encoder RateControlMode Problem

Hi Nvidia:
In example 03_video_cuda_enc,Add ctx->enc->SetRateControlMode(enum v4l2_mpeg_video_bitrate_mode mode) in main Function to change encoder’s Rate_Control_Mode, but when mode is V4L2_MPEG_VIDEO_BITRATE_MODE_VBR and the mode is V4L2_MPEG_VIDEO_BITRATE_MODE_CBR, the encoded files size is similar equal, of course,the input file is the same one. the expected result is in mode V4L2_MPEG_VIDEO_BITRATE_MODE_CBR, the encoded file size should maxer than its in V4L2_MPEG_VIDEO_BITRATE_MODE_VBR mode, and when Bitrate grow larger,the difference between in V4L2_MPEG_VIDEO_BITRATE_MODE_CBR and in V4L2_MPEG_VIDEO_BITRATE_MODE_VBR is larger and larger.Is it right?

 Bitrate              V4L2_MPEG_VIDEO_BITRATE_MODE_CBR                          V4L2_MPEG_VIDEO_BITRATE_MODE_VBR
 0.1M                              5911KB                                                    5932KB
 0.2M                              5916KB                                                    5942KB
 0.3M                              5925KB                                                    5909KB
 0.5M                              5925KB                                                    5890KB
 0.8M                              5934KB                                                    5902KB
 1.0M                              5931KB                                                    5883KB
 2.0M                               7.3MB                                                     7.2MB
 4.0M                              13.8MB                                                    13.8MB
 8.0M                              27.8MB                                                    27.8MB
16.0M                              56.2MB                                                    56.2MB
32.0M                             110.8MB                                                   110.7MB

Expected your reply! Thank you!


Usually, you can achieve a smaller file size by encoding VBR than CBR for the flexibility of bitrate.
But this slightly relies on the context of the video.