Is there a reason why opus_repacketier_cat is limited to 120ms packets? It seems 120ms frames can only be combination frames, that is, only constructed by combining smaller frames of the same config. If this is true, then why stop at 120ms and not support longer packets? Regards, -- Daniel. -------------- next part -------------- An HTML attachment was scrubbed... URL: http://lists.xiph.org/pipermail/opus/attachments/20141219/622228d6/attachment.htm
Daniel K wrote:> true, then why stop at 120ms and not support longer packets?https://tools.ietf.org/html/rfc6716#section-3.2.5 "...the audio duration contained within a packet MUST NOT exceed 120 ms..."
On Fri, Dec 19, 2014 at 9:03 PM, Timothy B. Terriberry <tterribe at xiph.org> wrote:> Daniel K wrote: >> true, then why stop at 120ms and not support longer packets? > > https://tools.ietf.org/html/rfc6716#section-3.2.5 > > "...the audio duration contained within a packet MUST NOT exceed 120 ms..."If you're wondering why the limit exists, there are a number of reasons... e.g. the overhead is already negligible by 120ms and the larger the amount of audio in a packet the more memory a straight-forward normally layered implementation may need to have to deal with it.