Lessons Learned

By Joseph Holmes Back to

00182

Starting sometime in the mid-1990’s, it became clear that replacing color film with some form of digital capture was inevitable, but the wait for a workable replacement for my beloved Technika and its 4×5 film for scanning was a long one. About five years ago I met a clever fellow, Roger Howard, who turned me on to VR images built by stitching many frames from multiple rows of 360 degree captures from a Canon 20D. My gears started to turn.

I wasn’t interested in 3D spherical results for my own work, but the prospect of quickly making, say, nine exposures (three rows of three) with a Canon 5D, to subsequently stitch together into a single image of roughly 4×5 detail was more than a little intriguing. In early 1996 I still had never seen a properly executed 39 MP capture from one of the new backs, so that option seemed yet inferior to 4×5 film in terms of sheer detail— always an obsession for some kinds of photographers. The costly backs offered the best digital quality around which didn’t involve the cumbersome system required for a scanning back capture, but apparently not quite good enough (subsequent discoveries would change that perception).

So I put together a Canon outfit with a 5D, two zooms and three tilt lenses and went out traveling a few times to see what I could see. A favorite from among the earliest results is Giants of the Owens Valley, California, 2006.

Land of the Navajo, Arizona, 2006

It didn’t take long to realize that setting up additional rows was hard, but making more exposures within a given row was easy, so my plan changed to using either one or two rows of vertical captures when stitching was indicated. That in turn led to preferring the higher detail of a Mamiya 645 body with a P45+ back so as to keep it to single row capture when feasible, and still wind up with enough detail to make a sharp print of a horizontal image up to somewhere between 30 and 50 inches tall, by a larger width.

Single-row captures with that first Canon 5D were no slouch in terms of detail either, and the speed capabilities are great. Double-row captures are slow and quite limited by wind, movement and changing light, but when they work, they can be spectacular, for example, equivalent to three full frames of 4×5 color or more in a best case. But beyond the expected detail there were exciting properties of this approach that I had not anticipated:

Giants of the Owens Valley, California, 2006

1) The freedom of aspect ratio. I had never realized that the reason I hadn’t wanted to work with different image shapes as a rule was only to do with the hardware limitations. There were simply no solutions to the camera problem. When you’re adding frames to a single or double-row digital capture sequence, all you need to do to change the shape to a wider one is to add more frames. This adds detail, rather than taking it away. So my life has gotten far more complicated vis-a-vis the requirements of presenting my work, as it now comes in a large number of aspect ratios. In the near future my new web site will finally go live and a wide variety of shapes will be shown, each presented in an optimal way on the various web pages.

2) Including more subject matter in the final design than I had thought I wanted. My compositions have typically been thoroughly worked out behind the cam- era, so it was an eye-opener to find that in quite a few cases I preferred a wider view which the stitched pano made possible (by my having shot more at the sides if not also at the top and bottom) to the original design concept. These wider views were inevitably more complex, and I’ve especially enjoyed the added layer of complexity that this outcome has delivered. My photograph Aspens, Eastern Sierra, California, 2006 is an example of this.

pacific
Sunrise at the Pacific, Oregon, 2007

3) The joys of cylindrical rendering. When stitching, the two most likely geometries to impose on the result are the conventional rectilinear one (same geometry as a single-frame capture) and the cylindrical one, resulting in an image like one which you would get if you looked at the subject through a rotating vertical slit. This one is particularly useful for wider angles, like 90 degrees and more. As long as the camera is rotated around the vertical axis, the horizon will remain straight. Very wide views can be captured without the corners being enlarged and distorted, but horizontal lines other than the horizon will be curved.

Nobel Cabernet, Sonoma County, California, 2007

4) The ability to adjust the white point during the RAW conversion only, with total freedom from color crossovers. For the first time sensors have made the rendering of a perfect grayscale in color photography possible. Often this precision doesn’t play a very strong role in a picture, but sometimes it decidedly does, and I have found (as I think many have) that this precision of gray balance can lend to an image a wonderful sense of solidity, which we might never have realized was lacking had we not seen it. André Oldani, (one of the three ‘ALPAs’) made a picture of an experimental white Porsche against a gray background which exhibits this quality more plainly than any other image I’ve seen.

Looking at only the front end of the car, one could easily believe that it was a monochrome image. Although it seems ironic, a solid gray backbone can be a precious asset for a color image.

The aspen photograph actually illustrates a vital benefit of the post-exposure controllability of the white point. The primary light source was very cool blue sky, but by neutralizing it completely, I was able to bring the image to life and reveal the essential character of the trees. At other times I’ve needed ambient blue to counterbalance the yellow of aspens, but not here.

I’ve not used my view camera for over four years. The shift to two digital outfits has changed many aspects of my work that I’ve not yet mentioned. I’ve always found that the camera system (camera, film or sensor, rendering methods) dictate the envelope and that the envelope, in turn, almost automatically limits my seeing. I adapt seamlessly to what I know the camera can do. Some kinds of images are now gone, and some kinds have arrived anew. I’m much more adept at working with skies now, but less inclined to look for interesting close foreground elements in a vertical composition especially. I still don’t know just how I’ll feel about the sum of what has emerged from these last five years work, compared with my earlier results as time goes by, but it does work, I am happy with the best of it, and I do so love playing with the sky!

Independence Celebrated

Resources Cameras: Phase One – phaseone.com, Alpa – alpa.ch; Software: (Image Stitching) PTGui Pro 9 – ptgui.com, Adobe Photoshop – adobe.com


About the Author

Joseph Holmes
00206
Joseph Holmes, a native of Berkeley, California, holds a degree in the Conservation of Natural Resources from UC Berkeley. His photographs have been published in the series of Last Wildlands Calendars for Friends of the Earth and several books, including Joseph Holmes, Natural Light, Canyons of the Colorado. He has been designing RGB working spaces since 1997 and is the author of ColorBlind Prove It!, software for monitor calibration. Along the way he mastered silver gelatin black & white, dye transfer, additive Cibachrome, pigment transfer and finally digital printmaking. josephholmes.com