[Matroska-devel] Opus in Matroksa Cont.

Moritz Bunkus moritz at bunkus.org
Fri Jul 5 21:47:15 CEST 2013


Hey,

I think you misunderstood me. I did not mean to say that the Opus
specs (any of them) should be changed. I only meant the wording in the
_Matroska_ specs for the newly-introduced "CodecDelay" element[1]
which, at the moment, reads as follows:

> CodecDelay is the delay in nanoseconds to add to each Block timestamp.
> The value should be small so the muxing of tracks with the same actual
> timestamp are in the same Cluster.

I think this wording is extremely misleading and does not really allow
the use of this element to hold Opus' pre-skip value -- though it has
been our intention to store just that in it.

This is purely a matter internal to the Matroska team and specs and
has nothing to do with any of the Opus RFCs and drafts.

Kind regards,
mosu

[1] http://www.matroska.org/technical/specs/index.html

On Fri, Jul 5, 2013 at 9:34 PM, Timothy B. Terriberry <tterribe at xiph.org> wrote:
> Cross-posting to codec: N.B. that <http://www.ietf.org/about/note-well.html>
> applies to replies there.
>
> Moritz Bunkus wrote:
>>
>> an audio packet corresponds to a video packet with timestamp 0. It also
>> seems not so say what Opus needs it to say: that a number of samples
>> have to be dropped because they're simply invalid.
>
>
> Quoting from the first paragraph of Section 4.1: "Therefore, the first few
> samples produced by the decoder do not correspond to real input audio, but
> are instead composed of padding inserted by the encoder to compensate for
> this latency."
>
> What do you think the text should say to make it clearer that "a number of
> samples have to be dropped because they're simply invalid"?
>
> _______________________________________________
> Matroska-devel mailing list
> Matroska-devel at lists.matroska.org
> http://lists.matroska.org/cgi-bin/mailman/listinfo/matroska-devel
> Read Matroska-Devel on GMane:
> http://dir.gmane.org/gmane.comp.multimedia.matroska.devel
>


More information about the Matroska-devel mailing list