Any suitiable depth quality sensitivity function may be used. For example, the function of Eq. 2 may be used. The values of d_o and the tuning parameters c_o and c₁ may depend on the particular application in which the video encoder is used, e.g., the values may be constants in some applications, may be user-settable parameters in some applications, may be adapted in some way for scene changes in some applications, etc. In some embodiments, the value of d₀ may be determined for each multi-view frame as the maximum of the depths of the macroblocks in the multi-view frame. In some embodiments, the value of d₀ may be adapted periodically based on the maximum macroblock depth of a sequence of multi-view frames. In some embodiments in which the background of a scene does not change significantly over time, e.g., for video surveillance, the value of d₀ and the values of the tuning parameters c_o and c₁ may be constants determined based on the background of the scene and the expected depths of objects that may move through the scene. In some embodiments, the video analytics component 220 (see FIG. 2) may analyze the scene across muliple frames and provide information regarding scene changes to the video encoder component 218 that may be used to adapt the values of d₀, c_o and/or c₁ for scene changes. In some embodiments, the depth quality sensitivity function may not be linear. For example, a quadratic depth quality sensitivity function may be used when both closer and more distant objects are more visually important than objects in between.