In some applications, d₀=d_far where d_far is the farthest depth of a macroblock in the frame. If d₀=d_far is used, the implication is that the farthest object in a scene has the least quality sensitivity. The value of d_far may be, for example, computed as the maximum of the depths of the macroblocks in a frame. In some applications, rather than using d_far, the value of d₀ may be set by a user based on known characteristics of the video sequences for a particular application. In another example, video analytics may be performed as the video is captured to determine the depth range of the most visually important area in a scene and the value of d₀ adapted accordingly. The values of c₀ and c₁ may also be adapted based on scene analysis performed by the video analytics.

Other suitable depth quality sensitivity functions may also be used that represent the sensitivity of the perceptual quality of the depth of pixels in a macroblock to the relative depth of those pixels in the frame. For example, the depth-quality senstivity function may be a multi-order polynomial function of d and d₀.

Given a depth-quality sensitvity function for an MB, depth-based perceptual distortion can be modeled by