as described for option 2 (combining the TV adapter + eyepiece).  The resulting image is identical to the image obtained at the side photoports.

Update: Attaching the Camera Without Intermediate Lenses

With help from Polyvar SC owner Peter Webber, I explored a fourth method of coupling a camera to the Polyvar.  In this approach, the top photoport unit is completely removed to expose the beam-splitter.

When I hold the Canon camera (without lens) 50-51 millimeters above the beam-splitter, I obtain spectacular images such as the three shown below.  These images were created by projecting directly onto the sensor, with no intermediate lenses at all.  This result is a ultra-wide field of view, approaching the FOV observed through the eyepieces (and significantly larger than the FOV seen through the top photoports):

Metamorphic rock, as viewed in transmitted light.

Meteorite chondrule, as viewed in transmitted light.

Meteorite metal surrounded by silicate, as viewed in reflected light.

It seems that the Polyvar's intermediate image is similar in size to the Canon sensor.  The diameter of the Polyvar intermediate image, measured above the beam-splitter using a ruler and frosted glass plate, is ~25-30 mm.  The Canon T3i sensor is reported as 22.3 x 14.9 cm (26.8 mm diagonal).  I don't get any vignetting, showing that the intermediate image is larger than the sensor.

One Polyvar expert, a former Reichert employee, warned me against removing the top photoport unit due to the complex optics.  However, I find that the image is surprisingly nice, and the FOV is much larger than the FOV I get using the camera module.  The edges of the image are a bit blurred, but this is likely caused by my crude camera mount (my hand).

Peter Webber pointed out the chromatic abberations (blue fringes) are likely due to the fact that this FOV gets too close to the edge of the objective's useful image.  It may be preferable, therefore, to stick with the original top port (with reduced FOV) as recommended by the former Reichert employee.