Updated 26 October 2017
Use the browser "back control" to return here.
My Leitz Dialux Polarising Microscope is a research grade microscope with 5 objectives. The microscope is fitted with white LED illumination. Recently I purchased a 1.3 MP microscope camera from Omax. The choice of a microscope camera is difficult as the optical limitations of any microscope and display has a major effect on the ultimate resolution. Resolution is a major issue with microphotography, and there are real physical limits to the image quality obtained.
An interesting method for contrast improvement with polarised illumination of samples is described, with examples. An additional plastic compensating plate is used to improve image contrast. It does this by setting the background colour exactly to a narrow magenta region between first-order yellow and second-order blue. This is equivalent to slightly adjusting the thickness of a full-wave gypsum "sensitive-tint" compensator plate until the contrast is optimized.
Dark-field and circular-oblique illumination techniques are also described. They are both used to improve contrast for difficult subjects like diatoms.
There are brief descriptions of various macro-cameras, stereo-microscopes and a modified USB-microscope.
I revise these notes periodically to correct mistakes.
Leitz Dialux Polarising Microscope
My Leitz Polarising Microscope
My Leitz Dialux Polarising Microscope is in good condition, with a performance matching that of a modern microscope. The microscope can carry 5 objectives on the turret. It has a trinocular head which allows a camera to be fitted. The rotating stage is supplemented with a precision x-y adjustable slide holder. Two condensers are available with full adjustments. Correct condenser adjustment delivers excellent viewing results with this microscope. The adjustments, which are described in many textbooks, confine the light to nearly fill the lens apertures without touching adjacent surfaces. This gives the best resolution and good contrast.
First I changed the original tungsten light-bulb source for a 1 watt LED. The LED runs much cooler and the intensity can be varied without changing the colour balance. The LED was mounted in an old light-bulb base in series with with a 5 watt 33 ohm resistor. The new "light-bulb" now worked only in a single orientation in the socket. A diode was added in series to protect the LED against reversed voltages. A simple 1.4 to 15 volt power supply was built based on an LM317 voltage regulator. The original lamp power supply could also be used, but I did not have one. Switch-mode LED power supply modules may produce interference bands in the image if a digital camera is used. In addition the light intensity will be difficult to vary.
The LED was set at the same position as the old filament. The LED lens surface was lightly sanded to widen the angle of illumination. A typical narrow angle LED will result in uneven illumination at the edge of the field of view. My 480 lumen Cree torch also works well, without producing interference bands. LEDs have improved in efficiency so lower rated devices should be useable just with rechargeable batteries.
I now use a Cree XP-E LED glass bulb with a BA15D base which is a suitable replacement for the original tungsten bulb. The only problem with this lamp is that it is not easily dimmable, as it has a built-in stabilised power supply. Fortunately the camera I now use works with the wide range of light levels from different setups.
When converting a microscope to LED illumination excess brightness can be a problem. LEDs are not lasers, but they can be very bright. Coloured or diffusing filters may help. A means of dimming is helpful. If dimming is not possible then a white LED power rating to replace a 15 watt incandescent lamp should be less than 1 watt, and preferably much less when using modern LEDs.
For example I have used a 2.7 volt, 20 mA, white LED with a 12 volt power supply. A (12 volt - 2.7 volt)/0.020 amp = 470 ohm resistor was needed in series to keep the current to less than 20 mA. I found that for general use this provided more than sufficient light. I have used bare LEDs recycled from old light bulbs, although some light bulbs package multiple LEDs with higher voltage ratings (4 x 2.7 = 11 volts). This will alter the required series resistance to about 56 ohms. The dimming voltage range also changes from 12-2.5 volts to 12-10 volts.
Choosing a Microscope Camera
For the usual range of microscope objectives the observed horizontal resolution, expressed as pixels, ranges from about 1800 pixels at 3.5x to 1100 pixels at 40x. In oil the observed resolution is doubled to a maximum of about 2000 pixels at 40x, falling to 800 pixels at 100x. For more see Numerical aperture and resolution.
A good camera covering much of the field of view should therefore have a horizontal resolution of 1000 to 2000 pixels. A 5MP camera would exceed this specification and a 2MP camera would be sufficient for work in air. A 3MP camera meets this specification and would perform well as the sample is moved into position and for focussing. A USB3.0 interface would allow for faster image updates, although USB2.0 is usually adequate - provided the camera resolution is not too high. For current low-cost camera models the speed is mostly limited by the camera electronics, rather than the USB interface. A 21 inch computer monitor is typically 2MP and not all of that area would be used in practice.
Field Curvature and Resolution
Another point against excessive camera resolution is that, although the image is sharp to the eye, there is field curvature. There is always a region towards the edge of the image circle which is not so sharp. This effectively reduces resolution when the objective image is projected onto the flat sensor surface.
The eyes can adjust focus a bit, when using the eyepieces. Fortunately only the sharper middle-part of the objective image is sampled by the camera. Expensive plan objectives do produce nearly flat fields, so more camera resolution might be justified, particularly at low magnifications.
Bayer sensors have an overlay of coloured dots dominated by green. This halves the resolution of the sensor for a given colour. Some of that lost resolution is regained by interpolation, by using just the intensity data. It could be argued that this is a case for an even higher megapixel camera, but on a normal computer it will behave sluggishly while also exceeding the display resolution.
Resolution of Sampled Image
Because the sampled rectangle is smaller than the full microscope image circle, the required pixel resolution will actually be almost halved. So we are back where we started. Personally I would choose a camera that could almost fill a display that I could afford. For me that is 1.3 megapixels to fill the screen vertically or 2 megapixels horizontally, so perhaps I could get a 3 megapixel camera to produce a quality image after some cropping. More specifically I was interested in the AmScope 3MP High-Speed Camera SKU:MU300B. All three resolutions, 680x510, 1024x768 and 2048x1536 pixels have good capture rates on USB2.0. A 0.5x C-mount reduction lens is also included. This matches the projected image to the sensor size.
Resolution and LCD screens
If the just the vertical resolution of a typical LCD display, at 1024 pixels, is considered then a 1.3MP camera should be satisfactory. A 5MP camera could be used on its fast update, 1280x960 pixel, middle range for routine use and only used at 5MP for large display photos. None of the choices are that expensive. In general, the lower resolution cameras are less noisy and can cope with a wider range of illumination levels. Omax offer 0.75x, 0.5x and 0.37x C-mount reduction lenses which can be chosen to optimise the camera viewing area.
Not all the screen area is used. An imaging program typically has a display window which is smaller than the full LCD. For one program, that I use, the display window is 0.71 times smaller horizontally and 0.78 times smaller vertically. The useable window on a common 21.5 inch 1920 x 1080 pixel LCD display is approximately 1360 x 840 pixels or 1.15 Megapixels.
Resolution and Publishing
For some users, publishing requirements may dictate the required camera resolution. An image printed at 300 dots per inch on half an A5 sized page, with margins, would need to have a resolution of about 2 MP. A 3 to 5 MP camera would meet most needs. But typically many papers have montages of much smaller images. Each small image focuses on a particular detail. So again, 1.3 MP will do.
Cameras and Trinocular Adaptors
Some microscopes have inconvenient trinocular adaptors which do not fit standard objectives. This makes many microscope cameras difficult to fit. There are a few options:
Omax 1.3 MP microscope camera
I ended up purchasing an Omax 1.3 Megapixel camera SKU:A3513U at US$94.99 plus shipping. The camera performs well and delivers very nice images, some of which are shown below and at right.
I also purchased a 0.37x reducing lens for my original 2MP JEPCAM which was made from an adapted web camera. Some images from this camera are shown at right. Modern microscope cameras are sufficiently inexpensive, and of such high quality, that it is no longer worthwhile modifying web cameras to suit.
I use the supplied Touplite software on my iMac and ImageJ for cropping and adding scale-bars. The Windows version of the supplied software is ToupView. It is more comprehensive and has more editing features, including scale bars. Touplite on the iMac is fine for my purposes.
Diatom - Pleurosigma angulatum, dark-field illumination, 63x objective, Omax 1.3MP camera
Compensating Plates and Imaging
Polarised light is used in geology to help to identify minerals in thin section. Compensating plates add about 530 to 560 nm to the interference colour wavelengths produced by minerals. This means that low birefringent minerals with almost monochrome interference colours are now brightly coloured. Yellow and blue are often seen, but many colours can be produced. The colours and extinction angles can be used to help identify some minerals, or at least help to determine the optical sign.
Sensitive Tint Imaging Improvements
I have experimented with polarised light microscopy since 1974 when I worked in the DSIR. I found that modern plastic materials can behave like compensators. Some plastics used for packaging are typically biaxial and offer a maximum retardation of about 100 nm. This makes them ideal for trimming the response of the standard 530 nm full wave compensating plates. They can be used to improve viewing and photography through the microscope.
The following method gives well coloured images for samples which are transparent, with low contrast and low birefringence. This procedure is equivalent to trimming the thickness of the standard full-wave gypsum "sensitive tint" compensator. This produces the best possible contrast. For example an additional 1/4 wave compensator plate can be placed in the condenser, above the polariser, and rotated until the colours are suitable.
If a Berek or Quartz-Wedge compensator is inserted, a fairly sharp boundary is seen between first order yellow and second order blue. My procedure trims all the lighting exactly onto this boundary. Some low birefringent samples are seen as distinctive combinations of yellow and blue. The whole field of view now has the same magenta background colour.
For some gypsum "sensitive tint" plates this trimming may not be needed, but for the ones I have here from Leitz, Olympus and Swift, it is. This procedure is mainly an aid for photography. For quantitative work use just the gypsum "sensitive tint" plate.
The following image shows first-order yellow bordering second-order blue, using a Berek compensator which tilts a gypsum or magnesium fluoride plate. This creates a range of thicknesses equivalent to several orders of retardation. The corresponding interference colours are produced when the Berek compensator is inserted between crossed polars. Sample details are often seen more clearly at the magenta border between first-order yellow and second-order blue. The image is a little unsharp because of the tilted compensator plate above the objective.
Oamaru Diatoms, Berek compensator, 10x objective, Omax 1.3MP camera
For the photo below of the Diatom - Nitzschia circumsuta BW-113, prepared by Stuart Stidolph in 1982, I inserted a strain-free perspex plate, covered with a clear 0.01 mm thick mylar plastic film, above the condenser polariser. The mylar film was originally used in XRF sample preparation. It sticks to Perspex electrostatically. It has a maximum retardation of about 200 nm which is high, given the thickness. This is possibly due to the thin mylar being more severely extruded than packaging plastics. My Perspex sheet came from Bunnings and I was pleased to see no sign of birefringence between crossed polars. As far as I can tell it plays no role in the improved image properties observed. I inserted a full wave gypsum "sensitive tint" compensator into the standard slot above the objective.
The mylar film was oriented on the perspex until the background colour observed, bordered between first-order yellow and second-order blue. The background colour is set a bit more towards yellow than a normal full wave plate. Compensators made from some thicker plastics produced the same colours. In the past, I have also used mylar film from an oven bag, as a near full-wave compensator, for similar applications. Secure semi-rigid plastic packaging also works. Any slightly birefringent sample either shifts the observed colour towards blue or yellow, depending on the sample structure, and on the orientation. This can make biological samples much clearer as different structures now can have different colours. These colours will change with orientation. Some samples work well with just the normal polarising microscope compensators.
For microscopes without a filter holder above the condenser polariser, some mylar film can simply be adhered to the bottom of the slide. With the gypsum compensator inserted, the stage is rotated until a suitably coloured image is obtained. The camera is then rotated to restore the desired image orientation on the screen. The camera white balance setting can also be used to further improve contrast, if needed.
Diatom - Nitzschia circumsuta BW-113, polarised illumination with dual compensators, 63x objective, Omax 1.3MP camera
Sensitive Tint the Low Cost Way
Many microscopists do not have access to a polarising microscope but they do have some polaroid filters. One polaroid filter can be secured below the condenser and the other directly above the sample. Turn one filter until the view is fairly dark. 3 or 4 layers of mylar are adhered to a slide or a clear filter. Other plastics may need more layers. Sources of plastic include packaging, report covers and oven bags. If filters are cut from the same sheet, maintain all in the same orientation. This filter stack can be placed above the lower polaroid and rotated until the background has a magenta or pink colour. Adding or removing mylar or plastic layers and rotating the slide will produce the right colour. Use the camera white-balance to help with contrast.
For whole-slide macro photography I took many photos for students using a circular polariser below the sample. The intention was to produce a coloured general view of the slide. The camera lens side of the polariser faced the sample slide. A plain polariser (analyser) was used above the slide. A circular polariser could be used instead, with the camera lens side facing the slide, if an additional 1/4 wavelength retardation was required.
Minerals, which were identical and likely derived from a larger grain, were in similar orientations and showed the same colours. The colours observed were less dependent on the slide orientation than the usual crossed polar setup and extinction was rarely seen. This setup can be duplicated with the polarising microscope by adding 1/4 wave compensators, typically above and maybe below the sample slide. Sample birefringence will still be seen but it will be less sensitive to slide orientation. In this case the 1/4 wave compensator cannot rotate with each polariser, so the situation is only partially equivalent. Essentially this setup produces an almost extinction free, but still polarised, image which is somewhat insensitive to orientation.
Dark-Field and Circular-Oblique Illumination
I cut out some plastic disks from a black binder cover, using some hole punches. The disks range in size from 7 mm to 16 mm. One disk has a semicircular cutout at the edge to provide two levels of oblique lighting and the option of shading for contrast. The larger disks suit the more powerful objectives, although they are quite critical to align for good dark-field work. A 10x to 20x objective is a good choice for initial work.
Circular-oblique illumination is less critical. An annular ring of light can be seen by the objective but the image retains some shading, to show relief. The amount of shading depends on the disk alignment and its perfection. A crescent shaped cutout on one side of a disk can improve the image by providing light with different strengths and directions.
The disks are placed on a tabular perspex filter holder and secured with a spot of Bluetac. The filter holder is inserted into the condenser slot of the microscope. The Bluetac allows the disk to be moved and centred with the aid of the Bertrand-lens. Removing an eyepiece is another option for disk-alignment observation The setup is not dissimilar to the lighting setups used for portraiture and movies. More than one light source (the ring and the cutout) and some delicate shading (slightly off centre) can produce a nice image.
With the condenser top-lens near to the slide the illumination is usually dark-field. With the condenser higher still the illumination becomes circular-oblique. This lighting works best with the stronger objectives. An improvement in resolution is usually observed with this illumination. Some of the references at right describe both techniques. Microscopy-UK has many articles of interest.
The next item of interest to me is a dark-field condenser. Some use reflecting optics to produce a ring of light which converges on the sample. Others use a disk, as already described.
LED dark-Field Illuminator
I have also been experimenting with a small LED ring-light mounted just under the stage aperture. The LED ring-light mounts on the fold-away condensing optics and therefore uses the condenser adjustments. The tiny LEDs are recycled from faulty light-bulbs. They are mounted between tiny rectangles of circuit board which are glued inside a short 24 mm ID aluminium ring. 8 LEDs are wired in pairs with 390 kohm dropping resistors. The LEDs are controlled with a 4 channel rotary switch so all 16 combinations of 4 lights are possible. This means that not only dark-field illumination is possible but many combinations of oblique illumination are as well. A fixed aperture above the LEDs protects the microscope optics from too much stray light.
So far the conclusion is that good dark-field images can be obtained for survey work. These results are, however, no better than using a circular disk in the condenser. The range of oblique illumination is more interesting, for photographic work. The LED ring-light also makes a very good dark-field and oblique illuminator for a stereo microscope.
Diatom - pores, dark-field illumination, 63x objective, Omax 1.3MP camera
Diatom - pores, circular-oblique illumination, 63x objective, Omax 1.3MP camera
Pentax Macro Camera
The bellows lens extension on the Pentax digital camera uses a reversed 28mm lens as the objective. Turning the lens around means the optical paths through the lens are similar to those in normal photography. The lens functions as the designer intended and image quality is high. Longer focal length lenses can be used for lower object magnifications. Microscope objectives can also be used.
Illumination is provided by low cost white LED light sources. The LED power supplies have been modified to run at 1.5 volts, using a switched mode oscillator. This is efficient and the light output only varies a little as the battery voltage falls. A second spot light is a gooseneck LED lamp. It uses a 4 cell rechargeable battery pack with a USB output.
Canon Macro Cameras
I have macro adaptors for Canon A75, A590 and G5 cameras. These cameras fit onto a dedicated LED illuminator and closeup lens. This setup is small and is suitable for field use. The plasma ashed lemon and carrot leaf images were taken with a Canon A75 on this illuminator.
My Stereo Microscopes
I have two stereo microscopes. One is a Kyowa stereo microscope with 10 and 30 times magnification. I added a bike LED front light shown at right for strong incident illumination. I also have a Cooke stereo microscope which is capable of both incident and transmitted light viewing. The magnifications are 15.6, 44 and 125 times. This microscope has been slightly modified for polarised light viewing. The latter setup was particularly useful for looking at the crystallinity of plastic films. Strong LED illumination was needed for this application.
Digitech 5MP USB microscope
USB Microscope Upgrade
ImageJ and Scale-Bars
The microscope is now a reasonably professional instrument. I use ImageJ instead of the supplied software. I use the ImageJ scale bar tool along with two white marks added to the camera focusing wheel. The left hand mark is placed on the right edge of the focussing wheel when it is turned fully to the right. The right hand mark is placed on the left edge of the focussing wheel when it is turned fully to the left. The field width is 1600 Pixels. For various settings of the focus the field width in mm is:
I have created a shortcut so I can just press "0" to start viewing.
The transparent shield on the nose of the microscope could be used to temporarily capture small insects for observation. It can also protect samples from the wind when doing field work. This only applies at the listed magnifications.
Canon Powershot G5 macro camera
Two unusual Diatoms from New Zealand: Tabularia Variostriata a new species and Eunophora Berggrenii. Margaret A. Harper*, David G. Mann and John E. Patterson, Diatom Research (2009), Volume 24 (2), 291-306.
New diatom taxa from the world’s first Marine Bioblitz held in New Zealand: Skeletomastus a new genus, Skeletomastus coelatus nov. comb. and Pleurosigma inscriptura a new species. Margaret A. Harper*, John E. Patterson, John F. Harper, Acta Botanica Croatica 68 (2) 2009.
JEPSPECTRO - Home Page
Email: replace at with @
Light MicroscopyImageJ - Image processing and analysis in Java - PC, Mac or Linux
Resource Guide: Microscopes On a Budget
McCrone Research Institute
Microscopy Society of America
Microscopy Resource Center
The Quekett Microscopical Club
ADIAC Diatom Image Database
Royal Microscopical Society
Fiji - ImageJ together with Java, Java 3D and a lot of plugins
MicrobeHunter.com - Microscopy Magazine and Blog
Microscope - Microscopes For Every Application - Microscope World
Microscopy New Zealand
Modern Microscopy Journal
Microworld Resources and News
WWW Virtual Library of Microscopy
Biomedical Imaging Research Unit
Microscopes For Kids Resource Guide
Close Up PhotographyInexpensive Macro Photography
Lambert Smith's Insecta
Mounting Digicams With Zoom Lenses Over The Microscope
Insect Macro Photography
Microscope CamerasProscitech - Search for Cameras
AmScope Microscope Cameras
Omax Microscope Cameras
Microscope Cameras from Microscope.com
ProScope 5MP Microscope Camera
Numerical aperture and resolution
JEPCAM with DSE 2MP webcam inside
JEPCAM showing 10 micron spaced bars. 100x objective
JEPCAM - Oamaru Diatom, 20x objective
JEPCAM - Sponge Gourd stem, 40x objective
JEPCAM - Wheat grain, 40x objective with added scale bar in ImageJ
Plastic bag seal cross section
Diatom - Pleurosigma angulatum, dark-field image, 40x objective, 2MP JEPCAM
Diatom - Pleurosigma angulatum, oblique illumination, 63x objective, 2MP JEPCAM
Leaf stem - Nerium Oleander, polarised illumination with dual compensators, 10x objective, Omax 1.3MP camera
Pentax macro camera - old watch movement
Canon Powershot G5 macro camera - tooth section
Pentax macro camera
Macro sample positioner
LED Bike lamp illuminator
Digitech 5MP USB camera
Digitech 5MP USB camera, Barrytown sand sample