[opus] Newbie question on VAD

[Harri Pasanen]

https://tools.ietf.org/html/draft-ietf-codec-opus-03 seems to indicate 
that the Silk encoder in Opus would have a VAD in there.

But thanks for the pointer to webrtc VAD, I'll look into it.

/Harri


On 01/16/2013 09:24 PM, John Ridges wrote:
> Questions aside as to whether this is a dumb idea (I'm not really 
> qualified to answer that), Opus doesn't have a VAD because it's 
> outside the scope of the codec (Opus is meant for any audio, not just 
> voice). That said, I should warn you that the Speex VAD is quite poor 
> (it was never really finished) and is pretty much unusable. If I were 
> you, I'd try the webrtc VAD. It's also open source, has ARM 
> optimizations, and seems to work pretty well.
>
> --John
>
> On 1/16/2013 1:00 PM, Harri Pasanen <harri&mpaja.com> wrote:
>> Hi,
>>
>> I'm interested in speech recognition and it occurred
>> to me to try to use VAD to automatically detect the start of speech
>> prior to feeding the stream to the recognition engine.
>>
>> Speex has a preprocessor capability which presumably could be
>> used for this, but I did not see similar in my quick perusal of Opus
>> docs.
>>
>> First, is this is a dumb idea for some reason?   If not, should I 
>> just use
>> Speex preprocessor, or can Opus be coaxed to do this?
>>
>> I'm targeting mobile platforms, so ARM v7 in practice.
>>
>> I'm noob in codec domain, but not as a developer.
>>
>> Thanks,
>>
>> Harri
>
>