Accuracy

There are some interesting contradictions regarding “accuracy” in Pictorial Photography. There is an extensive discussion of pinholes in the chapter on lenses, and pinholes are favored for their infinite depth of field and focus, and “a very pleasing quality of definition, the amount of diffusion depending on the size of the hole” (50). Anderson describes his method for making pinholes:

In describing the method for making a pinhole the text-books usually give elaborate instructions for its manufacture, these directions looking towards an accurately gauged, sharp-edged hole. The writer finds, however, that for pictorial work it is sufficiently accurate to punch a hole with a pin or needle in a piece of black paper such as plates are wrapped in, as the purpose of the more accurate method is to approximate lens definition, which is precisely what the artist wishes to avoid. (51-52)

In this case “accuracy” has nothing to do with definition. The precise definition of what is meant by “pictorial” photography is ambiguous, and there are several divergent attributes applied to it throughout the course of the book. What I find most interesting is that the standards of “accuracy” are different between portraits and landscapes. In the chapter on choice of plates, Anderson takes a great deal of space to encourage photographers to use filters which promote “accurate” reproduction rather than false values—values are seen to be the key element of accuracy, rather than definition. Using an illustration comparing the responses of orthochromatic and panchromatic plates, Anderson chastises portraitists for using the inaccurate (and cheaper) ortho plates—though he grants that they are very useful for landscape work:

In portraiture, however, the case is somewhat different, and here the panchromatic plate is invariably superior to any other, for, though falsification of values is sometimes desirable in landscape for pictorial reasons, correct rendering is always desirable in portraiture. It is of course obvious that when it is said that the panchromatic plate is invariably superior to any other it is assumed that the photograph is to be made with white light or by some illumination containing a suitable proportion of yellow and red, for, naturally, if the mercury vapor arc is employed, a color sensitive plate presents no advantages over an ordinary, the light in question containing no rays except ultra-violet and violet. However, it is assumed that the reader knows better than to expect a correct rendering of colored objects by monochromatic light, since it makes no difference whether the plate is color blind or the object reflects no colored light. The skin of a healthy Caucasian is distinctly yellow, with an element of red in lips and cheeks, hair usually tends towards the warm brown, though in some individuals it may be yellow or red, eyes are usually of a bluish-gray, though sometimes brown, and skin blemishes are yellow or reddish. Wrinkles have a reddish tendency, due to small capillaries lying close to the surface, and the faint lines under the eyes are purplish in hue. Hence we may expect the ordinary orthochromatic plate to render the skin, hair, lips, shin blemishes, wrinkles, and lines under the eyes as too dark, and the eyes themselves as either too light or too dark. (63-64)

It is interesting to compare this with some later manuals from the 40s and 50s that recommend the use of orthochromatic plates for photographing men, rather than the (then standard—or ordinary) panchromatic plates. The promotion of darker, ruddier skin tones was never considered a good move for women—in either period. It is interesting that Anderson’s example of the “incorrect rendering” of ordinary or orthochromatic plates uses a woman.

A comparison of the rendering of the ordinary and the panchromatic is given in figure 36, which was made in an ordinary studio, A having been made on an ordinary plate, and B on a panchromatic plate with a ray filter. No retouching or modifying of any sort has been done on either of these negatives or prints, and attention should be called to the fact that B is an exact rendering of the sitter’s appearance. No ordinary or even orthochromatic plate can render correctly a subject containing so much yellow and red as the face, even with a ray filter, and if true values are to be obtained in portraiture a panchromatic plate and filter are absolutely necessary. It is proof of the remarkable complaisance and equally remarkable lack of observation on the part of the public that portraits made on non color-sensitive plates have been and still are being accepted by the sitters. In fact, it is doubtful if one-tenth of one percent of the professional portrait photographers in the country at the present time use panchromatic plates in regular studio work, whereas it may safely be said that if they realized the advantages of such emulsions all except the very lowest class would at once discard the ordinary and even the orthochromatic plate. (64-64)

There are several distinctions here worth pondering. First, Anderson presupposes that Caucasian skin tones are the most important—and second, he supposes that lower-class workers are insensitive to accuracy. However, the most important distinction I think is that portraits require a sort of tonal accuracy which is a completely separate issue from definition. Landscapes, on the other hand, are purely seen as evocative rather than as accurate renderings of a physical object.