While there are insects involved in this posting, it’s actually more about a new, low-cost method I’ve come up with for better flash macrophotography. I’m very pleased with the way it works.
So, here I am with a digital SLR and a macro lens (both bought used to keep the cost manageable), and some domesticated feeder crickets (most likely Acheta domesticus) that I want pictures of. While it’s possible to take pictures of the crickets with just the built-in pop-up flash on the camera, the images come out with a mixture of specular highlights and deep shadows that make it harder than I would like to see details clearly. I also have to open up the aperture enough that the depth of field is pretty shallow, making it hard to get a well-focused image. So, for example, this one has a lot of glare off of the wings, and the lower part of the body is a shadowed and a bit out of focus.

Up to now, I’ve been trying to live with this, mainly by using photo-editing software [1] to adjust the brightness and contrast, but this is not really ideal. I’ve been looking into the preferred methods of insect macrophotographers for some time now, and the recommendation is to get a good macro flash unit. In particular, people with Canon gear have said good things about the capabilities of the Canon Macro Twin Lite MT-24EX. Then I looked at the price.

“Suggested Retail Price: $829.99”

Holy Cow! That’s almost four times what I paid for the camera body! Ouch! And they don’t seem to be very available on the used market, either.

The next-most-recommended option is the Canon Macro Ring Lite MR-14EX. That one is cheaper . . . I guess . . . although $549.99 is still COMPLETELY OUTRAGEOUS! FOR CRYING OUT LOUD, IT’S A FLASHLAMP NOT THE LBNL ADVANCED LIGHT SOURCE! rather more than I really want to spend at this point. Granted, these show up somewhat more frequently on the used market, but even used ones are still pretty pricey.

So, it’s time for a serious rethink: the fundamental problem is to illuminate a small object, quite close to the camera, with light that is simultaneously very bright (to allow fast shutter speeds with a small lens aperture) and diffuse (to both fill in shadows and avoid specular reflections). Having this light actually electronically controlled by the camera is a big plus[2]. Both of the expensive flash solutions above are basically approaching the problem the same way: put multiple light sources practically on top of the subject, blast the entire area with huge amounts of light from several directions at once, and hope for the best. But, aside from the cost of the flash units, I don’t really care for the fact that this also takes a lot of power, which means a lot more fiddling around with batteries just so I can spray more light around in irrelevant directions.

But, consider: I already have a flash on the camera. It’s pretty bright, too. Quite sufficient for lighting up an entire room, in fact. It’s certainly capable of providing enough light to adequately expose a volume of a few cubic centimeters that happens to have a bug in it. It’s just that the light is coming from the wrong direction for macro work. Most of the light generated never comes anywhere near where I want it, as shown here:

And, the light that does hit the subject reflects straight back, making bright reflections off of any shiny surface and shadows elsewhere. And for extreme macro, the subject is so close to the camera that the lens gets in the way of the flash. But . . . What if we come up with a way to redirect the light so that, instead of the flash being a point source spraying photons all over the room, the light from the flash is captured and concentrated where we want it?

Hmmmm.[6] At this point, I realized that something I had played with years ago might be applicable here – a non-imaging light concentrator.

The basic idea of a non-imaging light concentrator is that if you just want bright light in a particular spot and don’t care about making the light form an optical image, you can design a reflector that will accept light from multiple directions, and concentrate it into the desired volume. The concentrated light then illuminates any subject in that spot from all sides, making a very bright yet diffuse light. This idea has already been used for concentrating sunlight for solar power applications, but as far as I can see nobody has applied it to photography. So, how about if I design one that will catch most of the light from my on-camera flash, and direct the bulk of it into the relatively small volume that I actually want to illuminate?

While an ideal light concentrator is made from mirrored compound parabolic surfaces, for my purposes we can do pretty well with just a white cone. If the cone is set so that light from the flash goes into the large end, the light that reflects off of the cone wall will end up concentrated in the volume at the small end, changing the distribution of the light to this:

So, what I ended up doing was to take a stick, fasten it to the camera with an appropriate bolt into the tripod mount hole, and tape on a cone made of white posterboard, like so:

Cone for standard macro lens

I then made a second one for higher-magnification macro photography, where the subject is only an inch or two in front of the lens:

Cone for close-focus macro lens

In both cases, the cone is cut so that I can still turn the focusing ring on the lens. The tip of the cone is about an inch short of where the object being photographed sits.

So, what difference does it make? First, let’s look at an adult cricket (a dead one, to avoid being distracted by chasing a cricket around), using the 100 mm lens at 1:1 magnification[3] with the lens aperture set at minimum size (f/32) to give the greatest depth of field (up to now I’ve been using a much larger aperture, because I didn’t have enough light to choke the aperature down this small without a very slow shutter speed). First we’ll try it without the flash cone, and then compare with the identical camera settings with the flash cone:

Adult without flash concentration

Adult with flash concentration

As you can see, with the cone the whole image is markedly brighter, shadows are mostly filled in, and it is generally much easier to see detail. Now, let’s look at a cricket nymph using the higher magnification setup (a reversed 50 mm lens mounted on the end of the 100 mm lens. This looks to give a magnification of about 2:1, which is more of an increase in magnification than it sounds like). Again, the camera settings are identical in each case:

Nymph without flash concentration

Nymph with flash concentration

Here, we see that the flash concentrator again makes a huge difference. Without it, the subject is dark and practically indistinguishable from the background. But with it, the cricket nymph is uniformly illuminated with all features clearly visible.

Overall, the effect is similar to using a “White Box“, except instead of being an immobile stage, it is a portable thing attached directly to the camera, making it easy to use in the field. The flash concentrator cone is a bit bulky, but no more so than the supplemental flash units that would be needed otherwise. And best of all, it is

– Cheap[4];
– Easy to make[5];
– Cheap;
– Pretty rugged;
– Applicable to most cameras with a built-in flash;
and
– Did I mention Cheap?

While I’m sure that the expensive flash solutions can do a number of things that the flash cone can’t (at those prices, they had better!), the cone still goes a long, long way towards solving my long-running lighting problems. So, go ahead and try it out, and if any of you develop further refinements, please let me know.
———–
[1] I use ImageJ, partly because I already use it for microscope image analysis at work, so I am used to it. But, I mostly use it because it is free.

[2] I have some cheap “studio slave flash” units that I had bought to use with the previous camera that could help with this, except for the fact that they don’t play well with the flash unit on the Canon 10D camera. The old Powershot A95 allows a full manual mode where the flash only fires once, but the 10D evidently pretty much forces the flash to fire twice for every picture – once to adjust camera settings and once to take the picture. There is an option where I can manually force the flash to fire the first time, and then take up to 15 seconds before actually taking the picture while the flash fires the second time, but this means that the slave flashes only have a few seconds to recharge. Anyway, they aren’t very nice to work with using the current camera.

[3] A 1:1 ratio means that the image of the insect on the camera sensor is the same size as the actual insect. And, since the sensor is pretty small, blowing up the image on the sensor to a photograph that you can see results in a very large picture of an insect. A ratio of 2:1 means that the image on the sensor is twice as big as the actual insect, so this is good for photographing even tinier things.

[4] Let’s see, parts were:
– a wooden yardstick ($0.79),
– a sheet of posterboard (also $0.79),
– a 1/4-20 thread screw ($1.00 for a capscrew, or $2.50 for a clamping knob which is a bit nicer to work with),
– a snap ring to keep the screw from falling out of the ruler ($0.49),
– a piece of adhesive-backed craft foam ($3.00, but only because I had to buy a whole pack to get the thin strip that I wanted),
– and some packing tape to hold it all together (maybe a couple of cents).
Total materials cost was therefore on the order of $6, with two-thirds of the yardstick, 3/4 of the posterboard, and a whole bunch of adhesive-backed foam left over afterwards to use for other things.

[5] To make one:
– First, measure how far it is from your camera to the closest focus point you want to use.
– Then cut your yardstick to be long enough to go from the camera’s tripod-mounting hole to just short of that focus point.
– Drill a 1/4 inch hole in the yardstick piece, and put in your 1/4-20 capscrew.
– Put the snap ring onto the capscrew to keep it from falling out of the yardstick piece.
– Cut a piece of adhesive foam to put onto the yardstick, make a hole so that the capscrew can go through it. This will make a base for the camera to sit on without wiggling.
– Cut an arc out of the posterboard, like so (measurements are approximate, and were mainly chosen for convenience):

– Roll the posterboard arc into a cone so that the large end is about 7 inches in diameter, and the small end is about 5 inches. Tape it to keep it from unrolling.
– Tape the cone to the piece of yardstick so that the small end of the cone lines up with the tip of the yardstick piece.
– Fasten it to your camera by screwing the capscrew into the camera’s tripod mount threads, and you’re done! If necessary, use scissors to trim the cone so that you can reach the controls of your specific camera better.

[6] This line of thought was partially started when I saw what happened when I was photographing this spider a few weeks ago, inside of a conical plastic bucket. The flash fired directly into the mouth of the bucket, and the bucket wall actually reflected and diffused quite a lot of light. The bucket was blue, and so the reflected light was also, which was not exactly ideal, but it was along the right lines. Which incidentally suggests that an even simpler solution is just to put your bug into a white plastic bucket with at least an 8-inch mouth, and just start shooting away.