2008/5/20 Stream Alerts Support <[EMAIL PROTECTED]>:

>
> > I just need few precision...
> >
> > - You are talking from the client side, aren't you ?
> > - This "buffer" you talked about is like the first bytes of the movie ?
> > Whatever the format is ? Or should I separate the header and the first
> few
> > packages first ? Because I still haven't found the way to get the header
> of
> > the movie...
> >
> > So then, what is the server doing ? Only sending regularly piece of the
> > movie file ? It wasn't how I imaginate the process, but maybe I was just
> > wrong.
> >
> > B.R.
> >
> > Thibault
>
> Hello,
> Yep, all of that is on the client side. So on the server side, you want to
> send the encoded audio/video packets to your client application and then
> have the client grab the packets and store them in that
> pFormatCtx->pb->buffer. Your server can just read chunks from the movie file
> (the raw bytes) and send them to your client. The header should be included
> in the raw data and when you put it in the context buffer, it will be parsed
> as if it was reading the file locally. You do need to know the file format
> ahead of time (FLV, AVI, etc) so you can initialize pFormatCtx->iformat but
> then it will automatically grab the correct codecs. Hope that helps!


Hi,

That's what I thought, but I really needed to reduce as much as possible
calculations on the client side. After mucking around a bit more, I finally
found that it was the "av_class" that I missed. When properly sent, I could
use the pCodecContext on the client side ! (maybe it could be usefull for
somebody else).

I know it's not the best way to do this, but it works and that's all I ask
so far.

Anyway, thanks a lot for your answers.

B.R.

Thibault
_______________________________________________
libav-user mailing list
[email protected]
https://lists.mplayerhq.hu/mailman/listinfo/libav-user

Reply via email to