Is It Time To Re-invent The Optical Microscope?
by Mol Smith (UK)

The optical microscope has been with us basically unchanged for at least 400 years and a bit. In the last twenty years, extraordinary advancements have been achieved in the field of optical imaging. For the enthusiast, the most interesting area of improvement lies within the realm of photography. To be over-simple: digital photography.

And, indeed many microscopists now exploit that technology to sample ever clearer and detailed images from their studies at the microscope. This is accomplished by fixing their cameras to the eyepiece tube of a microscope through one of a variety of adaptors suited to the various camera bodies.

The Restrictions And Issues Of Optical Microscopy

All enthusiasts (or amateur) microscopists quickly come to realise many limitations. For one, light itself and its wavelength restricts optical resolution to a 'useful' magnification maximum of around 1600x, beyond which - increasing magnification fails to resolve any finer details. Even at these high levels of magnification, oil immersion lenses and techniques must be used to eliminate the distorting effects of air between the objective lens and the subject being examined.

Depth of field
A second issue is the nature of something called 'depth-of-field'. In simple terms, it is the depth into the subject which is clearly focused. The narrower the aperture (the tiny hole through which light passes), the greater depth of field is obtained. But the trade-off is the narrower the aperture, the more intense light is needed to illuminate the subject sufficiently to capture a recorded image.

With still specimens, good images can be obtained using image-stacking software which determines a composite image from many photos captured at various depths into the specimen. But with moving specimens, tiny organisms living in ponds for example, this is not a solution.

Most things we study under the microscope don't really appreciate bright hot light. in the past, the filament bulbs used in the illumination system produced a lot of heat. Fortunately, new LED lighting is improving rapidly and is now able to produce very white dimmable white light with no heat at all.

So, we can do something about better illumination. We can work solving depth of field problems using image stacking software (with non-moving subjects). But we can't change the laws of physics (alas), and that means we are stuck with a maximum resolving magnification of around 1600x.

This leaves a single limitation which hampers study, despite new technology being available to solve the problem!

Solution To Constant Focusing
Advances in digital photography and video technology could be directly utilised to make a brand new kind of optical microscope which would greatly overcome the issues of trying to keep a moving specimen constantly in focus,
I believe.

As a film maker utilising prosumer and low budget cameras and equipment to make documentary and feature films, I am aware of similar problems existing with making movies. When an actor moves backwards or forwards to the camera, the operator must manually shift focus, itself a judgemental action, which often proves inaccurate, leaving the actor out of focus. Various gadgets and contraptions have been produced to help improve manual focus adjustments (focus tracking) but none of them are really very good. DSLR digital cameras which do have auto-focus have been unable to smoothly track focus on moving subjects. They either perform too slowly or are prone to 'hunting' an effect where they lose focus and constantly 'hunt' to achieve focus again before settling into the correct one. No good for film makers, I'm afraid.

When a new Canon camera was released last year claiming to eradicate focus tracking issues, I was delighted. Now having tested the camera, I realise similar methods and technological application could be used to constantly track focus on a microscope. But... it would require a purpose-built and redesigned instrument.

The Canon EOS 70D
The EOS 70D features a newly-designed Canon APS-C CMOS sensor with 20.2 million pixels. All effective pixels on the surface of the EOS 70D’s sensor consist of two individual photodiodes, which are read separately for phase-detection autofocus and together for imaging. Dual Pixel CMOS AF is possible over 80% of the width and height of the Live View frame and is highly accurate without compromising on image quality.

Unlike other methods of sensor-based autofocus, which allocate either autofocus or imaging functions to pixels on the sensor, the pixels on the EOS 70D’s CMOS sensor can be used as imaging pixels and phase detection AF pixels. As a result, no additional imaging processing is required around pixels dedicated for AF, which ensures both quick acquisition of focus and maximum image quality. To perform phase detection on the image plane the EOS 70D reads left and right photodiodes independently and then calculates the phase-difference of the two parallax images.

By using this phase-difference AF technique the EOS 70D predicts the position the lens is in focus and moves it. By contrast, the more conventionally used contrast AF method looks for the focal position while moving the lens back and forth, so AF speed ends up being slower and not as smooth. With face and object tracking engaged, Movie Servo AF follows a chosen subject as it moves, or when you recompose a shot. Alternatively, just select different focus areas by simply by tapping the touch-screen when recording.

A New Microscope

What is required then is an optical microscope where the rack and pinion focusing method of the objective lens is replaced by a servo-driven system, similar to that within the Canon SMS lenses. A CMOS sensor of the type used in the Canon 70D along with a suitable processor could then be used to drive the 'intelligence' of the focus system. Of course it would be mean there would no longer be an eyepiece (not without more sophisticated mirror systems), and the subject would be viewed on a electronic screen.

Macro Video
I had little time to test the Canon 70D in Macro work, but I did a quick and dirty test. I never used a Tripod and it was a breezy day. I attached a macro lens to the camera and did a quick shoot. It's jerky sure, but I think you'll see the camera is doing a great job of focusing and tracking the fly on this violently moving leaf. See below.


Quick and Dirty Macro Video using Face-tracking - auto-focus on the Canon 70D