Dynamic tone mapping is a generic term to describe an image processing algorithm that applies localized changes to recorded luminance levels in response to overall image contrast ratio. The intention is to preserve more visible image detail in shadow areas as well as highlights.
While features such as Apple’s SmartHDR combine information from multiple exposures in order to extend highlight information, this information must be mapped into a limited range of luminance values.
A dynamic tone mapping algorithm seeks to map shadow, midtone and highlight information within this limited range of values in a way that preserves as much information as possible for the viewer while still appearing natural and well balanced.
Dynamic tone mapping automates the optimization of an image to appear more pleasing to a viewer at the expense of the photometric relationship of recorded values to reality.