Digital Zone System

Ever since getting into large format photography, I've used the zone system to set up my exposures. This is a system of sensitometry used and taught by Ansel Adams, and was the culmination of years of testing and in-the-field experience by him and his colleagues. By being aware of the dynamic range of your film, and knowing what you could do in the darkroom to compress or expand that range, it let you meter a scene and very precisely predict how it would be rendered on film. I won't presume to go into a full description of the zone system because I can't do as good a job as Adams himself. For that, see his book, "The Negative".

I admit I'm lazy when it comes to digital photography, though. I don't use the zone system, and I slip into complacency when metering, trusting the camera's electronics to do a better job than I can. This is a terrible way to operate! The camera may have a very good idea of how to render a particular tone as an 18% gray, but it has no clue whatsoever what the photographer had in mind for a scene!

Here's a good example: You're shooting a sunset, and want to get a foreground palm tree to show up as a silhouette against the glowing sky. So you whip up your camera, half-press the shutter button to meter the scene, and expose. Trusting to fate, you go about your business, only to find out later that the tree is rendered as a muddy, dim, tree-colored mess, and the sunset sky is blown out. Grrr! Photoshop can fix this by fiddling around with levels and curves, but the resulting picture looks almost Technicolor because the dynamic range has been squeezed around so much.

Better to meter the scene, know how bright the tree is, and set your camera so you know the tree will come out black, and the sky will be rendered in all its glory. To get there, though, you first have to know the dynamic range of your camera.

Regardless of whether you're shooting film or digital, the way to test the dynamic range of your detector is to shoot a textured, uniform-toned subject at various exposure settings, and look at the shots. Below a certain exposure value, it should render as pure black. Above a certain exposure value, it should render as pure white. Between those values is the dynamic range of your detector. Film, CCD, CMOS, doesn't matter.

Most B&W films exposed and developed normally give you about ten stops of dynamic range. You can tweak the developing to change that, but that's the base point. Color films typically gave you about seven stops of range. Color slide, about five. But it's always best to test it yourself and not trust to fate! I always tested a new film before taking it out in the field. Why should digital be any different?

I set my camera, a Canon PowerShot A650 IS at ISO 100, and pointed it at my trusty textured white target, a block of styrofoam. (Hey, better for it to wind up in my studio than in a landfill!) With the camera set on manual and the aperture fixed at f/8, I ran exposures from 1/2000 to 2 seconds in 1 stop increments. The results were a little surprising.

• 1/2000 (+0) was black. Pixels ranged from 0,0,0 to 1,1,1 in value. Down in the noise.
• +1 stop had some value (up to 2,2,2) but nothing I'd call texture. Noise.
• +2 stops started to show real texture. Call this the beginning of the dynamic range.
• +3 - +10 showed texture.
• At 1 second and beyond (+11 stops), the image was maxed at 255,255,255. Beyond the range.

At ISO 100, this camera has a usable dynamic range of eight stops. So it behaves more or less (slightly more) like a color negative film camera. Nice!

Something else that can be had from a test like this is the response curve of the detector. So starting at 1/2000, I took the central 1000x1000 patch of image and looked at the mean value of intensity for the whole patch:

+0 0.41
+1 1.42
+2 3.68
+3 9.60
+4 20.38
+5 39.97
+6 72.46
+7 112.97
+8 154.53
+9 209.67
+10 248.84
+11 255.00

Keep in mind that each additional stop of exposure value should double the amount of light hitting the detector. Given that, you can see my camera is not the most linear thing in the world. So if you're using a digital camera as a quantitative photometric device, be forewarned! These things may not be linear! If in doubt, test your gear.

Armed with this information and a spot meter (which the A650 IS has, thank goodness) I can start making better decisions about metering and exposure. My A650 will never replace my monorail, but that doesn't mean I can't apply the same care and attention to detail when using it.

Tom