#17. The Accuracy of Reflected-light Exposure Meters
True or False?
Reflected-light exposure meters are often inaccurate because they can be “fooled” by excessive highlights or shadows in a metered scene.
Reflected-light exposure meters with a single photosensitive sensor (such as hand-held exposure meters) can integrate (average) the light reflected from all the surfaces in the scene to create a representative illuminance for the scene Es. The Illumination Equation is then used to convert that representative illuminance into a representative luminance Ls that is used in the common (midtone) Exposure Equation to get the camera settings (aperture number and shutter speed) that expose the representative luminance at the midtone exposure of the photosensitive exposure range in use.
These meters can be “fooled” by excessive highlights or shadows in a scene not because the measured illuminances or the calculated representative luminance are inaccurate, but rather because that representative luminance is matched to the midtone exposure of the photosensitive exposure range in use. Excessive highlights or shadows drag the representative luminance off of the midtone. That is no more inaccurate than calling those scenes as they are–high key or low key. Meters with one or two sensors do not have a way to analyze the distribution of tones in a scene and recognize various high and low key scenes. Their measurements and calculations are correct: It is the assumption about the scene that is not.
A photographer that knows that his single-sensor meter is matching a luminance representative of a scene to the photosensitive midtone of the photosensitive array that he is using can always shift this match using the exposure compensation or using another more appropriate shutter speed.
Reflected-light exposure meters with segmented photosensitive sensors (such as those found in cameras), however, can construct and analyze a distribution of tones in a scene. The mathematical operations that allow this statistical pattern matching also allow camera-held meters to shift representative luminance of the distribution or to signal the camera to compensate, so that a more typical luminance is matched to the midtone exposure of the photosensitive exposure range in use.
Shifting the representative luminance of a distribution of tones in a scene based on statistical pattern matching doesn’t always work perfectly because there is never enough samples to describe all possible distribution. Nevertheless, given thousands of images analyzed for the distributions of their tones and for an effective representative luminance, the statistical pattern matching that supports segmented meters works just about all the time.
Copyright 2008 Michael G. Prais, Ph.D.
For a readable but in-depth analysis of this concept along with many other concepts associated with photographic exposure, take a look at the book Photographic Exposure Calculations and Camera Operation. This book provides insight into the equations that govern exposure, exposure meters, photosensitive arrays (both solid-state and emulsion) and the Zone System as well as concepts associated with resolution, dynamic range, and depth of field.
The book is available through Amazon.com (ISBN 978-1-4392-0641-6) where you can Search Inside!™.
Check http://michaelprais.info under Photography for the table of contents, an extensive list of the topics and subtopics covered, the preface describing the purpose of the book, and a diagram central to the concepts in the book.
Should you have any comments or questions about this web site, please contact me. Thanks.