On 9/10/2016 12:56 AM, Timur Guseynov wrote:
Hi.
I am new to C++ and FFmpeg and wanted to write application that
streams video from camera (webcam at least) to some server.
I've seen StreamingGuide
<https://trac.ffmpeg.org/wiki/StreamingGuide> and wanted to know how
to implement it.
I think that basic flow is like this, please correct me if I'm wrong:
1. Get input device AVInputFormat from libavdevice
2. Open that input with avformat_open_input
3. Find its streams' codecs
4. Get decoder with avcodec_find_decoder
5. Decode it somehow
6. Encode it for stream
7. Write data with muxer
8. Send muxed data to server
So I imagine how to implement the first half of this list but not the
second.
2 questions that I have are:
1. Do I understand the streaming flow right? What are the nuances
that I must consider? What modules/methods should I look into to
implement it?
2. How can I do a preview of stream in a GUI using, for example, Qt
Quick? Is an input device blocked by one of the processes (either
FFmpeg or Qt)? If it is, should I somehow copy frames for GUI to
show them for user or just reference it?
Thanks in advance!
Kind regards,
Timur Guseynov
_______________________________________________
Libav-user mailing list
[email protected]
http://ffmpeg.org/mailman/listinfo/libav-user
That is pretty close if you want to display it in a gui.
To display it you just draw the decoded bytes to some hardware overlay.
And yes, typically only one application can acquire the capture stream
output at a time. You can register pseudo input devices which distribute
the output of an actual camera. There are products available that do this.
For the server, you could set up a output for rtmp or rtsp and then
write the encoded av packets to it.
Have Fun!
Andy
_______________________________________________
Libav-user mailing list
[email protected]
http://ffmpeg.org/mailman/listinfo/libav-user
