Recherche avancée

Sur d’autres sites (10848)

• Count frames in H.264 bitstream

  12 novembre 2013, par user620297

  How to count/detect frames (pictures) in raw H.264 bitstream ? I know there are 5 VCL NALU types but I don't know how to rec(k)ognize sequence of them as access unit. I suppose detect a frame means detect an access unit as access unit is

  A set of NAL units that are consecutive in decoding order and contain
  exactly one primary coded picture. In addition to the primary coded
  picture, an access unit may also contain one or more redundant coded
  pictures, one auxiliary coded picture, or other NAL units not
  containing slices or slice data partitions of a coded picture. The
  decoding of an access unit always results in a decoded picture.

  I want it to know what is the FPS of live stream out to server.

• Merge commit ’46430fd47c6239ef8742d0a34f9412d5060fa798’

  15 mai 2013, par Michael Niedermayer

  Merge commit ’46430fd47c6239ef8742d0a34f9412d5060fa798’
  * commit ’46430fd47c6239ef8742d0a34f9412d5060fa798’ :
  
    vc1dec : Don’t attempt error concealment on field pictures
  
    vc1dec : fieldtx is only valid for interlaced frame pictures
  
    aacenc : Fix erasure of surround channels
  
    aacenc : Fix target bitrate for twoloop quantiser search
  Conflicts :
  
  	libavcodec/vc1dec.c
  Merged-by : Michael Niedermayer <michaelni@gmx.at>
  

  • [DH] libavcodec/vc1dec.c

• What is the most efficient way to broadcast a live stream ? [closed]

  3 août 2020, par Harsh

  I want to build a live streaming system for a classroom. The amount on information on this subject is so confusing. These are the features/requirements that I want to have in my app :

  


  

  1. Room type system.

  


  2. One teacher - N students (N<200).

  &#xA;

  3. Broadcast video/audio. This needs to be only 1 way. (1T ---> 200S)

  &#xA;

  4. Audio chat should be possible if a teacher allows a student to speak.

  &#xA;

  5. Need not to record the session, though it would be a great feature to have.

  &#xA;

  &#xA;

  Now, from my research I have established there are many ways to go about it. The best one to me seems using WebRTC. In that case I do not have to worry about the platform that much.&#xA;WebRTC needs a STUN/TURN server, that can be easily set-up using the coturn project.&#xA;I'll also need a SFU which forwards my stream to the client, like Janus or Mediasoup.&#xA;But that's where I'm getting confused.

  &#xA;

  Can I not directly use a live stream, send it to the server, transcode it in real time using ffmpeg to HLS/DASH and publish it to a S3 bucket from where the users can access it. Wouldn't that be more efficient and able to handle much more students easily.

  &#xA;

  For the audio part I could just use the p2p functionality of webrtc in the browser itself, so no need to route that through the server.

  &#xA;

  That is how far I've come to understand the system. I still don't completely understand how SFU works and I'm confused about how many live streams can one server handle (say a 4C/8GB). Or if using ffmpeg on VPS is a bad thing and I should use the AWS services instead ?

  &#xA;

  Can someone please help me understand this ?

  &#xA;

  Thanks !

  &#xA;