The idea is to display bright scenes as bright and poorly lit scenes
as dark, making the differences just visible in the real world just
visible on the display. In other words the visibility in the scene is
preserved. The scale-factor is derived from the contrast sensitivity
studies conducted by Blackwell in the early 1970s (see Human vision
section). Blackwell established the following relationship between
the adaptation luminance 
and the minimum discernible difference in
luminance:
Ward used Blackwell's studies, but we should be aware that all models
described here and based on human perception are only an approximation
of the human vision system, since Blackwell's experiments were conducted
in perfect laboratory conditions that are far from the usual complex
viewing conditions of a typical work place. Ward wanted to find a
proportionality constant between display luminance and world
luminance that yields a display with roughly the same contrast
visibility as the actual scene. He wanted to find a multiplier m such
that:
where 
is display luminance at an image point, and 
is
world luminance at an image point.
The key assumption that enables the calculation of m is that:
where 
is the minimum discernible luminance change at
, is the display adaptation luminance and 
is
the world adaptation luminance.
This assumption makes differences just visible in the real world just
visible on the display, and this was the main goal of the whole
mapping. Solving the equation 4.15 for m gives:
In order to get the scale-factor which converts a raw image to a display
device input interval [0,1], the maximum display luminance, 
,
should be known. The display adaptation luminance of the viewer, ,
should be known as well. Ward suggests to take 
, as
he has found that this is close enough for most applications. The
final scale-factor n is then:
Two unknowns are the maximum display luminance which can be measured, and the world adaptation level. If no other information is available, the average value of the raw image can be taken as the world adaptation level although this is not the completely correct way. If the area of interest in the scene is known, this can be used as the adaptation level. In this way results can be interpreted as telling us how well a person would see in an environment while looking at this point. Note that the raw image should be rendered in absolute units to use this method. This is still not easy for a common user, but this is the only way the lighting atmosphere can be automatically taken into account. This simple linear scale-factor can be used for a wide range of applications where lighting simulation is important (architecture visualisation) and it renders great results. Color plates 9c, 9d, 13a, 13b, 17a, and 17b in results chapter illustrate this method.