Depth-Image-Based Rendering (DIBR) is a key technology for the processing and distribution of three-dimensional (3D) content. Given a two-dimensional image from a scene that was taken from a specific viewpoint and a corresponding depth or stereo-derived disparity map, DIBR enables the generation of images (i.e., novel views) that capture synthesized viewpoints of the scene. The ability of DIBR to synthesize novel views enables the generation of enhanced 3D content (i.e., additional views) for stereoscopic and multi-view displays and gives control over the 3D depth impression (e.g., adjusting the depth range). In DIBR the quality of an underlying depth map contributes to the quality of the novel views generated from it. For example, mismatches, misalignments of depth and color edges or over-smoothed depth edges can lead to visible artifacts in the novel views. We conduct a user study to investigate the effects of depth map post-processing on the perceived quality of 3D content that contains a novel view. A test environment for subjective quality assessment of the visual quality is introduced. In our study we find that filters based on local smoothing, i.e., the bilateral filter and the guided image filter, achieve significantly higher quality scores than filters based on local statistics or the unprocessed counterparts. In addition, our results indicate that the depth range within a scene has a strong impact on the visual quality of DIBR-based novel view generation and the effectiveness of depth post-processing. Furthermore, the obtained subjective results are compared against ten objective quality metrics. We observe only a weak correlation between subjective and objective quality results, which confirms the necessity of user studies in this field.