I did a dirty little hack to attempt to fix it. Don't call the dshow_add_device() just in build time, but run the graph and sleep() a bit, then call dshow_add_device() at the end of the dshow_read_header().
Sadly, it does not work, it does not goes beyond the dshow "chit-chat" between filters/pins. After ctrl+c however the right codec stats are displayed... Regards, Máté On Mon, Aug 24, 2015 at 10:29 PM, Michael Niedermayer < mich...@niedermayer.cc> wrote: > On Mon, Aug 24, 2015 at 02:09:28PM -0600, Roger Pack wrote: > > I've run into the case today where (if we understand it correctly) you > > setup a directshow graph, it advertises media types, then when you > > "start" the graph, it actually calls through and says "here's your > > *real* media type". > > Does ffmpeg internals have any concept of or support for a "changing > > media type" at runtime? (any other suggestions for how to handle this > > if it doesn't?) > > do i understand correctly that the type at read_header() stage is > wrong but the type later at read_packet() stage is correct ? > if that is the case, you can just wait with setting the type till it > is reliably known. > > [...] > -- > Michael GnuPG fingerprint: 9FF2128B147EF6730BADF133611EC787040B0FAB > > I have never wished to cater to the crowd; for what I know they do not > approve, and what they approve I do not know. -- Epicurus > > _______________________________________________ > ffmpeg-devel mailing list > ffmpeg-devel@ffmpeg.org > http://ffmpeg.org/mailman/listinfo/ffmpeg-devel > > _______________________________________________ ffmpeg-devel mailing list ffmpeg-devel@ffmpeg.org http://ffmpeg.org/mailman/listinfo/ffmpeg-devel
