JOHN'S 3D GUIDE

Back

 

CHAPTER 4:    SPECIALIZED BEAMSPLITTER TWINCAMS FOR MACRO 3D

Closeup photography in stereo requires special techniques and equipment.  The problem with macro is simply that in order to satisfy the 1/30 guideline you have to put the cameras (or L and R image axes) very close together. 

For example if a mushroom is 12 inches away the camera separation should be about 1/3 inch (assuming a normal 50mm lens).  For a still-life the single camera slide-bar is perfect.  Just shift the camera 1/3" sideways for the second shot.  If using a flash attached to the camera, you do need to worry that as the camera slides the illumination of the scene may vary.  For macro work where the separation is generally less than an inch (for objects closer than 30 inches).  Thus, the camera translation is much smaller than the width of the flash head and so this flash shift is not a big problem.  While the slide-bar technique works fine for non-moving objects, if flowers are blowing in the wind, bugs are moving, or birds are flying, you must use a method that can handle a small stereo base separation.  Various techniques are reviewed on our gallery site.

Fig. 4.1. Action 3D Macro

A LR pair of a female broadtail hummingbird.  Shot with flash.  Distance to the subject was about 12 inches, requiring a camera separation of about 1/3 inch.  The flash units were triggered by the right camera.  The left camera has to have a fully open shutter when the flash goes off.  Fresh batteries and 1/30 or 1/60 shutter speed (no higher) were required with the twinned Pentax's.

We built two different units to do this sort of thing (initially for fluids research, but they turned out to be more generally useful).  Both use the basic idea of placing a 50-50 (equal transmission and reflection) beamsplitter in front of the lenses.  This is shown schematically below.

Fig. 4.2  Macro 3D

Two cameras are mounted in a perpendicular orientation.  A 50% transmission, 50% reflection planar beam splitter is mounted at the intersection of the central rays.  The size of the splitter plate is determined by the focal lengths of the lenses you wish to use.  A large plate is required for wide angles.  

 

  
Fig. 4.3a  Realization of figure 4.2.  The twincam is constructed out of 1" aluminum channel.  Fine adjustment screws are located on the underside of the beamsplitter bracket. Fig. 4.3b  Top view, showing a slight offset than can be achieved in this unit.  Generally the stereo base = 0" to 3" .

Each camera is mounted on a slider channel as in chapter 3.  The channels are incorporated into a strong square bracket that has a foot sticking out on which the splitter is mounted.  The hard part is getting the two cameras to point exactly perpendicular and on center, and for the beam splitter mid-point to be precisely at the intersection of the central rays as shown in figure 4.1.  Considerable tinkering and shimming was required.  Once aligned this unit functions quite well.  Some important considerations:

1)  Use black non-reflective felt to insure that the cameras do not receive stray light from the bottom surface under the splitter plate.

2)  Light loss will be at least 1 stop (since the incoming beam is split into two).  Plan flash strength accordingly.

3)  This unit is bigger than it needs to be.  I had thought that 28mm lenses might be useful so the plate I used had to be pretty big.  To work with just 50mm macro lenses, the plate could be half as big in each dimension.

In some situations it is useful to have more distance between the subject and the camera.  Flowers don't particularly care, but fluids and hummingbirds, for example, do.  Therefore we built a second unit using 135mm internal focus telephotos.  Because the telephotos have narrow viewing angles, a smaller beam splitter, just about as high as the lens diameter is sufficient.  This rig is illustrated below.

 
Fig. 4.4a   Top view of a telephoto macro rig.  Cameras are hung off a flat plate using angle brackets that allow for horizontal shifts.  The beam splitter is attached in a housing.  The red line shows its position. Fig.  4.4b   Looking into the tele-macro TwinCam.  The beamsplitter has to be wide enough to accommodate sliding the rear camera to the left about 1 inch.

With 135mm lenses the required stereo base separations are small indeed.  For an object 20 inches away I use about 1/4 inch.  Careful alignment of everything is critical, and errors are tolerated less than in the system of figure 4.3.  

This chapter has suggested some ideas on how to do stereo twin-camera macro-photography of moving subjects.  This is a pretty technical area, but entirely feasible if you have some metalworking skills and good patience with calibration and shimming.  A better setup than those above would use 80mm lenses only.  This focal length is a good compromise between working distance (+) and telephoto depth compression (-).  Such lenses would require a smaller plate, perhaps 51mm x 76mm, and objects can be nearer to the macro rig if the plate is smaller.  Beam-splitters can be obtained from Edmund Scientific (e.g. part number L31-433, $43) and other precision glass optics suppliers.   Large splitters can be custom made at (somewhat) reasonable prices by Precision Glass Optics.

 

Return to 3D GUIDE