Photographing snowflakes can be great fun and a rewarding hobby.  It is a quiet, contemplative activity that takes you outdoors on snowy days to examine and record some of Nature's most stunning artistry.  Best of all, each picture you take will be different from all others, because each snow crystal is unique. 
     Unfortunately, snowflake photography is not cheap, and it requires that you to spend some quality time outside in the cold.  But if snowmobiling (which is also mighty cold and expensive) is not for you, then you may want to give snowflake photography a try.

     Before spending a lot of money on photographic equipment, I suggest that you first try some basic snowflake watching.  Carry a small magnifier around in your pocket for a little while and see what the crystals look like where you live.  If you like what you see and want to go further, then consider buying a microscope or going full-on and building your own snowflake photography set-up.

     In this section I give detailed instructions for building your own snowflake photo-microscope, shown at right.  A full parts list is included, and you can build everything on your living room table with no extra tools or previous experience.  All that's needed is a loaded credit card and some free time.  With this set-up, you can take pictures that are as good as those in my photo galleries (or close to it, anyway).
     Alas, the hardware is rather expensive.  Photography is generally an expensive hobby, and micro-photography is even worse.  But if you compare the price tag with what you have to spend on a snowmobile, or even for a single weekend of skiing, it isn't so bad.  Also, the equipment will last for many years, so you can enjoy photographing snowflakes long after you've forgotten what you paid for the hardware.
     If you set out to build this instrument, I recommend that you print out this page along with the important pictures.  I'm constantly changing the website, and I cannot guarantee that this page will be here tomorrow.  I also recommend that you pick up a copy of my Field Guide to Snowflakes.  This book has lots of valuable information about snow crystals and how to photograph them.

Parts List:
Aside from the camera, the rest of the parts can be ordered online from Edmund Optics, ThorLabs, and McMaster-Carr.  If you try ordering from this list and find any errors, please send me a e-mail.

Parts -- The Camera  This you must supply yourself.  The only real requirement is that it be an SLR (single-lens reflex) camera body with a removable lens.  I find digital is much superior to film, but film will work also.  If you change cameras in the future, the rest of the hardware will require little or no modification. 

If you don't already have a camera, I recommend the Nikon D50, which is a high-quality, reasonably affordable 6-megapixel digital camera.  (See Steve's Digicams for reviews of this and many other cameras.)  (The camera in the picture above is a Nikon D1X, in case you're interested.)
Cost for Nikon D50: about $650 online.

WARNING:  Cameras are not designed to work in below freezing temperatures, so I recommend that you keep your camera warm at all times.  This is important!  If your camera freezes and breaks, don't say I didn't warn you.  More on this below.

Parts -- Lens Tube:

Edmund #52-296 100 mm T-mount extension tubes (order two)      (for a 35-mm film camera, order three tubes)	$  38.50  (prices as of 10/2006) $  38.50
Edmund #42-835 T-mount to camera adaptor      (this is for a Nikon camera -- pick an adaptor to match your camera)	$  29.00
Edmund #52-300 Focus Tube	$  66.10
Edmund #52-301 T-mount to microscope objective adaptor	$  27.50
ThorLabs #RMS4X microscope objective	$172.00
Cost of the lens tube:	subtotal $371.60

Parts -- Base Plate and mounting hardware:

ThorLabs #MB12 12"x12" aluminum breadboard	$162.00
ThorLabs #P14 post	$  74.00
ThorLabs #C1500 compact post clamp (order two)	$  59.00 $  59.00
ThorLabs #AS8E25E thread adaptor (order two)	$    3.90 $    3.90
Edmunds #52-304 T-mount ring (order two)	$  60.00 $  60.00
ThorLabs #C1501 mounting post clamp	$  58.00
ThorLabs #CAM1 right angle plate	$  89.00
ThorLabs #BD-3/16 wrench	$    4.60
McMaster-Carr #92313A537 1/4-20x0.5 SS set screws	$    5.88 (for 25)
McMaster-Carr #92185A541 1/4-20x7/8 SS socket head cap screws	$    6.04 (for 10)
McMaster-Carr #90313A203 1/4 washers	$    7.81 (for 50)
Cost of the base plate and mounting hardware:	subtotal $653.13

The grand total is $650 (camera) + $372 (lens tube) + $651 (base plate) = $1673 (plus tax, shipping, etc.).  And you'll need a computer to work with the digital photographs.  If you already own a 35mm film camera, then you need only buy the lens tube and base plate, which cost $1060, and you won't need the computer. 
     Even better, If you have some wood-working experience (or know someone who does), and are willing to improvise, then you could save $650 by building your own base plate assembly.  In that case, you would need to build some kind of structure that will rigidly support the camera, lens tube, and a platform to hold the snow crystals.  How you might do that is up to you.
     The above kit is intended for someone with no previous hardware experience -- every nut and bolt is provided, and everything can be assembled without additional tools. 

Assembly:
     Putting the parts together is fairly straightforward.  The lens tube simply screws together by hand.  From top to bottom, it consists of the camera, camera adaptor, two 100 mm tubes, the focus tube, microscope adaptor, and the microscope objective.  Assembling the lens tube takes all of five minutes. 
     Using two 100 mm tubes works well with most Nikon digital SLRs, where the sensor size is roughly 1.5 times smaller than 35mm film.  If you are using a 35mm film camera, use three 100mm tubes together to get the same field of view (more on this below). 

     For the rest of the assembly, refer to the picture of the final instrument above.  Start by setting the MB12 breadboard on a level surface.  Put a set screw in the hole of your choice, and screw it in by hand so that half of the screw is in the breadboard.  While using a knife to keep the set screw from turning (see picture at right), screw the P14 post onto the set screw.  When this is done, half of the set screw should be in the base plate while the other half is in the post.  Tighten the post firmly by hand, and give it a good shake to make sure it is secure.
     Next attach the mounting post clamp (C1501) to the post using the wrench (BD-3/16) (clamp2 in the picture).  Don't attach the right angle plate yet.
     Next attach the two tube clamps (labeled clamp1 in the picture) to the post.  Each clamp consists of a C1500 compact post clamp, a AS8E25E thread adaptor, and a 52-304 mounting ring.  The parts screw together by hand, and you use the wrench to attach each clamp to the post.  This also takes about five minutes.
     Next put the lens tube into the two clamps and secure them.  You should remove the camera and microscope objective before doing this step, so you don't break anything, or scratch the optics, etc.  You will also have to take the focus tube off so the tube assembly fits into the clamps.  You will have to loosen the clamps and the post clamps to get it to fit together, but it works.  Make sure the tube clamps are tight (by hand), so the lens tube is secure.
     Next attach the right angle plate using some washers and 7/8" long socket cap screws (see picture at right).  Put enough washers in so the hole in the right angle plate is centered with respect to the microscope objective.  Be careful not to scratch the optics during this step.
     The whole assembly should take less than a hour, and the final instrument should look like the picture above.
    

     To start using the microscope, place a piece of printed paper on the platform, shine a bright light on the paper, and look through the camera eyepiece (you do not need the right-angle eyepiece shown in the overall picture above).  Move the platform up or down (using the post clamp) until the paper comes roughly into focus.  Fine adjustment of the focus is done using the Focus Tube.  If you see some highly magnified print on the paper, then hooray -- it all works, and you're ready to start taking some pictures.
     Of course, I recommend you play around with the hardware indoors where it's pleasant before venturing outside in the cold to photograph snowflakes.  Use bits of clear plastic as surrogate snowflakes for practice (remember that ice is clear and colorless).  For illumination, it's best to shine light up through the snow crystals from below.  One way is to shine a bright light on a white card under the platform (see picture at right).  Be careful not to apply too much heat near the platform, however, since snowflakes can melt or evaporate away quickly.
     There is much to say about illumination; using a light and white card is just a quick way to get started.  When you become serious about it, the quality of your snow crystal photographs depends to a large extent on how well you do the lighting.  More on this below.

Keeping Your Camera Warm
     Before you go outside to photograph snowflakes, I recommend that you keep your camera above 32F (0C) at all times.  Your manual probably has a line buried inside saying how the camera is only designed for use above 40 F or so.  That little disclaimer means that if your camera freezes and breaks as a result, then the warranty is voided and the manufacturer will not give you a refund or replacement.  This is one of the hazards of snowflake photography, so it's up to you to keep your camera warm.  On the other hand, be careful not to overheat your camera.  Temperature extremes in either direction could wreck your camera.
     One easy and effective heating technique is to place an inverted styrofoam box over the camera and apply some heat underneath, as shown at right.  I built this box out of 1" styrofoam sheet held together with silicone caulk (both are available at most hardware stores).  I also mounted a 20W light bulb inside a cardboard tube and taped it to the lens assembly (see picture).  I wrapped the bulb in aluminum foil so no scattered light gets into the camera.
     When the lamp is on, the heat it generates rises and is trapped inside the styrofoam box, where it keeps the camera warm.  Note that with the shutter release cable and right-angle eyepiece, as shown in the picture, I can use the camera with the box in place.  I did some tests and found that the inside of the box was about 40-50 degrees F warmer than ambient.  So at 30F, the camera is no warmer than 80F, which is okay.  At 0F, the camera is no cooler than 40F, which is also acceptable.  I recommend that you put a thermometer inside the box and keep an eye on the temperature, since you don't want to take any chances with your camera.
     Heating your camera this way is a pain in the neck, and it gets in the way of taking pictures.  Maybe your camera will work fine at low temperatures.  I always keep my camera above freezing, and I don't know what will happen to your camera in the cold. 

     Another handy tip is to put your camera in a zip-lock plastic bag before bringing it inside from the cold.  The bag will keep condensation off your camera.  Once the camera warms up, you can take it out of the bag.

Miscellaneous
      If you take a picture of a clear ruler (here a reticle like Edmund #56-352)  with this instrument, you will get something similar to that at right.  The numbers are millimeters, so the field-of-view is about 2.5x4 mm, which is a reasonable choice for snowflake photography.  If you want to change the field-of-view (or equivalently, the magnification), use more or fewer extension tubes.
     You may have noticed that the reticle picture is brighter in the center than around the edges.  The excess brightness is more easily seen when taking a picture of a resolution target (Edmund #38-256), as in the first picture below.  Note the brighter haze in the center of the picture.  The source of this extra brightness is light that scatters around inside the lens tube, and it tends to give your pictures a washed out look.  The fix is to add a "field stop", as shown in the second picture below.  Here I just added some black tape outside the field of view of the microscope, which blocks all but the light you want.  As you can see in the third picture below, the excess brightness is gone and the image has a higher contrast than the first.

     When you start taking pictures with this instrument, you will notice that the threads of the Focus Tube are a bit sloppy.  You can fix this by taking it apart and putting a bit of Vaseline on the threads.  Don't use too much, however, or it will be difficult to focus in the cold.  You can put the Focus Tube in your freezer (take the microscope objective off first, so you don't scratch it) to see how easily it will turn when it's cold.

Photographing
     Once you are happy with the performance of your snowflake microscope, you're ready for the real thing.  I recommend setting it up on a bench inside your garage, or in some other unheated outdoor building, if you have one.  You can even leave your camera attached, provided you always leave the heater on.  (If you detach the camera, cover the lens tube to keep dust out, and always keep a lens or cover on your camera.) 
     If you keep your snowflake photo-microscope set up in your garage, then it will always be cold and ready, so you can start photographing whenever you see some nice snow crystals.  This will also keep it under shelter, so falling snow doesn't get into everything.

     Another challenging aspect of snow crystal photography is handling.  Snow crystals are small, somewhat fragile, and they melt away with the slightest breath.  The crystals also tend to evaporate away or melt once they're under the lights of the microscope.  This means you only have a few minutes to bring the crystal into view and compose the shot.
     A good strategy for handling larger snow crystals is to first catch them as they fall on some convenient cold flat surface.  Any dark-colored cardboard surface works well, and I find blue foam-core is nearly ideal.  Scan the board for interesting crystals, and brush the flakes off regularly.  When you spot a good one, carefully pick it up using a tiny paintbrush (say 1mm diameter brush, with a sharp tip) and place it on a clean microscope slide.  Place the slide under your microscope and take the picture.
     Temperature matters a lot.  If it's just a few degrees below freezing, snowflake photography is difficult.  The crystals melt almost as soon as you look at them.  When it's colder, the crystals are easier to handle and they last longer, giving you more time to compose.  I get my best pictures when the temperature is around -15 C (5 F), which is where stellar plates form (see the Morphology Diagram).

     Another technique I like is to simply let snowflakes fall on clean microscope slides.  Support 2-8 slides by their edges so you don't get icy glop on the lower surfaces.  When a slide has a nice dusting of snow, put it under the microscope and scan around for nice crystals.  When you've seen what one slide has to offer, wipe it off, place it back outside, and pick up another.  I've taken many excellent pictures of smaller snow crystals using this simple technique.

     People often ask me how I took the pictures in the Photo Galleries, so let's take a look at my setup, which I call the SnowMaster 9000.  I tend to use expensive components to get the very best pictures, so this is something of a deluxe approach to snowflake photography.

     When I first began toying with the idea of writing a book about snowflakes (see Snowflake Books for the results), I found that there weren't any really nice pictures out there (in my opinion, anyway).  The state-of-the-art in snowflake photography was not what it could have been. This was an opportunity, so I set out to design and build a photo-microscope that was optimized for taking pictures of snowflakes. 
     There are a number of challenges in photographing snowflakes.  The crystals are small, so the usual techniques of macro photography don't work well; what you need is essentially a high-quality, low-power microscope.  In addition, the equipment must be robust enough to work outside, and it has to withstand sub-freezing temperatures.  Finally, to take good pictures, you have to take a lot of pictures.  Therefore, I wanted my hardware to be easy to set up and use.
     The eventual result of my design efforts is the snowflake camera shown at right.  The camera itself is housed in a heated enclosure (the white styrofoam box in the picture) so that it can stay outside in the cold indefinitely.  The whole assembly is mounted in an instrument case so I can take it on an airplane (since there aren't many snowflakes to photograph here in Southern California).  Finally, I put some effort into illumination techniques, mainly the use of transmitted light tailored by a variety of color filters for different effects.

     The overall layout for my setup is shown at left.  My camera is a Nikon D1X, which was a good choice five years ago when I bought it.  Digital is much better than film for snowflake photography, in my opinion -- digital pictures are less grainy than film, the overall image quality is better, it's easier to take lots of pictures, you can check the pictures as you take them, you can easily do post-processing, etc.  I find that my camera is actually not such a limiting factor, but nevertheless I intend to move up to a 10-12 megapixel camera one of these days.
     I use three microscope objectives - the Mitutoyo M Plan Apo 2X, 5X, and 10X objectives (available from Edmund Optics) - mounted on a home-made turret.  I like these objectives, partly because they have an extra-long working distance, but they are expensive (about $600 each).  I use the 5X most of the time, the 2X when the crystals are large, and the 10X only rarely. 
     The camera and objectives are separated by a length of rigid tubing, which is baffled to cut down on scattered light.  (There is also a lens mounted inside this tube above the objectives, which is a bit of a technical detail.  The objectives are infinity-corrected, so a lens is needed to keep focus when I rotate the turret.)
     The snowflake is placed on a platform attached to a translation stage for focusing.  As is typical for microscopes, it is more convenient to move the subject than to move the lens.  I attached a flexible shaft and a large knob to the stage, making it easier to focus with cold fingers.
     To illuminate the crystals, I use a lamp mounted below the crystal, as shown in the diagram.  Above the lamp is a heat-rejection filter and a piece of ground glass (not shown), and on top of those components I can place any of a number of color filters.  The field lens focuses the color filter onto the microscope objective (so the filter sits at a pupil).  What this does is discussed below.

     A number of people have asked me to share the design drawings for my hardware, so they can copy this set-up.  While I would like to oblige, I really don't have machine drawings or a parts list, at least not in a form that would do anyone much good.  Projects like this tend to be build-as-you-go, trying different things to see what gives the best pictures.  It would be quite time-consuming to document every nut and bolt in the system, so I haven't.  Instead, I put together the do-it-yourself kit in the previous section, which is almost as good for a lot less money.