opus - Oct 2016 - encoder with FEC+DTX enabled but not detecting noise

Hi,

When we pass around 9K samples of only ambient noise (no voice), the
encoder which is enabled FEC+DTX is detecting only some 140 frames as
non-voice (returning only TOC, no frame content). We were expecting all or
more to be identified as non-voice.

Our idea was to check how the decoder re-generates the original ambient
noise during the silence duration (when we feed NULL to decoder) when the
FEC+DTX enabled.


Please clarify.

Thanks
Suresh
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.xiph.org/pipermail/opus/attachments/20161004/657b32f6/attachment.html>

Hi,

I changed the VAD sensitivity by varying VAD_SNR_FACTOR_Q16 and now it
shows the difference in silence frames.

Thanks
Suresh

On 4 October 2016 at 17:57, Suresh Thiriveedi <sthiriveedi at gmail.com>
wrote:
> Hi,
>
> When we pass around 9K samples of only ambient noise (no voice), the
> encoder which is enabled FEC+DTX is detecting only some 140 frames as
> non-voice (returning only TOC, no frame content). We were expecting all or
> more to be identified as non-voice.
>
> Our idea was to check how the decoder re-generates the original ambient
> noise during the silence duration (when we feed NULL to decoder) when the
> FEC+DTX enabled.
>
>
> Please clarify.
>
> Thanks
> Suresh
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.xiph.org/pipermail/opus/attachments/20161005/5125b5a3/attachment.html>

I suspect the VAD has a harder time doing its job when it has never seen
any speech in the audio it receives. When in doubt, it errs on the side
of saying "speech".

	Jean-Marc

On 05/10/16 06:01 AM, Suresh Thiriveedi wrote:
> Hi,
> 
> I changed the VAD sensitivity by varying VAD_SNR_FACTOR_Q16 and now it
> shows the difference in silence frames. 
> 
> Thanks
> Suresh 
> 
> On 4 October 2016 at 17:57, Suresh Thiriveedi <sthiriveedi at gmail.com
> <mailto:sthiriveedi at gmail.com>> wrote:
> 
>     Hi,
> 
>     When we pass around 9K samples of only ambient noise (no voice), the
>     encoder which is enabled FEC+DTX is detecting only some 140 frames
>     as non-voice (returning only TOC, no frame content). We were
>     expecting all or more to be identified as non-voice. 
> 
>     Our idea was to check how the decoder re-generates the original
>     ambient noise during the silence duration (when we feed NULL to
>     decoder) when the FEC+DTX enabled.
> 
> 
>     Please clarify.
> 
>     Thanks
>     Suresh
> 
> 
> 
> 
> _______________________________________________
> opus mailing list
> opus at xiph.org
> http://lists.xiph.org/mailman/listinfo/opus
>