Re: [codec] Conformance with unusual sample rates

[Stephan Wenger <[e-mail]>]

Uh.  Scary discussion.
I believe we may well be unchartered territory here.  Remember the "green
light / red light" discussion we had here in Taipei?  I'm fairly certain
that both 16 ms frames and 44.1 kHz sample rate would fail the colored
lights test, with the text vectors considered today, even if we were
defining conformance very loosely.  In such a case, I don't think that you
can necessarily rely with any certainty on the IPR declarations provided.
Stephan

On 11.19.2011 00:20 , "John Ridges" <[e-mail]> wrote:

>Thank-you for clarifying that. I also have another application where I
>need to use 16 ms frames and it looks like Opus-custom would allow me to
>do that.
>
>John Ridges
>
>On 11/18/2011 8:56 AM, Benjamin M. Schwartz wrote:
>> On 11/18/2011 09:29 AM, John Ridges wrote:
>>> Is it possible to create a conforming (and
>>> thus unencumbered) Opus codec from the reference implementation that
>>>can
>>> use a sample rate of, say, 44100 Hz?
>> It depends what you mean.
>>
>> One answer is "no".  Opus always operates at a fixed nominal samplerate
>>of
>> 48000 Hz.
>>
>> Another answer is "yes", because you are of course always free to
>>resample
>> your inputs before encoding, and to resample the outputs after decoding.
>> If you figure out a way to use less CPU by integrating your resampler
>>into
>> the decoder then that's just a trivial implementation detail.
>>
>> Another answer is "yes, in Opus-custom".  Opus-custom is a
>> non-interoperable Opus-derived codec development system, defined as part
>> of the Opus standard [1].  Opus-custom can be driven at different
>> samplerates, including 44100 Hz.
>>
>> --Ben
>>
>> P.S. I do agree that clarifying our text on this point might be helpful.
>>
>> [1] http://tools.ietf.org/html/draft-ietf-codec-opus-10#section-6.2
>>
>
>_______________________________________________
>codec mailing list
>[e-mail]
>https://www.ietf.org/mailman/listinfo/codec