Photographic Technique for Pixel Coordinate Analysis

It is important to know how high the camera is positioned relative to the highest ground on the summit, if the results are to be of value. In some cases, a few inches can make the difference between success and failure; in others, there is no need to know the height of the camera to a degree of accuracy any better than several feet. It depends on the error in knowing the camera height compared to other sources of error and to the rival peak elevation difference.

It is tempting to place the camera on the highest point, since that provides assurance that the photo was taken from the same level as the highest point. Experience has shown, however, that this is not recommended. That may cause it to focus on the foreground, and the background will then be out of focus. It may also cause some otherwise useful background peaks to be hidden below the foreground.

If the highest boulder is flat enough to stand on, take the picture while standing on it, holding the camera at eye level at normal standing height. Be sure to know the height of the person taking the picture. Most people's eye level is 4 to 5 inches lower than they are tall.

If the highest boulder is sloping, pointed, or difficult to stand on for any reason, have one person stand next to the highest rock, holding the camera at the same level as the highest point, while a second person sights across the camera and highest point to align them with the horizon beyond; this is to be sure that the camera is held on the same level as the highest point (within a few inches).

It is OK if the camera is tilted; the analysis can handle that just fine.

Ideally, the rival summit should be near the center of the image, with background peaks on both sides. If one side of the rival summit is devoid of useful background peaks, you can consider offsetting it somewhat, if you would get more background peaks on the other side. Due to projection error, however, there are limitations as to how much offset there can be without compromising the result. It is strongly recommended that all background peaks, and all rival peaks one intends to analyze, be within 10 of the center of the image, and preferably within 5.

A good background peak is not necessarily very prominent, nor is it one that you can identify from the field. Just about any peak can be identified in the post-mortem image. The best background peaks are relatively sharp, so that you can pinpoint their pixel coordinates exactly. Broad, rounded summits are usually not very good. Little sharp blips, even if they appear insignificant from the summit, often end up being good background peaks if you can identify them on a map later.

Back to the Geopix home page