Expert Systems
by Galen Rowell
Outdoor Photographer, September 2001

Throughout history, professionals in every human endeavor used to get along fine without direct knowledge of what industrial psychologists now call expert systems. Times have changed. In the old days of long apprenticeships before venturing out on one’s own, expert systems became ingrained into the behavior patterns of every skilled artist or craftsman. Today, photographers who buy expensive equipment with automated features expect to rapidly match the aesthetic results of photographers of old. After all, they had such primitive equipment, it’s a wonder they could accomplish anything at all.

Riding a bicycle is the classic example of a skill that a person never forgets. Actually, the art of hopping on a bicycle and riding it away without loosing your balance is actually a series of expert maneuvers. Taken individually, they’re simple to perform, but the rapid sequence of standing on a pedal with one foot, getting the bike rolling, swinging the other leg over the bar and setting it onto the opposite pedal without falling over doesn’t come naturally. I never saw a kid jump on a bicycle for the first time and ride away smoothly, nor have I seen a photographer pick up a brand new camera and successfully shoot a fast moving situation, whether it be a landscape of light and form or a moving subject, without screwing up at first.

Advertisements imply that clever, automated features guarantee great results. For average situations they do a passable job. Relying on them, however, breeds an absence of the expert systems needed to handle complex new situations, those rare moments that offer the finest opportunities for photography.

Automated features are at their best in the highly controlled situations that sports and wildlife photographers experience when shooting with long lenses from the sidelines or a blind. As observers, not participants, their creative choices are brought to the forefront by equipment that takes care of focus tracking, exposure, and bracketing. In these fields, the quality of results has gone up tremendously over the past decade due to advances in equipment. In nature photography, however, fine 1970s images hold their own. Both street and nature photographers who relied on expert systems to narrow the need for conscious thought about camera settings and lens choices consistently outperformed early users of automated features in the eighties and nineties.

Why should this be true if these features also free the mind of technical concerns and allow one to concentrate more on creative choices?

A veteran columnist for now-defunct Modern Photography magazine hit the nail on the head about twenty years ago when she commented that her first trials with her new automated camera gave her consistently sharp and well-exposed pictures that simply weren’t as interesting or provocative as those she had made in the past with her trusty manual camera. She didn’t know why.

Expert systems narrow the search zone toward creative choices in a very different way than automated features. Multi-segment metering attempts to do the best.

This in no way means that I switch to manual when something is happening fast. Quite the opposite. I leave my auto-exposure and auto-focus on unless I make a conscious decision to switch to manual settings. However, I don’t unconsciously except the built-in sequence of automatic camera operations anymore than I always used the same sequence of manual settings in the days before automatic cameras.

The use of expert systems involves combining memorized skilled behavior patterns that are executed as unconsciously as walking across the room and opening the refrigerator with conscious choices that govern the artistic and creative outcome of a photograph. Photography can be extremely fast moving and complicated, even when the goal may be a seemingly static landscape photograph. To make conscious choices about everything is to loose the moment and watch the shadows fall, the bird fly away or the wind blow the flowers before you have a chance to click the shutter. Clicking the shutter prematurely creates a bad picture because the necessary creative steps weren’t performed.

That buzzword of modern life, “multi-tasking,” is something of a misnomer. Our brains can’t multi-task. They simply can’t pay attention to more than one task at a time. Psychologists suspected this after extensive tests with illusory art forms that indicated that a person could see one concept or another, but never both at the same time. The classic sketch of a vase that can also be seen as two faces in profile is a great example. Neuroscientists have recently done brain scans of subjects who verbally report whether they are observing the vase or the faces at the exact moments that different areas of their brains light up in the imagery. What this means is that regardless of our high regard for our personal ability to multi-task with the best of them, we are really only capable of performing one creative choice at a time.

If we use that moment to do something mundane, such as thinking about how to turn the manual focus on that zoom lens without grabbing the zoom control and changing the focal length, then you may have lost the moment in which you would have sensed just where to compose that wild animal walking across a clearing for a matter of seconds. The extra choices and complexities of zoom lenses have often caused me to loose the moment in fine landscape photography. For example, I might get lazy and carry only a wide-angle zoom to the shore of Mono Lake to catch the sunrise, and then I set up on a tripod with the proper choice of graduated filter positioned in a Cokin holder just so across the horizon, but when the sun peaks up and I have the chance to have diffraction beams without lens flares coming off the horizon, I push the shutter release decisively only to have the auto-focus start searching in the new area of glare and the graduated filter tip to the side. With moments to spare, I need to know whether it’s best with this particular lens and camera combination to try using the auto-focus and to regain the setting and filter position I had, or whether to instantly revert to manual. More often than not, I loose the picture.

Had I brought the proper focal length of fixed lens in the first place, everything would have gone like clockwork in a sequence of conscious choices mixed with behaviors wired into my motor nerves. Give me a camera and a couple of lenses I know well, and I can be shooting a completely novel situation in less than thirty seconds after opening my camera bag, including mounting the body on a tripod and setting the proper graduated filter in just the right position. Throw me a curve with new equipment or even my favorite stuff packed in the wrong order in a side pocket, and I’m as confounded as a Golden Retriever thrown six tennis balls at once. My mind blocks up and the sequence stops until I figure out what went wrong.

The “Wild Iris” photograph is an example of the successful use of conscious expert systems. In other words, I made a point of thinking about my options and my sequence of operations before deciding on the final image, combining the process with procedures as wired into my motor nerves as hopping onto a bicycle. Photographers who use expert systems well walk decisively across the landscape to the positions where they take their photographs, while their less accomplished counterparts often wander aimlessly waiting for their eye to randomly catch something worth photographing. I began with reasoning that the most evocative photograph of wild iris bluing beneath the Eastern Sierra would take advantage of the morning alpenglow on the peaks for which the region is justly famous. But to do so would mean a radically different exposure for the iris in deep shadow. I reasoned that a soft-edged graduated filter wouldn’t do the job, and that, depending on the brightness of the sunrise, I would need a 3-, 4- or 5-stop hard-edged graduated filter. If I used such an extreme filter, I would need to have a very straight, unobstructed horizon to place the edge. If the iris themselves or any tree or rock rose beyond the level of the edge of light on the peaks, it would create an unacceptable dark stripe with an open exposure below.

Thus I created a mental template for my search. I would only consider unobstructed horizons with vast fields of iris. In order to encompass the flowers and the peaks while holding great depth of field, I needed a wide-angle, but not too extreme a wide-angle, because the peaks would become less prominent on the horizon. So I walked confidently from my car with my camera already on a tripod set up with a 24mm fixed lens and graduated filter holder. The next order of the day was to find a unique composition. Just any random group of blooming iris wouldn’t do. From experience, I knew that the flowers nearest the foreground edge of the frame would take on a greatly increased importance in the two dimensions of film, and that I needed to carefully select an aesthetic scene that singled out something special in that area.

Knowing the exact time of the sunrise, I had driven out with just five minutes to spare. As the first rays touched the peaks, I turned on my spot meter to compare that exposure with one I had already taken from the unchanging light in the foreground shadows. The difference was 4-stops. I slipped on the special filter that I had had custom made by Singh-Ray and lined it up with the edge of crimson light, favoring a slight darkening of the shadowed area, rather than any possibility of letting a hint of overly bright light slip out at the horizon, which is visually quite objectionable. After bracketing three exposures on the meter and 2/3rds of a stop up and down, I knew I had a good image in the bag, but I asked myself, could I do better? Maybe.

With the depth of field preview held down to my picture-taking aperture of ƒ16, I reset the filter just on the upper edge of the light, a gamble that in the past had resulted in as many blown out horizons as seamless melds of light and shadow. Only days later, when I checked my film on a light box, did I know that my gamble had paid off. The 4-stop extreme filter had formed a virtually seamless merger, even though I hadn’t been sure looking through the viewfinder. Why? Because the greatly reduced 8-to-1 acceptable brightness range of slide film compared to the 2,000-to-1-brightness range of the human eye greatly reduces a person’s ability to judge how a filter will appear in the final image.

One of the pitfalls of becoming an “expert” photographer is the false assumption that you know every thing so well that you’re going to get it right the first time. It’s an attitude that comes naturally with expertise, but the better you are, the less likely that your first shot will be your best shot. Working through an idea creatively and exploring the options may deliver an image that is both technically and aesthetically superior to that first go. I know this is true when I print some of my work from thirty years ago and wish that I had used my present conscious awareness of expert systems to expand my horizons and create more viable options, while at the same time limiting my visual search of the natural world to situations most likely to come across in the differing visual language of film.

Galen’s columns are reproduced on our web site courtesy Outdoor Photographer.
106 South Main Street :: Bishop, CA 93514 :: 760.873.7700 :: fax: 760.873.8229
Licensing: mlpictures@mountainlight.com :: Orders & Info: gallery@mountainlight.com
All images and content are © Galen Rowell / Mountain Light Photography, Inc., unless otherwise noted. All rights reserved. No form of reproduction, including copying or saving of digital image files, or the alteration or manipulation of said image files, is authorized unless accompanied by a written license issued by Mountain Light Photography, Inc. For information regarding commercial or personal uses, please contact: Mountain Light Photography, 106 South Main Street, Bishop, California 93514. Telephone 760-873-7700, or email us at mlpictures@mountainlight.com. All prices subject to change without prior notification. Original web site design by Barbara Cushman Rowell.