[GST] gStreamer pipeline 분석

gStreamer pipeline 분석

 

1. prerequisite

 

~$ sudo apt-get install graphviz ~$ export GST_DEBUG_DUMP_DOT_DIR=/tmp/

 

2. run your pipeline

 

~$ gst-launch audiotestsrc num-buffers=1000 ! fakesink sync=false or ~$ gst-launch-1.0 filesrc location=/red_rookie.mp3 ! mpegaudioparse ! nvmediamp3auddec ! audioconvert ! audioresample ! pulsesink or ~$ gst-launch-1.0 playbin uri=file:///red_rookie.mp3

 

3. generate diagram

 

1) check output dot files

~$ ll /tmp

 

   0.00.00.972540004-gst-launch.NULL_READY.dot

   0.00.01.051387461-gst-launch.READY_PAUSED.dot

   0.00.01.074729712-gst-launch.PAUSED_PLAYING.dot

   0.00.12.187852589-gst-launch.PLAYING_PAUSED.dot

   0.00.12.201485839-gst-launch.PAUSED_READY.dot

   psplash_fifo

 

2) generate diagram

~$ dot -Tpng 0.00.12.187852589-gst-launch.PLAYING_PAUSED.dot > pipeline.png

 

3) see diagram

eog pipeline.png

 

4. helper script

 

DOT_FILES_DIR="fs/fs/graphs" PNG_FILES_DIR="fs/fs/graphs" DOT_FILES=`ls $DOT_FILES_DIR | grep dot` for dot_file in $DOT_FILES do  png_file=`echo $dot_file | sed s/.dot/.png/`  dot -Tpng $DOT_FILES_DIR/$dot_file > $PNG_FILES_DIR/$png_file done

 

5. app에서 생성하기

 

1) define 추가

모든 element의 생성 및 link 이후에 다음의 코드를 수행

GST_DEBUG_BIN_TO_DOT_FILE(pipeline, GST_DEBUG_GRAPH_SHOW_ALL, "pipeline")

2) 앱 실행

~$ GST_DEBUG_DUMP_DOT_DIR=. ./application

 

3) image 생성

앱 실행 위치에 dot file 생성

위에서 설명했던 것처럼 png file 생성 후 check

 

 

참고:

 

https://developer.ridgerun.com/wiki/index.php/How_to_generate_a_Gstreamer_pipeline_diagram_(graph)