RTSP stream too slow

Please provide complete information as applicable to your setup.

• Hardware Platform (Jetson / GPU) GPU
• DeepStream Version 7.1
• JetPack Version (valid for Jetson only) None
• TensorRT Version 10.6
• NVIDIA GPU Driver Version (valid for GPU only) 560.35.05
• Issue Type( questions, new requirements, bugs) Questions
• How to reproduce the issue ? (This is for bugs. Including which sample app is using, the configuration files content, the command line used and other details for reproducing)

I am trying to grab frames from 16ch rtsp streams and forward multiple detectors with nvidia refrence app deepstream_parallel_inference_app.

If input streams are video files w/o any detector, fps is around 60 and gpu usage ‘dec’ is high.

# source_file.csv
enable,type,uri,num-sources,gpu-id,cudadec-memtype
1,3,file:///workspace/vidoes/m28s1.mp4,1,0,0
1,3,file:///workspace/vidoes/m28s1.mp4,1,0,0
1,3,file:///workspace/vidoes/m28s1.mp4,1,0,0
...

However, when input streams are RTSP, FPS drops to 5~6.

# source_rtsp.csv
enable,type,uri,num-sources,gpu-id,cudadec-memtype
1,4,rtsp://RTSP_URI,1,0,0
1,4,rtsp://RTSP_URI,1,0,0
1,4,rtsp://RTSP_URI,1,0,0
...

Here are my problems,

  1. Decoding frame rate too small like 5 or 6 per second.
  2. Deepstream has 2 second delay compare to vlc player with same uri.
  3. in some point, blurring/ pixelation in some area while frame in vlc player is clean.

What config or option should I check…?

I would greatly appreciate your assistance.

• Requirement details( This is for new requirement. Including the module name-for which plugin or for which sample application, the function description)

Plesae provide all the configuration files

1 Like

f21_metamux.txt (1.2 KB)
f21_config.txt (4.7 KB)

config file is f21_config.yml but changed to f21_config.txt due to file uploader recognizs txt file only.

No primiary gie is running now.
and command is
deepstream_parallel_infer -c f21_config.yml

then estimated fps is as below:

**PERF: 7.07 (5.06)	7.07 (6.88)	7.07 (6.28)	7.07 (5.06)	7.07 (6.88)	7.07 (7.74)	
**PERF: 7.00 (5.63)	7.00 (6.94)	7.00 (6.57)	7.00 (5.63)	7.00 (6.94)	7.00 (7.37)	
**PERF: 6.98 (5.94)	6.98 (6.95)	6.98 (6.69)	6.98 (5.94)	6.98 (6.95)	6.98 (7.25)	
**PERF: 7.08 (6.14)	7.08 (6.96)	7.08 (6.76)	7.08 (6.14)	7.08 (6.96)	7.08 (7.18)	
**PERF: 6.98 (6.27)	6.98 (6.97)	6.98 (6.80)	6.98 (6.27)	6.98 (6.97)	6.98 (7.15)	

even if I feed only two rtsp stream, change bs of streammux

streammux:
  batch-size: 2

still shows

**PERF: 6.31 (6.36)	6.31 (7.22)	
**PERF: 6.19 (6.66)	6.19 (7.11)	
**PERF: 6.44 (6.45)	6.44 (6.75)	
**PERF: 7.01 (6.58)	7.01 (6.81)

dec in nvidia-smi dmon is 0~1.

However, if a single rtsp played with vlc-player dec varies

# gpu    pwr  gtemp  mtemp     sm    mem    enc    dec    jpg    ofa   mclk   pclk 

    0     31     61      -      0      3      0     40      0      0   7000   1515 
    0     43     62      -      6      1      0      0      0      0   7000   1515 
    0     38     60      -     13     10      0      1      0      0    810    900 
    0     35     61      -      6      2      0      1      0      0   7000   1425 
    0     41     61      -      2      1      0     55      0      0   7000   1515 
    0     44     62      -      4      1      0      8      0      0   7000   1905 
    0     61     63      -      3      1      0      0      0      0   7000   1905 
    0     67     63      -      4      1      0      1      0      0   7000   1905 
    0     64     62      -      5      2      0      1      0      0   5000   1215 
    0     36     62      -      7      2      0     15      0      0   7000   1620 
    0     48     62      -      5      1      0      0      0      0   7000   1620 
    0     39     60      -     14     11      0      1      0      0    810    960 
    0     39     61      -      6      2      0      0      0      0   7000   1515 
    0     41     60      -      6      2      0      0      0      0    810   1020 
    0     30     61      -      5      2      0     19      0      0   7000   1515 

It seems that network bandwidth caused this delay.

Sorry to bother you…

Have a nice day!

Thank you

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.