Using Graduated Neutral Density Filters

by John Shaw

No film can record the contrast range that our eyes can see. If you are standing in a location with contrasty lighting --for example, a forest on a bright sunny day-- you can see detail in both the darkest shadow areas and the brightest highlight spots simultaneously. Our eyes are very accommodating; most people's vision can handle a contrast range of about 12 to 14 stops. This is not true of film, which at best can only deal with about a 5-stop range in any one scene. Most beginning photographers ignore this limitation, only to be sadly disappointed in their photographs. Recognizing this compressed contrast range and learning to view subjects the same way that film records contrast, are vitally important to your photographic growth.

The newest cameras, the most expensive lenses, the latest films ... all these won't change this most basic photographic fact. If you must photograph in a high-contrast situation, a scene that goes beyond those five stops from light to dark, you can't record all the detail on the film. If you make an exposure favoring the dark area, the highlights will will blow out, turning into detailess, washed-out areas. Expose for the brightest areas and the shadows block up, resulting in a solid, featureless black blob. Split the difference, and nothing looks good.

If you're not sure you can recognize this situation, it's fairly easy to use your camera meter for a precise answer. Meter both the brightest area of your scene and the darkest area, placing them tonally where you want them to be. Now compare the two meter readings. If they are beyond five stops, you've got a problem.

For example, suppose you want to photograph snow-capped Mt. Rainier. It's a beautiful morning of a blue-sky day, with puffy white clouds slowly drifting past. You're standing in a meadow filled with wildflowers that is still shadowed at this time of day. The mountain is bathed in fantastic sunlight, and you can easily envision a great picture. But can film record the scene?

Take a meter reading of the snow on the mountaintop. Remember to use a narrow-angle meter, such as your spot meter, as you want to meter the snow and only the snow. Your meter gives you the direct values of 1/125 sec. at f/16, but you know that this will render the snow as a medium tone. To make the snow a pure white, you must open up 2 stops to 1/125 sec. at f/8.

Now meter the foreground --the shadowed meadow. It's basically a medium toned area. There are some slightly lighter flowers and a few darker leaves, but it averages out to a medium value. Take a meter reading here using an overall metering pattern. The values read out as 1/60 sec. at f/2.8.

Now count how many stops difference there are between these two readings. Well, you slowed the shutter speed by one stop and you opened the aperture by three stops, for a total of four stops. If you expose the film so that the white snow is indeed rendered as a white, the meadow will be recorded four stops darker. Four stops below white will yield an "extremely dark" tonality.

The meadow will become on film two stops darker in tonality than it appears to your eyes. Not good; it's a great meadow and you want all the flowers to be visible. But if you expose it properly as a medium-toned subject, what happens to Mt. Rainier? You still have that four-stop difference, so if you set your camera to record the meadow as a medium, the snow on Mt. Rainier will be four stops lighter. Four stops lighter than medium is so light that nothing will record on the film, so there will be a blank section of film base where the snowy top of Mt. Rainier should appear.

At this point you have several choices. You could give up in frustration and not take any pictures at all. Or you could take a photo of only the meadow and a second photo of only the mountain. But if you want both in the same frame, there's only one possible answer: you need to somehow cut the light on the sunny portion of your frame by two stops, or add two stops of light to the foreground meadow area. You need a 2-stop graduated neutral density filter.

After all this discussion, the question of how to actually use these graduated ND filters in the field remains. Let me refer to my earlier Mt. Rainier example, in which you counted the number of stops you needed to compress the tonal range. In that case, you needed to reduce the contrast by two stops, since you wanted the correct two-stop difference to remain between white and middle tone. You needed a two-stop graduated ND filter oriented so that the dark half covered the brightest part of the scene.

The best, most precise way to meter with a graduated ND is to mount the filter over your lens so that the dark part is entirely in front of the lens. Now meter the bright area and place it at its correct tonal value. So for Mt. Rainier, meter the snowy area through the dark side if the filter, then open up the exposure by two stops to make the white snow record white. Metering through the dark side takes care of any possible discrepancies in the tonality of the filter. Is your filter exactly two stops neutral density? 2 1/3 stops? 1 7/8 stops? It doesn't matter, since you've just metered right through the dark portion. This works for all brands of graduated filters, but I will say that working with the Singh-Rays is particularly easy because the stop ratings are correct.

Now that you've placed the bright area, recompose the scene as you want it to appear, then reposition the filter. This is not always easy to do. When you're viewing through a lens, the graduation line is very difficult to see, especially when the lens is wide open. Stop down your lens with the depth-of-field preview, and slightly jiggle the filter. Doing this permits you to see exactly where the graduation occurs so that you can precisely align the filter as you desire.

Be careful that you line up the graduation exactly with the shift of contrast in the scene itself. Position the density so that it overlaps into the foreground area, and you create a dark band running across the area where the highlights and shadows meet. If you position it so that the graduation stops while while still over the highlight area, you get a bright highlight streak.

Be exacting and precise in aligning the filter. when you look at a graduated ND filter, the greatest change from dark to clear seems to occur in the lighter area. However, what shows up on your film relates more to the dark side of the graduation.

By the way, the more you stop down for your final exposure, the greater the filter's effect. This is particularly true with wide-angle lenses, which are the lenses you'll use most with these filters. If you work with your lens wide open, the filter will have little effect, since it lies far outside the depth of field of the wide-open aperture. Stick with apertures around f/16 for the best results.

One last hint: Don't try to position the filter by looking at where the graduation occurs in front of your lens. Doing so will give you a false impression of the final results. how much density covers the front element does not directly equate with how much of the film image is affected. There is absolutely no substitute for looking through your lens at your shooting aperture.

This material has been excerpted from John Shaw's 160-page classic:
Nature Photography Field Guide (pp59-61).