Hi,
Thanks for the reply.
I infer from it, that there is no routine in the api that would populate
AVCodecContext on the basis of feeding it sps and pps packets only, right..?
Regards,
Sampsa
On 29.11.2016 13:11, Carl Eugen Hoyos wrote:
2016-11-29 12:07 GMT+01:00 Sampsa Riikonen <[email protected]>:
When reading AVPackets from a file, it is straightforward to pass correct
parameters to the ffmpeg/vdpau decoder, i. e., we call
av_format_find_stream_info
and get the relevant data to "AVStream(s)". Then we do:
av_codec_parameters_to_context(AVCodecContext,AVStream.codecpar[AVCodecParameters])
That populates correct parameters into AVCodecContext, namely,
AVCodecContext->width, height, etc.
This works also with "raw" H264 streams (i.e., with ".h264" files), that are
not "container" formats, such as matroska, etc..
.. so the only conclusion is, that "av_format_find_stream_info" does some
kind of sps/pps packet parsing, right?
It calls libavcodec that does sps/pps parsing (this is the reason libavformat
depends on libavcodec).
Carl Eugen
_______________________________________________
Libav-user mailing list
[email protected]
http://ffmpeg.org/mailman/listinfo/libav-user
_______________________________________________
Libav-user mailing list
[email protected]
http://ffmpeg.org/mailman/listinfo/libav-user
