Please provide complete information as applicable to your setup.
• Hardware Platform (Jetson / GPU) - GPU • DeepStream Version - 7.1 • JetPack Version (valid for Jetson only) • TensorRT Version - 10.3.0.26 • NVIDIA GPU Driver Version (valid for GPU only) - 565.57.01 • Issue Type( questions, new requirements, bugs) - not stable rtsp output without detections • How to reproduce the issue ? I attached code with logs and config
Description
Detector: Deepstream YOLO, CUDA 12.7
Python app ( pipeline.py)
I checked rtsp output, it is not stable with drops it is being destroyed, additional I can’t see detections, I ve checked frame info from tracks it is empty (see ds.log).
ffplay logs:
[rtsp @ 0x785a78000cc0] max delay reached. need to consume packet[rtsp @ 0x785a78000cc0] RTP: missed 17 packets[rtsp @ 0x785a78000cc0] max delay reached. need to consume packet[rtsp @ 0x785a78000cc0] RTP: missed 1 packets[rtsp @ 0x785a78000cc0] max delay reached. need to consume packet[rtsp @ 0x785a78000cc0] RTP: missed 15 packets[rtsp @ 0x785a78000cc0] max delay reached. need to consume packet[rtsp @ 0x785a78000cc0] RTP: missed 1 packets[h264 @ 0x785a78e4cb00] out of range intra chroma pred mode f=0/0[h264 @ 0x785a78e4cb00] error while decoding MB 51 6[h264 @ 0x785a78e4cb00] concealing 7438 DC, 7438 AC, 7438 MV errors in I frame[rtsp @ 0x785a78000cc0] max delay reached. need to consume packet[rtsp @ 0x785a78000cc0] RTP: missed 1 packets[rtsp @ 0x785a78000cc0] max delay reached. need to consume packet
```
The problem repeats itself with deepstream-app config, not stable rtsp without detections
Your input source is a RTSP source. Please enable “select-rtp-protocol” as TCP protocol to improve the rtsp source quality. select-rtp-protocol
Your [streammux] group set “batched-push-timeout=-1” configuration which is not proper for live source. Please set the value as “1/fps of your input RTSP source”.
set select-rtp-protocol=4, batched-push-timeout=17000, 33000, 40000 (I have 60 FPS on input rtsp) but result is the same, frame drops, destroyed pictures
[h264 @ 0x74f65c1cef00] concealing 1789 DC, 1789 AC, 1789 MV errors in I frame
[rtsp @ 0x74f65c000cc0] max delay reached. need to consume packet26
[rtsp @ 0x74f65c000cc0] RTP: missed 17 packets
[rtsp @ 0x74f65c000cc0] max delay reached. need to consume packet
[rtsp @ 0x74f65c000cc0] RTP: missed 10 packets
[h264 @ 0x74f65c1953c0] corrupted macroblock 53 63 (total_coeff=-1)
[h264 @ 0x74f65c1953c0] error while decoding MB 53 63
[h264 @ 0x74f65c1953c0] concealing 596 DC, 596 AC, 596 MV errors in P frame
[h264 @ 0x74f65c1b2140] corrupted macroblock 53 56 (total_coeff=16)
[h264 @ 0x74f65c1b2140] error while decoding MB 53 56
[h264 @ 0x74f65c1b2140] concealing 1436 DC, 1436 AC, 1436 MV errors in P frame
[rtsp @ 0x74f65c000cc0] max delay reached. need to consume packet26
[rtsp @ 0x74f65c000cc0] RTP: missed 18 packets
[rtsp @ 0x74f65c000cc0] max delay reached. need to consume packet26
[rtsp @ 0x74f65c000cc0] RTP: missed 1 packets
[h264 @ 0x74f65c143880] Invalid level prefix897KB sq= 0B f=27/27
[h264 @ 0x74f65c143880] error while decoding MB 32 38
[h264 @ 0x74f65c143880] concealing 3617 DC, 3617 AC, 3617 MV errors in P frame
[rtsp @ 0x74f65c000cc0] max delay reached. need to consume packet27
[rtsp @ 0x74f65c000cc0] RTP: missed 9 packets
[h264 @ 0x74f65cc630c0] Invalid level prefix486KB sq= 0B f=27/27
[h264 @ 0x74f65cc630c0] error while decoding MB 100 65
[h264 @ 0x74f65cc630c0] concealing 309 DC, 309 AC, 309 MV errors in I frame
[rtsp @ 0x74f65c000cc0] max delay reached. need to consume packet27
[rtsp @ 0x74f65c000cc0] RTP: missed 11 packets
[rtsp @ 0x74f65c000cc0] max delay reached. need to consume packet27
[rtsp @ 0x74f65c000cc0] RTP: missed 11 packets
[h264 @ 0x74f65c1646c0] out of range intra chroma pred mode f=27/27
[h264 @ 0x74f65c1646c0] error while decoding MB 100 31
[h264 @ 0x74f65c1646c0] concealing 4389 DC, 4389 AC, 4389 MV errors in P frame
[h264 @ 0x74f65c1953c0] ref 32 overflowvq= 320KB sq= 0B f=28/28
[h264 @ 0x74f65c1953c0] error while decoding MB 43 24
[h264 @ 0x74f65c1953c0] concealing 5286 DC, 5286 AC, 5286 MV errors in P frame
[rtsp @ 0x74f65c000cc0] max delay reached. need to consume packet28
[rtsp @ 0x74f65c000cc0] RTP: missed 106 packets
26.47 M-V: 0.044 fd= 3 aq= 0KB vq= 488KB sq= 0B f=28/28
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 3 times
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 128 times
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 356 times
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 201 times
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 165 times
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 97 times
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 222 times
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 78 times
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 153 times
[rtsp @ 0x753fb0000cc0] max delay reached. need to consume packet
[rtsp @ 0x753fb0000cc0] RTP: missed 6 packets
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 87 times
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 100 times
[rtsp @ 0x753fb0000cc0] max delay reached. need to consume packet
[rtsp @ 0x753fb0000cc0] RTP: missed 56 packets
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 57 times
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 46 times
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 83 times
[rtsp @ 0x753fb0000cc0] RTP: dropping old packet received too late
Last message repeated 83 times
Last message repeated 83 times
I ve checked it on local machine, the stream has become more stable, but artifacts are visible, it is being destroyed, when adding streams, the artifacts increase and stability disappears.
Maybe there are some params for setting stable multi re-streamming in config?
I have additional details about python app code, when I start it, all stages of pipe run successfully, but in logs I see (3 fps) on single stream and very low gpu usage, can you check this code or should i start a new topic?