Gazebo kinect does not generate point cloud for entire object

I have simulated a gazebo kinect camera on the end-effector of a manipulator. Here is the plugin for the camera:true20.0${60.0*M_PI/180.0}B8G8R8640480cameratrue10rgb/image_rawdepth/image_rawdepth/pointsrgb/camera_infodepth/camera_infocamera_depth_optical_frame0.10.00.00.00.00.00.4 But when I visualize the point cloud in RViz, the object is visible only partially.  Another kinect camera simulated which is stationary, generates a proper point cloud. I am using the kinect.xacro for this stationary camera:  Why is one plugin able to generate point cloud for entire object, while the other only partially?