On Thu, Jun 02, 2016 at 04:28:13PM +0200, Vlad Tarca wrote:
> Pro-MPEG Code of Practice #3 release 2 forward error correction for
> rtp_mpegts streams
>
> Signed-off-by: Vlad Tarca
> ---
> Changelog | 1 +
> doc/general.texi| 1 +
> doc/protocols.texi | 35
>
> also why are some called _size and some _len ?
I will use length_recovery as it's named exactly like that in the code of
practice and use better names for the rest.
The size check will also take into account UINT16_MAX and not INT_MAX since
the recovery field is only 16-bit.
> -ret = ffurl_
On Wed, Jun 01, 2016 at 11:19:38AM +0200, Vlad Tarca wrote:
> Pro-MPEG Code of Practice #3 release 2 forward error correction for
> rtp_mpegts streams
>
> Fixes:
>
> [prompeg.c]
> - proper RTP size check to fit the largest buffer
> - UDP port overflow check
> - replaced ffurl_close calls with ff
On Mon, May 30, 2016 at 11:04:03AM +0200, Vlad Tarca wrote:
[...]
> +static int prompeg_init(URLContext *h, const uint8_t *buf, int size) {
> +PrompegContext *s = h->priv_data;
> +uint32_t seed;
> +int recovery_len, bitstring_len, rtp_buf_len;
> +int i;
> +
> +s->fec_buf = NULL;
On Wed, May 18, 2016 at 04:18:11PM +0200, Vlad Tarca wrote:
> >
> > Have you tested this on actual receivers?
> > Are you aware this won't work effectively because of the bursty nature of
> > FFmpeg?
> >
> > Kieran
> >
>
> Hello,
>
> I tested the FFmpeg version on a software receiver, not a hardw
On Wed, May 18, 2016 at 02:46:33PM +0200, Vlad Tarca wrote:
> Fixed a type issue in xor added after the last modification and cleaned up
> unnecessary NULL checks for av_dict_free
[...]
> +static int prompeg_init(URLContext *h, const uint8_t *buf, int size) {
> +PrompegContext *s = h->priv_da
>
> Have you tested this on actual receivers?
> Are you aware this won't work effectively because of the bursty nature of
> FFmpeg?
>
> Kieran
>
Hello,
I tested the FFmpeg version on a software receiver, not a hardware one. I
initially implemented this for one of my applications based on VideoLAN
On Wed, 18 May 2016 at 13:52 Vlad Tarca wrote:
> Fixed a type issue in xor added after the last modification and cleaned up
> unnecessary NULL checks for av_dict_free
>
>
Hi,
Have you tested this on actual receivers?
Are you aware this won't work effectively because of the bursty nature of
FFmpe
Hi
On Fri, May 13, 2016 at 04:30:34PM +0200, Vlad Tarca wrote:
> Pro-MPEG Code of Practice #3 release 2 forward error correction for
> rtp_mpegts streams
>
[...]
> +static const AVClass prompeg_class = {
> +.class_name = "prompeg",
> +.item_name = av_default_item_name,
> +.option
