Snow Crystal Photography

by E.M. Kinsman, Rochester, New York

This article is titled Snow Crystals, even though many people would like to call the crystals snowflakes. Scientifically the term snowflake relates to large clusters of dendritic flakes that fall in clumps. The single crystals are called snow crystals. Ice crystals are formed from freezing liquid ice. During this project I photographed both snow crystals and snowflakes.

I would also like to state that this short article is an introduction to an amazing topic that can easily absorb a whole life of study. The many aspects of the "simple ice crystal" are nothing but complex, intriguing, and fascinating. I encourage serious students to look up the two references below as a starting point. It would take several text books to cover all the aspects of just ice, not to mention the complex nature of liquid water.

Polarized light shows different ice crystals frozen from liquid water - I should mention this shot was rejected from a prestigious weather photography competition as being "too scientific". An honor I consider better than the winning prize!

A snow crystal that has been outlined in Photoshop®.
Dendritic form.

For the past 8 years I have been experimenting with ways to document snowflakes. I have tried to preserve them in optical ultraviolet cement, built a laser system to record the diffraction patterns, and more recently used video techniques to make a survey of the crystals.

It became obvious from the experiments performed during the 1999-2000 season that to correctly record snowflakes, a photographic system needed to be built that had the following parameters.

I chose to use the classic Olympus OM2N camera on a B&L microscope body.

Although ice does rotate the polarization field as many crystals do, snow crystals are too thin to use this property of light to show the differences in the crystals. Polarized light only has the effect of changing the color of the background. For this reason I have designed the photographic apparatus to only use bright-field and dark-field illumination, not polarization.

In many storms the precipitation is often just about all identical and many times not worth photographing. My goal on this project is to collect high resolution images that are pretty, although I could not resist collecting some images of scientific value. I found extreme close-ups using a 10X objective (62x at the film plane) to be my favorite. If I wanted the whole snowflake I used a 5X objective that gave about 8x on the film plane. Observant readers will realize that the last two values are not in ratio to each other - this is due to the microscope having an adjustable tube length. A highly useful characteristic of this particular microscope.

A common lake affect snow storm. The snow crystals have collided with super cooled liquid water. The crystals grow fast and are formless. Dark-field illumination.
I use a tungsten illuminator, so I chose Kodak tungsten 64 ISO film. Now many microscopists would take great amounts of time getting the color balance correct for the application. Oddly, if the snowflakes turn out a bit blue it only adds to the artistic nature of the images. Due to the automatic exposure on the camera, I like to over expose the bright-field illuminated snowflakes, and under expose the dark-field flakes.

Of all the snowflakes I have photographed this winter, the crystals of greatest interest are the large dendritic forms that I believe fall into two classes; dendritic and the dendritic rime forms. Dendritic crystals are built of water that is evenly deposited and the flake is optically clear and behaves like a lens. Dendritic rime flakes look the same in form, but under a microscope you can see that they have a large number of microscopic ice spheres attached to them. These flakes will appear black under bright-field illumination and thus are better photographed under dark-field.

Close up of a dendritic flake. It behaves like a lens.

Dendritic crystal beginning to be coated by the supercooled water;
the frozen droplets behave like micro lenses and really scatter the light.

Large dendritic crystal covered with rime. White light illumination.

Dark-field illuminated dendritic crystal.

Here in Rochester, NY we are just a few miles from one of the great lakes (Lake Ontario) that greatly affects our weather. One of the common forms of snow happens when the warmer lake's air mixes with a cooler air mass. The resulting snow is called Lake Effect Snow, it is characterized by rapidly growing crystals that are basically formless spheres and elongated masses. The pretty dendritic forms are created in airborne temperatures of approximately 13°C. Once the crystals can form they fall through warmer air and build up ice under different crystal orientations. The opposite is also true - where warmer ice can fall through a cooler layer and an ice crystal starting in an ugly mass can start to grow the perfect dendritic arms. Snowflakes can also fall through layers of different relative humidity where they can both build up and lose mass. Ice can also sublimate. Here in New York, I estimate that we get 15 good storms when it is both cold enough to photograph and there is good snowflake development.

Due to my proximity to the Great Lakes, there are a number of snowflake types that are either very rare, or just do not form in this location. These include the very cold forms, and the top-hat shapes.

You can basically think about photographing a snowflake as being identical to optically testing a small piece of glass or micro-lens for defects.

Dendritic, 10x objective.

Dendritic, 10x objective.
Image used for false color tests below.

Rarely do snow crystals exist as a flat object, flakes are very three-dimensional objects and due to their thickness some of the snow crystals will be out of focus. Some forms of dendritic flakes go in all directions and are impossible to get a picture of, unless you hack off some of the protrusions.

I only photograph snow crystals when the ground temperature is at or below 25°F. Warmer temperatures cause the snowflakes to quickly melt around my body's microclimate of warmer air and higher humidity. The illumination system of the microscope, even though it has an infrared filter to block out heat energy, it radiates enough heat to warm up the snowflake a few degrees and cause melting.

The process of snow crystals photography is quite simple. I use a sheet of masonite to collect snow crystals. The sheet of masonite is placed just outside the garage door, while the microscope and camera are in the garage. The microscope is kept in the garage to keep it cool, while the camera is kept in a plastic container and moved to the garage when there is good flake development. When the photography is done, the camera is placed into the plastic box and moved into the house. The plastic box keeps water vapor from the warm house condensing on the camera. The camera system can be used if the microscope is moved into a warmer environment.

Good snow crystals are picked up with a thin pin taped to the end of a pencil. The flakes are transferred to a pre-cleaned microscope slide. I typically try to pick out the best dendritic flakes that fall on the sheet of wood. If the flakes fall on the microscope slides the chances of getting a nice flake are relatively few. The majority of flakes that fall will collide with other flakes and loose arms, or will grow unevenly. Since the snowflakes have an electrostatic charge, they often clump together and form falling masses. Some times the masses can form more interesting images than solitary flakes. Collecting flakes in a snowstorm is often a challenge, any snow is accompanied by huge collections of broken flakes, bits, and little particles or dust, as well as micro crystals. All of these particles will end up on the slide.

Once a snowflake is photographed it can be cleared off the microscope slide with an artist's paintbrush, and the slide can be reused a few times. I like to photograph snow anytime the flakes are good, but the cooler temperature required, mandate that I photograph late at night. With good crystal development and a quick setup you can take about 30 images an hour. Many times the rates are nowhere near this number. For every 30 images taken I would estimate that you might get 2 good images. You cannot really tell how good a flake is until you get it under the microscope. Even broken flakes can have some interesting parts on them, so I have a 10x objective that I swing into place to see if I cannot get an interesting image out of a bad flake.

I have a lamp post in my front yard that I can quickly look at night to see if it is snowing and what type of snow is falling. Dendritic flat flakes reflect light so if you look a few degrees away from the light and see little reflections then there is good snow falling. Hard snowstorms have too many broken flakes and bits to bother going out and photographing. The most desirable storms are when it is just barely snowing and there is no clumping of the flakes yet. Many times a storm can have windows of different types of flakes. Some of these windows of good flakes will only last a few minutes, so I collect my ½ square meter of masonite wood with the good flakes and move it into the garage (I also keep the garage door open to equalize the temperature), then I can take my time to photograph the good flakes while a polycrystalline mess falls just a few feet away. Of course the collected flakes will change depending on the humidity conditions on the ground (grow larger, smaller, condense).

If you observe a snowstorm and decide that the flakes are not desirable, just wait 20 min and look again. The snow forms can change quickly and you must be ready to act quickly. This waiting and hoping for good snowflakes can lead to sleepless nights.

I am quite surprised at the lack of snowflake photographers. After corresponding with numerous interested photographers, several important variables to make this process work have come to my attention.

The photographer must meet the following criteria:

The first point eliminates 50% of the photographers, while the second point eliminates another 49.99% while the last four points eliminate just about everyone else. I would be interested to hear from anyone else actively involved in snowflake photography.

An image is converted to black and white, then false colors are applied to highlight details.

Another false color image.

Future Experiments

The idea of thinking of snowflake photography as a form of optical testing is of particular interest. There are a number of both optical and digital techniques that can be employed to document snowflakes.
I am also intrigued by the idea of growing heavy water (deuterium oxide) crystals. Since D2O freezes at 3°C in theory I should be able to grow frost underwater. In this case thicker crystals can be photographed by using polarized light, schlieren photography, or a modification of optical lens testing called Ronchi Testing.
I am also interested in photographing more snowflakes with a much smaller F# so as to yield a much greater depth of focus.

Comments to the author Ted Kinsman are welcomed.

For additional snowflake / snow crystal images check out the author's web site, Kinsman Physics Productions:


Field Guide to Snow Crystals
Edward R. LaChape
University of Washington Press 1969
This is a nice little text showing the changes in snow crystals due to time, humidity, and evaporation. The results are applied to the field of predicting unstable snow fields or avalanche prediction.

Snow Crystals: Natural and Artificial
Ukichiro Nakaya
Harvard University Press 1954
This is the definitive textbook on snow crystals. Nakaya reproduced just about every natural form in his cold laboratory. A wonderful researched book by a brilliant scientist. This is a must read for any serious snow crystal researcher.

Microscopy UK Front Page
Micscape Magazine
Article Library

© Microscopy UK or their contributors.

Images must not be used without the author's permission.

Published in the May 2001 edition of Micscape Magazine.

Please report any Web problems or offer general comments to the Micscape Editor,
via the contact on current Micscape Index.

Micscape is the on-line monthly magazine of the Microscopy UK web
site at Microscopy-UK


© Ltd, Microscopy-UK, and all contributors 1995 onwards. All rights reserved. Main site is at with full mirror at