ACELP | ADPCM | AMBE | AMR | AMR-NB | CELP | Codebook | DTMF | G.723 | G.726 | G.729 | GSM | Interpolation | LPC | LSF | LSP | MELP | PCM | Perceptual | Pitch | PSOL | QCELP | Quantization | SMV | VAD | Vocoder

hi,

I am using AMR 4.75 and AMR 5.15 for Voip application over very small
bandwidth line.
I would like to identify low enery frame I order to Skip over them on
transmition.
I'am using black box encoder/decoder so I can't get to the raw voice.

Is there any indication of frame enrgy?

how, in general can energy be built?
can I assume that the gain composed of the filter gain and the excitation
gain?

any other Idea on how to calculate an approximation of the gain?
thanks,
nir

______________________________
New Year Gift for Members of DSPRelated.com.  Details here.

(You need to be a member of speechcoding -- send a blank email to speechcoding-subscribe@yahoogroups.com )

Nir Elkayam-

> I am using AMR 4.75 and AMR 5.15 for Voip application over very small
> bandwidth line.
> I would like to identify low enery frame I order to Skip over them on
> transmition.
> I'am using black box encoder/decoder so I can't get to the raw voice.

Are AMR output frames stand-alone; i.e. independent of each other?  My understanding is this is
not the case, that
some parameters are interpolated between frames, which would mean that skipping output frames
could have a degrading
effect on voice quality.

-Jeff

> Is there any indication of frame enrgy?
>
> how, in general can energy be built?
> can I assume that the gain composed of the filter gain and the excitation
> gain?
>
> any other Idea on how to calculate an approximation of the gain?
------------------------------------

______________________________
New Year Gift for Members of DSPRelated.com.  Details here.

(You need to be a member of speechcoding -- send a blank email to speechcoding-subscribe@yahoogroups.com )

Hi nir,
 
In my opinion, you should able the DTX mode in the encoder command line if
you want to skip some low energy frame. DTX mode will transmit only small
bits to the decoder. So the average bitrates will be lower than single mode.
And there is almost no degradation happens. Hope this information will help
u.
 
Dejun
 
  _____  

~{7"<~HK#?~}: s...@yahoogroups.com [mailto:s...@yahoogroups.com]
~{4z1m#?~} nir elkayam
~{7"KMJ1<d#?~}: 2008~{Dj#?~}9~{TB#?~}1~{HU#?~} 20:16
~{JU<~HK#?~}: s...@yahoogroups.com
~{VwLb#?~}: [speechcoding] Identify low energy AMR frames
 
hi,
 
I am using AMR 4.75 and AMR 5.15 for Voip application over very small
bandwidth line.
I would like to identify low enery frame I order to Skip over them on
transmition.
I'am using black box encoder/decoder so I can't get to the raw voice.
 
Is there any indication of frame enrgy?
 
how, in general can energy be built? 
can I assume that the gain composed of the filter gain and the excitation
gain?
 
any other Idea on how to calculate an approximation of the gain?

thanks,
nir
 
------------------------------------

______________________________
New Year Gift for Members of DSPRelated.com.  Details here.

(You need to be a member of speechcoding -- send a blank email to speechcoding-subscribe@yahoogroups.com )

That is VAD detection, you need to check from vocoder if there is voice act=
ivity. For wireless application in DTX mode, there are some extra code set =
like SID_UPDATE/SID_ONSET etc which will udpate the background noise level =
and handshake the voice/silence period transition. The degradation is small=
.=20

Yueheng
--- On Mon, 9/1/08, =E5=BC=A0=E5=BE=B7=E5=86=9B <z...@huawei.com> wro=
te:

> From: =E5=BC=A0=E5=BE=B7=E5=86=9B <z...@huawei.com>
> Subject: Re: [speechcoding] Identify low energy AMR frames
> To: "'nir elkayam'" <n...@gmail.com>
> Cc: s...@yahoogroups.com
> Date: Monday, September 1, 2008, 8:33 PM
> Hi nir,
>=20=20
> In my opinion, you should able the DTX mode in the encoder
> command line if
> you want to skip some low energy frame. DTX mode will
> transmit only small
> bits to the decoder. So the average bitrates will be lower
> than single mode.
> And there is almost no degradation happens. Hope this
> information will help
> u.
>=20=20
> Dejun
>=20=20
>   _____=20=20
>=20
> =E5=8F=91=E4=BB=B6=E4=BA=BA=EF=BC=9F: s...@yahoogroups.com
> [mailto:s...@yahoogroups.com]
> =E4=BB=A3=E8=A1=A8=EF=BC=9F nir elkayam
> =E5=8F=91=E9=80=81=E6=97=B6=E9=97=B4=EF=BC=9F: 2008=E5=B9=B4=EF=BC=9F9=E6=
=9C=88=EF=BC=9F1=E6=97=A5=EF=BC=9F 20:16
> =E6=94=B6=E4=BB=B6=E4=BA=BA=EF=BC=9F: s...@yahoogroups.com
> =E4=B8=BB=E9=A2=98=EF=BC=9F: [speechcoding] Identify low energy AMR frame=
s
>=20=20
> hi,
>=20=20
> I am using AMR 4.75 and AMR 5.15 for Voip application over
> very small
> bandwidth line.
> I would like to identify low enery frame I order to Skip
> over them on
> transmition.
> I'am using black box encoder/decoder so I can't get
> to the raw voice.
>=20=20
> Is there any indication of frame enrgy?
>=20=20
> how, in general can energy be built?=20
> can I assume that the gain composed of the filter gain and
> the excitation
> gain?
>=20=20
> any other Idea on how to calculate an approximation of the
> gain?
>=20=20
>=20=20
> thanks,
> nir
>=20=20
>=20
>

------------------------------------

NEW!  You can now post a message or access and search the archives of this =
group on DSPRelated.com:
http://www.dsprelated.com/groups/speechcoding/1.php



______________________________
New Year Gift for Members of DSPRelated.com.  Details here.

(You need to be a member of speechcoding -- send a blank email to speechcoding-subscribe@yahoogroups.com )