Page 3. Version 1.3, ©2006 by Dale Cotton, all rights reserved.
Fig. 8: Typical exposure series for finding X
The next step is to determine X for your particular camera, where X is distance between the blown highlights mark and the exposure level your camera suggests via its spot or centre-weighted meter. Experience shows that this is not simply the latitude (DR) of your camera divided by two. The latitude of my LX1 is somewhere around 7.5 stops, but the spot meter weighs in at 1.5 stops under the highlight clipping point. Fortunately, 'tis simplicity itself to find your camera's X value. Here's the drill:
That exposure compensation value is X. X happens to be 1.5 stops on my LX1 and 3.0 stops on my Pentax DS. (You only have to do this once per camera and it isn't rocket science.)
Fig. 9: Brightest portion of scene is pure white paint under the deck, metered at +1 2/3 with LX1, and allowed to spike.
So let's say you're shooting with a Nikon D50, you've found that X=3.0 for your camera (I have no idea what the actual X would be for a D50), now you want to put this to use. We know that the concept is to spot meter the brightest portion of the scene then add 3.0 stops to whatever the meter recommends. Since exposure compensation on the D50 is a whopping -/+ 5 stops, one can use aperture or shutter priority, leave exp. comp. set to +3, press the AE Lock button when the meter is pointed at the brightest (non-specular) spot in the scene, then recompose and shoot as desired. Other cameras have the more typical -/+ 2 stops exp. comp., so if X is greater than 2.0, one would have to use manual mode, dialing in either shutter or aperture, over the meter's recommendation.
Fig. 10: First guess was shed roof but sky proved brighter; specular glare on windshield ignored.
Both these procedures sound much more complicated written out than they actually are. A bit of practice will make them second nature. Generally, the most tedious part is switching back and forth between this and multi-metering for action shots. Hopefully, you can hit on some equally efficient method for achieving the same results with your own model camera. If so, e-mail it to me and I'll append it to this tutorial and include your name as discoverer.
(For a more comprehensive exploration of metering theory and practice, see my tutorial: Existing Light Exposure Metering.)