The photographic tone mapper converts actual pixel luminances (in candelas per meter squared) into image pixels as seen by a camera, applying camera-related parameters (like f-stops and shutter speeds) for the exposure and applying tone mapping that emulates film/camera-like effects.
The photographic tone mapper has two basic modes: Photographic and Arbitrary.
In “Photographic” mode, the tone mapper assumes that input values are or can be converted to candelas per meter squared.
In “Arbitrary” mode, scene pixels are not considered to be in any particular physical unit but are simply scaled by a factor to fit in the display range of the screen.