To understand the essence of this problem, why the first team works, and the second does not. They create several different files with different contents:
ffmpeg -i test.wav -f ogg -acodec libvorbis test.a.ogg
ffmpeg -i test.wav -f ogg -acodec libvorbis - > test.b.ogg
test.a.ogg will play correctly and there will be no problems. test.b.ogg starts in the middle of the original audio and has stops and spaces in the audio. It also does not report the length of the track.
I want to transcode the source files on the fly to ogg for the program I am writing, and I am trying to pass stdout from ffmpeg to my program. Putting the results in a mediation file will kill performance, since transcoding must be done on demand.
source
share