Tissue segmentation, implicit in the assignment of opacity, is a typical prerequisite for volume rendering of medical images. Characteristics of 3D ultrasound images impede reliable segmentation and thus make volume rendering challenging. Lowpass filtering and image enhancement methods like anisotropic diffusion improve rendering appearance in some specific cutaway views, but still do not allow comprehensive all-round visual perception of 3D echocardiographic data. Starting from our automatic endocardium segmentation, we have developed a method of assigning opacity to each voxel based on its distance from the segmented endocardium and on intensities within the kernel of low-pass nonlinear filter. We achieved a highly interactive, clear, comprehensive perception of the dynamic ventricular wall motion for real-time 3D echocardiography.