Hi all.
I am writing an application, capturing an MPEG-2 transport stream from the DVB-
S card and saving its parts (for example, a video stream and one of
accompanying audio streams) to disk files, according to the user defined
schedule. The files should be playable with a usual software video player (say,
Windows MediaPlayer).
I have already developed a code which extracts from the transport stream the
packets with needed PIDs (desired elementary PIDs, carrying audio and video
streams).
I am rather new to FFmpeg. Thanks to Andreas Öman, I have obtained some
knowledge about its data structures.
Now I have come to the following algorithm.
Accumulate packets with desired PIDs (in separate buffer for each PID)
Try to decode the buffers periodically using
avcodec_decode_audio2 and avcodec_decode_video.
When decoding was successful,
av_init_packet
[ fill packet's fields ]
av_interleaved_write_frame
However, I have some doubts about if I need decoding.
According to the task, I don't have to decode or re-encode video or audio
frames.
I think I need simple multiplexing of audio and video data, come in the stream.
FFmpeg is certainly capable to multiplex audio and video data.
Could anyone direct me to the necessary functions?
_______________________________________________
libav-user mailing list
[email protected]
https://lists.mplayerhq.hu/mailman/listinfo/libav-user
