Remember: you can click on any of the image to open a full screen version.
In the first installment of this 2nd series on creating a camera system that allows the use of a DSLR as a capture device combined with optical movements for depth of field and distortion control, I began to experiment (and learn) about the Cambo Ultima 35/D system view camera by using simple depth of field correction. THat first test was a total success.
What this system consists of is a view camera base that is VERY high precision plus a means to attach a 35mm or DSLR camera as the rear standard, a bag bellows between the two to allow for the close positioning required of wide angle lenses, and a front standard for use with large format or medium format lenses. For a grand total (The Ultima is a modular system so you buy all the parts a la carte) of about $6219 (plus DSLR and lenses) you have converted your DSLR into a true view camera with complete front and rear optical movements.
With the latest models of high resolution DSLRs this becomes a workable rig for all but the most high-resolution needs where a medium format back or large format scanning back is all that will suffice. Cheap? No! But the fees for doing the sort of work this will potentially allow you to accomplish would make it worthwhile. And besides, this is not a cheaply built rig either, it truly is an impressive bit of engineering and manufacturing by this Dutch company. The final and most serious question to be answered however is whether or not it is worth it, compared to the costs and ease of use of options such as using tilt/shift lenses and the somewhat less expensive option of putting a DSLR adapter plate on a normal view or technical camera.
The answer is that it really depends on what you will be shooting, where you will be shooting, your shooting style, and, to be brutally, materialistically frank, whether or not you can make a “business case” for it when factoring in the level of client expectations and payment for your work. As with nearly ALL things photographic, there is no “one-size-fits-all” solution and despite the best efforts of salesmen to convince you otherwise, sometimes the fastest, latest, most expensive item is NOT the best for your personal needs.
Of course if money is no object, then rush right out and get one; at the very least it will blow away the others on your next photo trip especially when you introduce them to the Sherpa needed to haul it around. I’ll wait until the tests are over and I’ve used this setup in several scenarios before presenting my own conclusions.
After the first test (Part 1) using front movements (swings) for depth of field adjustments I decided to do two more test set-ups. One would be an architectural shot to play with distortion control to solve keystoning/foreshortening typical of that type of work with a fixed lens camera, and the other would be a small product shot where complex or compound movements using both front and read standards might be required.
If you recall in the last post I was not sure how to make the camera shoot in a vertical position. Turns out it needs an additional part. After scrounging through the stuff in the check-out locker at school we discovered that the answer to shooting vertically is a proprietary designed “L” bracket which had been left in one of the parts boxes. So with that in hand go from horizontal to vertical with the DSLR.
To be sure, it is not as easy as the Fotodiox back (mentioned in the first part) for the 4×5 view/technical camera. That clever device simply allows you to turn the camera 90 degrees while attached, (or of course you could simply rotate the back). But still, the Ultima’s transfer is not all THAT difficult. To switch from horizontal to vertical position you unhook the bellows from the back of the front standard (simple sliding lock), dismount the camera from the mounting block, attach the L bracket and camera, reattach the bellows and then mount the L bracket back on the mounting block.
Like I say, it’s not instantaneous… but not all that difficult either at least in a studio. Whether or not it would be so easy out in the weather with cold hands is another issue… But there was an issue I had not anticipated but which makes this rig of less value to me. Take a look at the photo below and note the shift position.
Look at the indicators for left shift; this position was necessitated because the supplied “L” bracket placed the camera too far to the right. A full left shift centered it and made it ready to shoot. But only if you were going for a single frame and did not need any further left shift was it OK. One could always drill and tap a new hole to better center the camera on the mounting block. But, I’m sorry, i think a $6,000 camera rig ought to have holes drilled in the right spots! I also needed to lower the rear mounting block but that would have no real effect on 99% of any normal movements needs.
Unfortunately, with it at full left shift just to center it, my plan for using the rear shifts to mimic the sliding DSLR adapter I have for the 4×5 is effectively eliminated. For most shooters, relying on a single frame for everything is hardly a problem. But for me, the idea of stitching 3 vertical frames into a nearly 50 megapixel file (or, more with one of the latest generation 20+ megapixel DSLR sensors) was attractive because it virtually obviated the need for a medium format back.
Anyway… If you have used view cameras before using the Ultima you will find it not only easy but it is a delight. Almost everything is smoothly geared and movements take place on the optical Axes so that you do not lose focus. This is just like the high end view cameras from Sinar, Linhof, Arca-Swiss, etc.
However, if you have never used a view camera before, there is no way around the fact that you will have a steep learning curve. On my website’s school page, found at (www.ndavidking.com/sdccd.htm), in the Photo 240 class section are several handouts to get you started learning how they operate and why. But, even for old timey experienced large format shooters, it is a new head-trip entirely trying to compose and focus through the DSLR’s little viewfinder instead of using the wonderful large ground glass of the 4×5, 5×7, and 8×10 cameras. The good news is that you don’t need a dark cloth.
Shooting with a view camera of ANY sort requires far more deliberation than simply running around in “spray and pray” mode. Now, with nearly complete optical control of the image you really need to think in terms of MAKING a photo rather than just TAKING one. Plus of course, there is the set up time to prep for a shot which allows and requires some forethought.
Perhaps because most of my career was involved with view cameras that slower approach to shooting is ingrained in me so I tend to be pretty deliberative even with a hand held DSLR. But if you are used to filling card after card with every possible optional view, angle, exposure, whatever, then this will slow you right down and force you to think first. I think that is a good thing but it would drive some photographers crazy.
San Diego has some interesting architecture to play with so after a little bit of location scouting I settled on The corner of the Hilton Hotel on Mission Bay for this test shot. As much a resort as a hotel this lodging facilities is in a prime spot. But more to the objective of this test, it has a lot of vertical lines I could use to determine whether or not the camera could effectively straighten them.
Here is a shot of the camera movements all in zero position and looking up at the building.
And here is the shot from the camera with no movements employed. You can already see the converging lines due to perspective. The top of the building is farther away from the image plane than the lower floors so they are smaller resulting in the edges of the building leaning slightly inward. What you can also begin to see is a slight vignette on the right side of the frame (remember image on sensor is upside down and backwards) from the far left shift.
Right away a problem emerged. The medium format Hasselblad lenses are designed to project an image circle sufficient to cover a 2.25 x 2.25 (6cm x 6cm) square negative with very little additional coverage area. To set the movements properly for this shot moved the image area OUT of the available circle and so cut it off. And a taller building would have been simply impossible to deal with using this type of lens.
Now began a game of compromises. By dialing back the corrections and cheating a little with the front rise I was able to make a workable shot; but it has major issues. Though it is straight it is not properly focussed. I shot this one at f5.6 so you could still see the remaining focus issues. With full optical corrections, even if shooting wide open the plane of focus should cover the front of the building. But with the limitations of the covering power of the lens, searching a setting in the rises and tilts where the full hotel was in the frame and the edges were more or less straight, left me out of movements to fully bring the front standard to a setting parallel to the building front.
Now the building’s converging lines are straightened since the outer edges of the building in the frame are now parallel to the image plane due to the rear standard (the image plane) being parallel to the building. As a bonus, what should happen if BOTH standards are parallel to the building, because, as we all remember from high school geometry, parallel lines meet at infinity. Therefore we normally would have actually satisfied both Herr Scheimpflug and M. Carpentier ( the image plane, the lens plane, and the plane of focus all are parallel and therefore intersect at infinity) and the entire front of the building lies within the depth of field plane. But in this case making it all work was physically impossible and the plane of focus remained at an angle to the plane of the building. Stopping down to f22 brought the whole building in focus but it would have been at the expense of some softening due to defraction and that defeats the whole purpose so I’m showing it to you without stopping down so you can see the problem.
So test two was not a total success. Oh good grief, that is putting it kindly. I would have had to go back and get the view camera to do this shot properly. I was also disappointed at the need to shift the L bracket to center the camera not only because that disappointment was based on my potential use for the camera system, but also because it created the edge vignette as well.
The big lesson was, however, if I bought one of these for myself I would get the lensboards (or make them) designed for view camera lenses that project a coverage circle enough wider than even the 4×5 central image area to allow extensive movements in that format. And although it has a bag bellows on it, I would get recessed lens boards so there was a little space between the standards to allow all or most of the movements this rig is capable of. With a 4×5 view camera lens, bag bellows, and recessed lens boards you probably could not get enough movement out of the bellows and base to get you outside of that area of coverage.
So now after two tests, for me, the score is tied up, 1 and 1. For the final test in Part 3 I’ll do a small product that will demand more complex movements. And then be able to draw some conclusions about this option for creating a workable DSLR-based camera system with optical movements.