During the initialisation process, the depth buffers are filled with the same value, MAX_DEPTH which is the
maximum value that they can store (for example, 0xFFFFFFFF for 32 bit representation). So, if there are no second
objects (that is no shadow receivers), no value is written into the secondary depth buffer and it remains initialised.
Consequently, the ∆ value is computed only when there is a difference to compute, that is when there are values in
the primary and second buffers, different from the initialisation value. For the same reason, the maximum and minimum
and updated when a ∆ value is computed. Thus, each cell of the primary and secondary depth buffers are processed
together, producing a delta value ∆ according to their difference