25 February 2006

Double standards of scientific photography

In 1977, Alfred A. Blaker, a professional scientific photographer, wrote1:

"...errors do occur, and besides, it is not always possible to make a single exposure that will be optimal for the entire image. I refer to the solution of this problem as differential printing, though this concept is commonly thought of as two separate procedures called dodging and burning in. Dodging consists of interrupting the light between the lens and a small area of the printing paper so as to give less exposure to that area, whereas burning in is the opposite-interrupting the light to the entire paper except for a small area, so as to give more exposure to that area. These procedures are illustrated in Figure 6. [reproduced below]"

Instructions you will not need anymore. But note the sentence I underlined: "When correctly done, differential exposure is not detectable in the final print." Go explain that to the editors of Nature.

In our age of digital photography, however, the established techniques of scientific film photography that were once completely acceptable would get your manuscripts rejected. Item #5 in Nature's new "Provisional guide for digital images":

"Processing (such as changing brightness and contrast) is appropriate only when applied equally across the entire image and when it is applied equally to controls. Contrast should not be adjusted so that data disappear. Excessive manipulations, such as processing to emphasize one region in the image at the expense of others (for example, through the use of a biased choice of threshold settings), is inappropriate, as is emphasizing experimental data relative to the control."

Differential printing, I suppose even if it is applied equally to controls, is now inappropriate.

I recommend that Nature change the title of their guidelines to "Paranoid guide for digital images".

1. Alfred A. Blaker. 1977. Handbook for Scientific Photography. W.H. Freeman.


Duane said...

My own more serious work, which is of course, quite a bit different than yours, often requires that certain areas of the object under study be adjusted for brightness and contrast. As far as I can determine, the expectable standard is to indicate the nature of any "enhancements" and, if appropriate, offer the "unenhanced" version of the picture. There is almost no light source or direction of lighting at will show a clay tablet in sufficient detail to be able to read all four edges of the tablet and still have the flat surface legible. The ability to read the table from a photograph depends on the ability to see the shadows caused by the wedges with clearly. If you have the tablet in your hand you can rotate it to get provide the best lighting conditions. Of course, one can, and often does, take pictures from differing directions or under differing lighting conditions and "stitch" them together to make a composite photo. In this case, also, one is supposed to tell the reader what you did. Something like an ostraca is even more difficult - composite illumination frequencies over a fairly large spectrum (including x-ray and infrared) are often used.

I own take on the whole issue rests on the goal of the photograph. What is important is extent to which one tells one's readers what that goal is and the extent to which one offers enough information about what one did to meet that goal so they can make an independent judgment concerning the "validity" of the photo. If the goal is clear and the methodology is spelled out and is appropriate to the goal then almost anything is OK.

Of course, the whole thing may be different in the hard sciences.


On the contrary, your photographic needs & concerns are closer to mine: photography of shells. Nature editors, on the other hand, seem to be more concerned with the fudging of images of gels or cells.

Blaker does talk about the photography of tablets, by the way.

pascal said...

The versimilitude of photography was so adamantly defended in the early days of film, that the flexibility and fluidity of digital imaging has made editors a little gun shy - which is a shame, because digital imaging often needs lots of help in burning and dodging aspects.