Hi there,
i am writing some sort of a media player right now, but when i convert
the audio samples into the target format i need - the memory is always
increasing even though when i call av_frame_unref() und
av_packet_unref() when i am done with the frame/packet. If i disable the
audio sample software conversion, the memory is fine. I am new to FFMPEG
development, so i must do something horrible wrong...
Code is written in C++ but with C style and uses my own platform
abstraction library. The way i do things is different from other
examples. I have 4 separate threads, one for packet reading and queuing
up packets into the proper places, one for video frame decoding + queue,
one for audio samples decoding + queue and finally one for outputting
the decoded audio samples to the audio device. The audio samples
conversion is done in the output thread and queued up in one audio
buffer array.
Can you please look into my demo source? Maybe you see something stupid
what i dont see since days:
https://github.com/f1nalspace/final_game_tech/blob/master/demos/FPL_FFMpeg/fpl_ffmpeg.cpp
Thanks,
Final
_______________________________________________
Libav-user mailing list
[email protected]
http://ffmpeg.org/mailman/listinfo/libav-user
