in Uncategorized, Arts and Photography

Rephotography of Historical Sites

            This is a guide to rephotography which I found online a few years ago. Now the link is down and I can't find a new link for it. So I am saving some of it here.The photographs which were on the site are not working now, when I view it with the <a href="https://web.archive.org/web/20170824204058/http://home.onemain.com/~home_range/doc_phot/index.htm#contents">Wayback Machine</a>.

AN INTRODUCTION TO DOCUMENTARY PHOTOGRAPHY

A GUIDE AND FIELD HANDBOOK

Gerald T. Davidson

4. Rephotography of Historical Sites

Rephotography evolved into a distinct discipline of history and environmental studies during the second half of the 20th century, as old photographs were rediscovered and people began to wonder how the landscape had changed over the past several hundred years. It was exciting, but no great surprise, to see just how the man-made landscape had changed in some places of western North America. But it was a surprise to see—in a way that could be accurately documented—just how much the natural landscape had changed.

The two photographs on the cover are both of the same spot, taken almost 100 years apart in time. The location is the site of a coal mine at Red Lodge, Montana. The first shot was made early in the twentieth century, and shows a typical turn-of-the century industrial waste-land-scape. The second shot was made at the end of the century. Almost every trace of the mine is now gone. The large building has been replaced by the Senior Citizens’ Center. The railroad yards and sidings were long ago abandoned by the railway; the green in the foreground is an athletic practice field. About the only landmark that can be recognized in both photographs—apart from the distant mountains—is the tiny house, just to the right of the center, almost hidden by the trees and shrubs.

A.   Uses of Historical Rephotography

History is about loss, loss of a past that can never be recovered.  However the past has given us many artifacts—including paintings, drawings, and photographs—that tell us something about the way things were.  The vanishing way of life of the native peoples of North America was among the first applications of what we now call documentary photography.   Edward S. Curtis was only following in the footsteps of painters like George Catlin when he set out in the early 20th century to make hundreds of photographs of the Indian ways, as he saw them.  This may be considered an example of re-photography, where the originals were paintings and drawings.

Edward S. Curtis’ photographs can be most easily appreciated in the numerous books devoted to reproductions of his work.  The photographs are mostly in the public domain, so there have been many such books, some better than others.

In the instance of Curtis’ photographs the subjects are long gone, and only their landscape remains.   That landscape is interesting for later photographers, particularly when it contains more-or-less durable works of man, such as buildings.   Photographers can document what has been lost in the landscape, environment, and ways of life by the photography of modern subjects.

Documentary photography serves history well when it concentrates on man’s environment.   Changes in our buildings, our transportation systems, and all the artifacts around us tell much about how our customs and ways have changed.   This is particularly apt in urban landscapes, where even the building styles have a story to tell.

A very effective use of rephotography is in the application to ecology, the natural landscape, and climate change.  The disappearance of the free ranging bison is a well known part of the history of the American West.  But there have been many other, sometimes quite subtle changes in the flora and fauna of America.  Of course some of the changes in the landscape and natural vegetation could be attributed to the actions of man, through suppression of wildfires, elimination of animal species, damming of rivers, and spread of agriculture and non-native plants.   Nature also had a trick of its own.   We now know that the 19th century was subject to some wild climatic excursions, including the end of the Little Ice Age and at least one great drought at the mid century.  One fact stands out: almost everywhere in western North America there are more trees and shrubs than 150 years ago.

There are many uses uses of rephotography.  Among the most important are

  • Changes in the landscape, due to man’s activities
  • Changes in cultures, customs and ways of living
  • Changes in the environment and vegetation
  • Climate change
  • Changes in the urban landscape

One can practice rephotography in an entirely ad-hoc way, using the simplest of cameras, or one can employ all the techniques available to the photographer and to the historian.

The techniques of historical rephotography are simple to describe; but in practice they often require much labor and lucky guesswork.   Finding the general location of an old photograph may require extensive library research, to establish its date, origin, and probable location.   That may narrow the search down to a few square miles.   Further narrowing the search, especially in rough, broken country, may demand days of tramping over every part of those few square miles.   Clearly rephotography is only valuable and useful if some important results are expected.

Klett et al tell about some of the difficulties they encountered in The Rephotographic Survey, including vegetation that had grown so dense that the original site was almost inaccessible.  Dense vegetation proved a problem also in some of the examples shown here, where it was quite impossible to set up the camera precisely on the site of an earlier photograph.   They describe some of the strategies they used to locate the spot where the original photograph was made, including elaborate triangulation.  Sometimes there may be only one tiny spot where the photograph could have been made.   Usually the search must be narrowed by triangulation of objects and prominent landmarks.   Several practitioners have devised methods whereby the viewpoint might be determined in the field, by the use of cut-out templates and other visual aids to triangulation.

A problem that has not always been recognized is the possibility that many old photographs and prints may be distorted.   Many 19th century camera lenses had an unknown amount of image distortion.  More likely is distortion of prints caused by faulty processing and the effects of aging.  When possible it is best to work from glass-plate negatives or make new prints from the glass negatives, which are very dimensionally stable.  Distortion of the originals is usually revealed after the new photographs are made, and carefully compared with the old.  The problem is not usually a major difficulty, because most outdoor photographs in the 19th century were made with relatively small lenses at narrow apertures, which lessens the effects of image distortion.

Since we expect to make digital photographs when we have found the site, it is appropriate that digital processing of photographic images be used to aid the location process.  One useful tool is a transparency that can be held up against the scene.  Such a transparency should, of course, be large enough to be handled easily, and it must indicate the most important features in the original photograph.  So, we make a digital copy of the original.   The more sophisticated image processing software contains one or both of two special tools, generally called “edge detection” and “high pass filter”.   Either of these washes much of the information out of the photograph, leaving an enhanced trace of the most sharply defined features.  The example here shows the result of such image processing, applied to a simple photograph of a mountainside and some trees in the foreground.   The trial photograph, which represents an copy of an old photograph, was made at a normal focal length.

B. Useful Techniques: Finding the Location

The techniques of historical rephotography are simple to describe; but in practice they often require much labor and lucky guesswork. Finding the general location of an old photograph may require extensive library research, to establish its date, origin, and probable location. That may narrow the search down to a few square miles. Further narrowing the search, especially in rough, broken country, may demand days of tramping over every part of those few square miles. Clearly rephotography is only valuable and useful if some important results are expected.

Klett et al tell about some of the difficulties they encountered in The Rephotographic Survey, including vegetation that had grown so dense that the original site was almost inaccessible. Dense vegetation proved a problem also in some of the examples shown here, where it was quite impossible to set up the camera precisely on the site of an earlier photograph. They describe some of the strategies they used to locate the spot where the original photograph was made, including elaborate triangulation. Sometimes there may be only one tiny spot where the photograph could have been made. Usually the search must be narrowed by triangulation of objects and prominent landmarks. Several practitioners have devised methods whereby the viewpoint might be determined in the field, by the use of cut-out templates and other visual aids to triangulation.

A problem that has not always been recognized is the possibility that many old photographs and prints may be distorted. Many 19th century camera lenses had an unknown amount of image distortion. More likely is distortion of prints caused by faulty processing and the effects of aging. When possible it is best to work from glass-plate negatives or make new prints from the glass negatives, which are very dimensionally stable. Distortion of the originals is usually revealed after the new photographs are made, and carefully compared with the old. The problem is not usually a major difficulty, because most outdoor photographs in the 19th century were made with relatively small lenses at narrow apertures, which lessens the effects of image distortion.

Since we expect to make digital photographs when we have found the site, it is appropriate that digital processing of photographic images be used to aid the location process. One useful tool is a transparency that can be held up against the scene. Such a transparency should, of course, be large enough to be handled easily, and it must indicate the most important features in the original photograph. So, we make a digital copy of the original. The more sophisticated image processing software contains one or both of two special tools, generally called “edge detection” and “high pass filter”. Either of these washes much of the information out of the photograph, leaving an enhanced trace of the most sharply defined features. The example here shows the result of such image processing, applied to a simple photograph of a mountainside and some trees in the foreground. The trial photograph, which represents an copy of an old photograph, was made at a normal focal length.

The features of the landscape are recognizable, but the image is neither very attractive nor very useful. What we want to do is enhance the contrast so that this image can be used as a template for triangulation of the site.

The second example is the same image, to which has been applied a process called “posterization”, which reduces the number of gray levels, in this case to 5. Then the contrast has been increased so the image looks like a line drawing.

The final effect is rather artistic, but art is not the intent. An image such as this can be printed on a sheet of 8.5 x 11 inch transparent foil, such as used for projection transparencies. If the contrast is sufficient, the transparent image is opaque only at the edges of objects. It can then be held at arms length while moving about until the image and the scene match. In a trial using the image shown here, it was possible to re-locate the position from which the original photograph was made within about 10 feet. This is very precise, considering that the mountain ridges are about 2 to 4 miles away and the large tree near the center is about 400 feet away.

This method could be greatly improved by improvising a metal or wood frame to hold the template at a distance such that it subtends the same field of view as the original. But, since the original field of view may not be known, the frame would need some sort of sliding adjustment to change the distance from the eye to the template.

Once the location has been established within 10 or 20 feet, digital photography can make the final step of the process easier by enabling a series of test photographs, until the exact point has been found. This part of the process is largely trial and error, much as an artillery gunner might find his target by first aiming too high, then too low, and estimating the intermediate setting.