Re: speech degradation in a GSM network transmission (Ira Panzer )


Subject: Re: speech degradation in a GSM network transmission
From:    Ira Panzer  <ilpanzer@xxxxxxxx>
Date:    Thu, 6 Mar 2008 18:53:26 -0600
List-Archive:<http://lists.mcgill.ca/scripts/wa.exe?LIST=AUDITORY>

This is a multi-part message in MIME format. ------=_NextPart_000_001A_01C87FBB.5C8AEEC0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit Content-Disposition: inline Hi All, ITU-T Recommendation G.191, Software Tools for Speech and Audio Coding Standardization includes the mnrudemo program. The recommendation is available from the ITU website (www.itu.ch). This library of software tools is very useful if doing work in speech and audio coding. Regards, Ira Panzer Dynastat -----Original Message----- From: AUDITORY - Research in Auditory Perception [mailto:AUDITORY@xxxxxxxx Behalf Of adil raja Sent: Thursday, March 06, 2008 6:18 PM To: AUDITORY@xxxxxxxx Subject: Re: speech degradation in a GSM network transmission Hi, You may want to try to use signal-correlated noise. ITU-T recommends on using Modulated Noise Reference Unit (MNRU) for this (please google it). I guess it is also available in ITU-T's standard tools library (?) (dont have refs at the moment). However, it is quite easy to implement it in Matlab. Signal correlated noise also affects the spectral envelope of the signal. As a matter of fact various researchers have endeavored to model its effect using measures such as cepstral deviation etc. Another way to induce spectral distortion in to the signal is to treat it with a codec (after all thats what codecs are good at (among other things)). Signal correlated noise is a feature of waveform codecs (such as u-law compander etc). I hope that helps? Regards, Adil Raja ----- Original Message ---- From: Faten B.A. <fatenpesq@xxxxxxxx> To: AUDITORY@xxxxxxxx Sent: Wednesday, March 5, 2008 11:48:19 PM Subject: [AUDITORY] speech degradation in a GSM network transmission Hi list, I want to introduce some artificial degradation to a speech signal in order to simulate speech degradation in a GSM network transmission. The addition of a background noise seems to be simple but I don't know which type or noise distribution I have to use. I want also introduce frequency degradation and I have no idea about the way this must be done. Can anyone advice me or send me some documentations about this topic? Best Regards. F.B.A. ---------------------------------------------------------------------------- -- Never miss a thing. Make Yahoo your homepage. ------=_NextPart_000_001A_01C87FBB.5C8AEEC0 Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <HTML><HEAD> <META http-equiv=3DContent-Type content=3D"text/html; charset=3Dus-ascii"> <STYLE type=3Dtext/css>DIV { MARGIN: 0px } </STYLE> <META content=3D"MSHTML 6.00.2800.1589" name=3DGENERATOR></HEAD> <BODY> <DIV><SPAN class=3D812123200-07032008><FONT face=3DArial color=3D#0000ff si= ze=3D2>Hi=20 All,</FONT></SPAN></DIV> <DIV><SPAN class=3D812123200-07032008><FONT face=3DArial color=3D#0000ff=20 size=3D2></FONT></SPAN>&nbsp;</DIV> <DIV><SPAN class=3D812123200-07032008><FONT face=3DArial color=3D#0000ff si= ze=3D2>ITU-T=20 Recommendation&nbsp;G.191, Software Tools for Speech and Audio Coding=20 Standardization&nbsp;includes the mnrudemo program.&nbsp; The recommendatio= n is=20 available from the ITU website (<A=20 href=3D"http://www.itu.ch">www.itu.ch</A>).&nbsp; This library of software = tools=20 is very useful if doing&nbsp;work in speech and audio coding.&nbsp;=20 </FONT></SPAN></DIV> <DIV><SPAN class=3D812123200-07032008><FONT face=3DArial color=3D#0000ff=20 size=3D2></FONT></SPAN>&nbsp;</DIV> <DIV><SPAN class=3D812123200-07032008><FONT face=3DArial color=3D#0000ff=20 size=3D2>Regards,</FONT></SPAN></DIV> <DIV><SPAN class=3D812123200-07032008><FONT face=3DArial color=3D#0000ff si= ze=3D2>Ira=20 Panzer</FONT></SPAN></DIV> <DIV><SPAN class=3D812123200-07032008><FONT face=3DArial color=3D#0000ff=20 size=3D2>Dynastat</FONT></SPAN></DIV> <BLOCKQUOTE> <DIV class=3DOutlookMessageHeader dir=3Dltr align=3Dleft><FONT face=3DTah= oma=20 size=3D2>-----Original Message-----<BR><B>From:</B> AUDITORY - Research i= n=20 Auditory Perception [mailto:AUDITORY@xxxxxxxx<B>On Behalf Of </B>= adil=20 raja<BR><B>Sent:</B> Thursday, March 06, 2008 6:18 PM<BR><B>To:</B>=20 AUDITORY@xxxxxxxx<BR><B>Subject:</B> Re: speech degradation in a G= SM=20 network transmission<BR><BR></FONT></DIV> <DIV=20 style=3D"FONT-SIZE: 12pt; FONT-FAMILY: times new roman, new york, times, = serif"> <DIV=20 style=3D"FONT-SIZE: 12pt; FONT-FAMILY: times new roman,new york,times,ser= if">Hi,<BR>&nbsp;&nbsp;=20 You may want to try to use signal-correlated noise. ITU-T recommends on u= sing=20 Modulated Noise Reference Unit (MNRU) for this (please google it). I gues= s it=20 is also available in ITU-T's standard tools library (?) (dont have refs a= t the=20 moment). However, it is quite easy to implement it in Matlab. Signal=20 correlated noise also affects the spectral envelope of the signal. As a m= atter=20 of fact various researchers have endeavored to model its effect using mea= sures=20 such as cepstral deviation etc. Another way to induce spectral distortion= in=20 to the signal is to treat it with a codec (after all thats what codecs ar= e=20 good at (among other things)). Signal correlated noise is a feature of=20 waveform codecs (such as u-law compander etc).<BR><BR>I hope that=20 helps?<BR>Regards,<BR>Adil Raja<BR><BR> <DIV=20 style=3D"FONT-SIZE: 12pt; FONT-FAMILY: times new roman,new york,times,ser= if">-----=20 Original Message ----<BR>From: Faten B.A. &lt;fatenpesq@xxxxxxxx&gt;<BR>= To:=20 AUDITORY@xxxxxxxx<BR>Sent: Wednesday, March 5, 2008 11:48:19=20 PM<BR>Subject: [AUDITORY] speech degradation in a GSM network=20 transmission<BR><BR> <DIV> <P class=3DMsoNormal=20 style=3D"BACKGROUND: white 0% 50%; MARGIN: 0cm 0cm 0pt; moz-background-cl= ip: -moz-initial; moz-background-origin: -moz-initial; moz-background-inlin= e-policy: -moz-initial"><SPAN=20 lang=3DEN-GB style=3D"FONT-FAMILY: Arial">Hi list,</SPAN></P> <P class=3DMsoNormal=20 style=3D"BACKGROUND: white 0% 50%; MARGIN: 0cm 0cm 0pt; moz-background-cl= ip: -moz-initial; moz-background-origin: -moz-initial; moz-background-inlin= e-policy: -moz-initial"><SPAN=20 lang=3DEN-GB style=3D"FONT-FAMILY: Arial"></SPAN>&nbsp;</P> <P class=3DMsoNormal=20 style=3D"BACKGROUND: white 0% 50%; MARGIN: 0cm 0cm 0pt; moz-background-cl= ip: -moz-initial; moz-background-origin: -moz-initial; moz-background-inlin= e-policy: -moz-initial"><SPAN=20 lang=3DEN-GB style=3D"FONT-FAMILY: Arial">I want to introduce some artifi= cial=20 degradation to a speech signal in order to simulate speech degradation in= a=20 GSM network transmission.</SPAN></P> <P class=3DMsoNormal=20 style=3D"BACKGROUND: white 0% 50%; MARGIN: 0cm 0cm 0pt; moz-background-cl= ip: -moz-initial; moz-background-origin: -moz-initial; moz-background-inlin= e-policy: -moz-initial"><SPAN=20 lang=3DEN-GB style=3D"FONT-FAMILY: Arial">The addition of a background no= ise seems=20 to be simple but I don't know which type or noise distribution I have to= =20 use.</SPAN></P> <P class=3DMsoNormal=20 style=3D"BACKGROUND: white 0% 50%; MARGIN: 0cm 0cm 0pt; moz-background-cl= ip: -moz-initial; moz-background-origin: -moz-initial; moz-background-inlin= e-policy: -moz-initial"><SPAN=20 lang=3DEN-GB style=3D"FONT-FAMILY: Arial">I want also introduce frequency= =20 degradation and I have no idea about the way this must be done.</SPAN></P> <P class=3DMsoNormal=20 style=3D"BACKGROUND: white 0% 50%; MARGIN: 0cm 0cm 0pt; moz-background-cl= ip: -moz-initial; moz-background-origin: -moz-initial; moz-background-inlin= e-policy: -moz-initial"><SPAN=20 lang=3DEN-GB style=3D"FONT-FAMILY: Arial">Can anyone advice me or send me= some=20 documentations about this topic?</SPAN></P> <P class=3DMsoNormal=20 style=3D"BACKGROUND: white 0% 50%; MARGIN: 0cm 0cm 0pt; moz-background-cl= ip: -moz-initial; moz-background-origin: -moz-initial; moz-background-inlin= e-policy: -moz-initial"><SPAN=20 lang=3DEN-GB style=3D"FONT-FAMILY: Arial"></SPAN>&nbsp;</P> <P class=3DMsoNormal=20 style=3D"BACKGROUND: white 0% 50%; MARGIN: 0cm 0cm 0pt; moz-background-cl= ip: -moz-initial; moz-background-origin: -moz-initial; moz-background-inlin= e-policy: -moz-initial"><SPAN=20 lang=3DEN-GB style=3D"FONT-FAMILY: Arial">Best Regards.</SPAN></P> <P class=3DMsoNormal=20 style=3D"BACKGROUND: white 0% 50%; MARGIN: 0cm 0cm 0pt; moz-background-cl= ip: -moz-initial; moz-background-origin: -moz-initial; moz-background-inlin= e-policy: -moz-initial"><SPAN=20 lang=3DEN-GB=20 style=3D"FONT-FAMILY: Arial">F.B.A.</SPAN></P></DIV></DIV><BR></DIV></DIV= ><BR> <HR SIZE=3D1> Never miss a thing. <!-- <A=20 href=3D"http://us.rd.yahoo.com/evt=3D51438/*http://www.yahoo.com/r/hs"> -= ->Make Yahoo=20 your homepage.<!-- </A> --> </BLOCKQUOTE></BODY></HTML> ------=_NextPart_000_001A_01C87FBB.5C8AEEC0--


This message came from the mail archive
http://www.auditory.org/postings/2008/
maintained by:
DAn Ellis <dpwe@ee.columbia.edu>
Electrical Engineering Dept., Columbia University