Mostly there’s three topics: long exposures, stereo photography, and gear. The last two overlap.

I’ve been attending the local camera club’s meetings (Eaton Rapids Photography Club, find them on Facebook), and quite a lot of attention has been paid over the course of several meetings to the topic of long exposure photography. That’s something I used to have nailed down as a technique. How nailed down did I have it? I was using the Zone System for long exposure work, that’s how nailed down. Of course, the primary thing to nail down way back when was the unpleasant non-linear and even non-logarithmic business of “reciprocity failure”.

What’s reciprocity failure?, you may well ask. Digital sensors don’t have this peculiarity, which amounts to the photographic medium becoming less responsive to exposure as the light level approaches darkness. Films had a particular sensitivity to light given certain assumptions (mostly to do with how development in processing was handled). Under most conditions of exposure in daylight outdoors through indoor artificial light situations, you could select exposure that traded, stop for stop, between duration of the exposure (shutter speed control) and the intensity of light during the exposure (aperture control). Once light got very dim, though, this trade-off no longer worked the same way. Once you got to a point, varying by the film used, around 1 second in exposure, trying to trade a doubling of time for a halving of intensity (going for, say, 2 s at f/8 instead of 1 s at f/5.6) would get you a somewhat underexposed negative or slide. The effect gets worse as you try to increment time further. I spent a lot of time dialing-in exactly what reciprocity failure did when using Kodak Panatomic-X black-and-white film, and I had a set of handy time conversions committed to memory. I don’t think I have them readily accessible for recall now, but essentially given a light meter reading that would indicate, say, 4 minutes exposure at f/8, I’d use a much longer time instead, say, 10 minutes exposure at f/8 (not meant to be accurate; you’re going to have to calibrate your own film). For common times, I had corresponding adjusted exposure times mapped out of 10s, 30s, 1 min, 4 min, 10 min, and 45 min. Sorry, 45 minutes is where I run out of patience with long exposure.

The primary annoyances for time exposure in the digital domain have to do with battery life of the devices and how noise reduction is often handled. Because over a time exposure, the sensor electronics remain very busy, this tends to reduce battery life by quite a bit. The digital cameras I have tend to use a dark field subtraction method that means that once the photographic exposure is done, the camera immediately goes into a mode where it is busy creating a dark exposure at the same duration as the photo just taken. The two exposures are used to compensate for the actual noise contributed by the sensor. So if I choose to make a 10 minute digital exposure, my camera is going to be tied up for 20 minutes. I can’t review my work for that long. That’s still a better turnaround than for film, but digital spoils me for more immediate feedback.

I mentioned stereo photography. This was once a much bigger thing than it seems to be now. You can still get ViewMaster viewers and 3D slide sets, but the demand isn’t as high now as it used to be. Back in the 1950s and 1960s, you could get a serious piece of photographic gear, the Stereo Realist camera, that featured a pair of taking lenses separated by about the usual human interocular distance. One press of the shutter made simultaneous paired exposures, and one could use glasses for looking at the resulting pairs of prints to see things in 3D.

I have been thinking about stereo photography off and on for a long time. You see, I like to run a flashlight around in those long exposure photographs mentioned above. Sure, that leaves light trails that are pretty cool on their own in the photo, but I figure that if I can solve the stereo thing, that opens up a new dimension (the third one, natch) in how I can use light in long exposure, among other good things.

Recently, I was able to pick up a used Olympus E-M5 camera and grip cheaply. (That goes to the gear part of this post.) It seems to be in fine operational condition and even had a relatively low shutter count. That gave me a second E-M5 body. And that allowed a big advance on the stereo photography front.

You see, techniques exist for single-camera stereo photography. One that I had used in the past had an optical attachment consisting of two sets of mirrors so that a single exposure would yield a stereo pair, each part of the pair being on the left or right of the exposure. That’s limited to starting from a normal lens or longer focal length, otherwise it would vignette like crazy. But I somehow lost my attachment, and it never was that satisfactory in any case. Another technique is to take two exposures, but shift the camera by a set amount horizontally. This technique works well for still-life or macro photography, but obviously can’t be used for anything that moves.

What I’ve been looking to put together is essentially two separate optical systems that can be made sufficiently “like” each other and that can be triggered for close-to-simultaneous exposure, so I can capture stereo photographs of just about anything.

For the physical rig, the cameras have to be put baseplate-to-baseplate to minimize the distance between the centers of the lenses. I have a prototype rig made out of a couple of camera flash mounts, which at basis are flat metal bars with a slot in the middle where a 1/4-20 thread screw can go to attach the camera by its tripod mount. I’ve used a couple of 1/4-20 screws as standoffs to hold these near each other. This ends up being a bit thicker than ideal, but I have some ideas on minimizing that distance later. When taking stereo photos with longer than human inter-ocular distance between the lenses, I expect the brain of the viewer to interpret the things in the photo as either somewhat closer or somewhat smaller than what was actually the case. But it shouldn’t (and doesn’t) prevent fusion of the images.

The simultaneous triggering I think I have more-or-less solved with a pair of Pixel Oppilas wireless trigger receivers set to the same channel. I can use one transmitter to fire both cameras.

The optical systems are not quite identical. I’m using a Rokinon 12mm f/2 lens on one camera, and an Olympus 12-50mm zoom lens on the other. At least judging by merging pairs displayed on the back LCDs of the cameras, this looks like it may be close enough.

I was curious about the potential for using flash with the rig. Initial tests look pretty decent. I have a FlashPoint R2 flash on an extension cable attached to one of the cameras. I put both cameras in manual mode and use a relatively long shutter speed (1/20 to 1/30th of a second), and the few test shots I’ve taken have had the flash show up in both images. (More recently, I’ve gotten some failures, so there’s still some experimentation in order on this.)

While the whole rig is cumbersome, it provides something I’ve been looking to implement for some decades now. One of the things I like about this rig is that I’m able to use wide-angle lenses and get stereo pairs. In general, consumer stereo photography has largely been restricted to moderate wide to normal focal lengths on optics. I’m using 12mm on this, which gives an equivalent view of about 24mm lenses on full-frame 35mm.

On with the examples. These were put together with the assistance of a nifty Windows application called Stereo Photo Maker. It allows importing a left/right pair and provides various tools to help crop, align, correct, and export in various formats, including the stereo pair format I’m using here.

Viewed 34937 times by 1783 viewers