<div dir="ltr"><div><div><div>Thanks for you explanation.<br></div>I understand this much clear now.<br></div><br>If so, I think it's better let the original Opus decoder writer to do the modification.<br><br></div><div>
It seems that all the problem lies in the current opus decoder.<br><br></div></div><div class="gmail_extra"><br><br><div class="gmail_quote">On 2 December 2013 19:06, Rémi Denis-Courmont <span dir="ltr"><<a href="mailto:remi@remlab.net" target="_blank">remi@remlab.net</a>></span> wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">On Mon, 2 Dec 2013 18:01:59 +0800, Yupeng Chang <<a href="mailto:changyp6@gmail.com">changyp6@gmail.com</a>><br>
wrote:<br>
<div class="im">> Hi Denis,<br>
> I read VLC code and finally understood the relationship between<br>
packetizer<br>
> and decoder.<br>
><br>
> I think that to implement Opus Raw decoding, if we need a packetizer, we<br>
> should ADD OggOpus header in the packetizer instead of stripping the<br>
header<br>
> out.<br>
<br>
</div>I think the packetizer/decoder separation does not fits well for RTP<br>
(de)packetization. That is not really what it was meant for.<br>
I also think that generating Ogg headers is a bad idea. It might help some<br>
decoder libraries in the short term, but it will likely create more<br>
problems down the line, notably with stream output. For instance the RTP<br>
output will need raw Opus too. Furthermore, decoder libraries really should<br>
be fixed to accept raw Opus; it is architecturally wrong to require Ogg<br>
headers at decode.<br>
<br>
I would rather suggest writing common helpers to strip the Ogg headers in<br>
the *demuxers*, that is to say Ogg and MKV. As Ilkka already wrote, I think<br>
the Opus FOURCC should be reserved to raw Opus. Indeed, from a quick glance<br>
at draft-ietf-payload-rtp-opus-01, I cannot see any critical out-of-band<br>
data that would need to be carried other than the maximum sample rate and<br>
channels count (both fit in es_format_t).<br>
<div class="im HOEnZb"><br>
--<br>
Rémi Denis-Courmont<br>
Sent from my collocated server<br>
</div><div class="HOEnZb"><div class="h5">_______________________________________________<br>
vlc-devel mailing list<br>
To unsubscribe or modify your subscription options:<br>
<a href="https://mailman.videolan.org/listinfo/vlc-devel" target="_blank">https://mailman.videolan.org/listinfo/vlc-devel</a><br>
</div></div></blockquote></div><br></div>