Re: P-Centers (RoyP )

Subject: Re: P-Centers From: RoyP <rdp1@xxxxxxxx> Date: Wed, 12 Oct 2016 14:44:09 +0100 List-Archive:<http://lists.mcgill.ca/scripts/wa.exe?LIST=AUDITORY> This is a multi-part message in MIME format. --------------FE2D47AB8CFD0310EB7078C0 Content-Type: text/plain; charset=windows-1252; format=flowed Content-Transfer-Encoding: quoted-printable X-MIME-Autoconverted: from 8bit to quoted-printable by edgeum2.it.mcgill.ca id u9CDiHCE008241 Dear Daniele, Tim Ives wrote a robust routine for specifying the p-centers of CV and=20 VC syllables in 2005. It is described in *I**ves, D. T., Smith, D. R. R. and Patterson, R. D.=20 (2005).*"Discrimination of speaker size from syllable=20 phrases,"/J.//Acoust//. Soc. Am.//*118*//(6), 3816-3822./ The relevant text reads The syllables were normalized by setting the RMS value in the region of=20 the vowel to a common value so that they were all perceived to have=20 about the same loudness.We also wanted to ensure that, when any=20 combination of the syllables was played in a sequence, they would be=20 perceived to proceed at a regular pace; an irregular sequence of=20 syllables causes an unwanted distraction. Accordingly, the positions of=20 the syllables within their files were adjusted so that their=20 perceptual-centers (P-centers) all occurred at the same time relative to=20 file onset. The algorithm for finding the P-centers was based on=20 procedures described by Marcus (1981) and Scott (1993), and it focuses=20 on vowel onsets. Vowelonset time was taken to be the time at which the=20 syllable first rises to 50 % of its maximum value over the frequency=20 range of 300-3000 Hz.To optimize the estimation of vowel onset time, the=20 syllable was filtered with a gammatone filterbank (Patterson /et al/.,=20 1992) having thirty channels spaced quasi-logarithmically over the=20 frequency range of 300-3000 Hz.The thirty channels were sorted in=20 descending order based on their maximum output value and the ten highest=20 were selected.The Hilbert envelope was calculated for these ten channels=20 and, for each, the time at which the level first rose to 50 % of the=20 maximum was determined; the vowel onset time was taken to be the mean of=20 these ten time values.The P-centre was determined from the vowel onset=20 time and the duration of the signal as described by Marcus (1981). The=20 P-center adjustment was achieved by the simple expedient of inserting=20 silence before and/or after the sound. After P-center correction the=20 length of each syllable, including the silence, was 683 ms. I can provide a pdf of the paper if you would like to see it. Sincerely, Roy P On 11/10/2016 15:41, SCHON Daniele wrote: > > Deal all, > > I wrote some Matlab code to automatically extract the perceptual=20 > centres from natural speech and I realized that there are several=20 > options and choices to be made. > > I would appreciate your input on this topic (ie how you would=20 > generally proceed after speech-syllable segmentation) and I'll try to=20 > summarize the comments in a future post with my implemented=20 > solution(s) and the link to the code (with all the bugs!). > > > best > > daniele > > > Daniele Schon /- Chercheur - INS INSTITUT DE NEUROSCIENCES DES SYSTEMES/ > INSERM & Aix-Marseille Universit=E9 - Timone - 27 BVD Jean Moulin -=20 > 13005 Marseille > T=E9l: +33(0)4 91 38 46 32 / 55 77 > Site : https://sites.google.com/site/danieleschon/home=20 > <https://sites.google.com/site/danieleschon/home> - Email :=20 > daniele.schon@xxxxxxxx <mailto:daniele.schon@xxxxxxxx> > /Afin de respecter l'environnement, merci de n'imprimer cet email que=20 > si n=E9cessaire./ > > --=20 Roy D. Patterson Department of Physiology, Development and Neuroscience University of Cambridge, Downing Street, Cambridge, CB2 3EG email: rdp1@xxxxxxxx phone +44 (1223) 333819 http://www.pdn.cam.ac.uk/directory/roy-patterson http://www.pdn.cam.ac.uk/groups/cnbh/ http://www.AcousticScale.org --------------FE2D47AB8CFD0310EB7078C0 Content-Type: text/html; charset=windows-1252 Content-Transfer-Encoding: quoted-printable X-MIME-Autoconverted: from 8bit to quoted-printable by edgeum2.it.mcgill.ca id u9CDiHCE008241 <html> <head> <meta content=3D"text/html; charset=3Dwindows-1252" http-equiv=3D"Content-Type"> </head> <body bgcolor=3D"#FFFFFF" text=3D"#000000"> Dear Daniele, Tim Ives wrote a robust routine for specifying the p-centers of CV and VC syllables in 2005. It is described in I<strong style=3D"font-weight: bold; color: rgb(34, 34, 34); font-family: verdana, arial, sans-serif; font-size: 12px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: rgb(255, 255, 255);">ves, D. T., Smith, D. R. R. and Patterson, R. D. (2005).</st= rong><span style=3D"color: rgb(34, 34, 34); font-family: verdana, arial, sans-serif; font-size: 12px; font-style: normal; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: rgb(255, 255, 255); display: inline !important; float: none;">=A0"Discrimination of speaker size from syllable phrases,"=A0<em style=3D"font-style: italic; color: rgb(34, 34, 34); font-family: verdana, arial, sans-serif; font-size: 12px; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: rgb(255, 255, 255);">J.=A0<i style=3D"color: rgb(34, 34, 34); font-family: verdana, arial, sans-serif; font-size: 12px; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: rgb(255, 255, 255);">Acoust<em style=3D"font-style: italic; color: rgb(34, 34, 34); font-family: verdana, arial, sans-serif; font-size: 12px; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: rgb(255, 255, 255);">. Soc. Am.=A0</= em><em style=3D"font-style: italic; color: rgb(34, 34, 34); font-family: verdana, arial, sans-serif; font-size: 12px; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: rgb(255, 255, 255);">118=A0<em style=3D"font-style: italic; color: rgb(34, 34, 34); font-family: verdana, arial, sans-serif; font-size: 12px; font-variant-ligatures: normal; font-variant-caps: normal; font-weight: normal; letter-spacing: normal; orphans: 2; text-align: start; text-indent: 0px; text-transform: none; white-space: normal; widows: 2; word-spacing: 0px; -webkit-text-stroke-width: 0px; background-color: rgb(255, 255, 255);">(6), 3816-3822. The relevant text reads <meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3Dwindows-1252"> <meta name=3D"ProgId" content=3D"Word.Document"> <meta name=3D"Generator" content=3D"Microsoft Word 15"> <meta name=3D"Originator" content=3D"Microsoft Word 15">      <style>  </style> The syllables were normalized by setting the RMS value in the region of the vowel to a common value so that they were all perceived to have about the same loudness.=A0 We also wanted to ensure that, when any combination of the syllables was played in a sequence, they would be perceived to proceed at a regular pace; an irregular sequence of syllables causes an unwanted distraction. Accordingly, the positions of the syllables within their files were adjusted so that their perceptual-centers (P-centers) all occurred at the same time relative to file onset. The algorithm for finding the P-centers was based on procedures described by Marcus (1981) and Scott (1993), and it focuses on vowel onsets. Vowel onset time was taken to be the time at which the syllable first rises to 50 % of its maximum value over the frequency range of 300-3000 Hz.=A0 To optimize the estimation of vowel onset time, the syllable was filtered with a gammatone filterbank </spa= n>(Patterson et al., 1992) = having thirty channels spaced quasi-logarithmically over the frequency range of 300-3000 Hz.=A0 The thirty channels were sorted in descending order based on their maximum output value and the ten highest were selected.=A0 The Hilbert envelope was calculated for these ten channels and, for each, the time at which the level first rose to 50 % of the maximum was determined; the vowel onset time was taken to be the mean of these ten time values.=A0 </span= >The P-centre was determined from the vowel onset time and the duration of the signal as described by Marcus (1981). The P-center adjustment was achieved by the simple expedient of inserting silence before and/or after the sound. After P-center correction the length of each syllable, including the silence, was 683 ms. I can provide a pdf of the paper if you would like to see it. Sincerely, Roy P On 11/10/2016 15:41, SCHON Daniele wrote: <blockquote cite=3D"mid:25303_1476247655_57FDC067_25303_162_1_1476196860853.52194@xxxxxxxx= amu.fr" type=3D"cite"> <meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3Dwindows-1252"> <style type=3D"text/css" style=3D"display:none"></style> Deal all, I wrote some Matlab code to automatically extract the perceptual centres from natural speech and I realized that there are several options and choices to be made. I would appreciate your input on this topic (ie how you would generally proceed after speech-syllable segmentation) and I'll try to summarize the comments in a future post with my implemented solution(s) and the link to the code (with all the bugs!). best daniele <div id=3D"Signature"> <div class=3D""> <div id=3D"div_contenu_0" style=3D"z-index:1000"><img moz-do-not-send=3D"true" id=3D"VisuLogoSign_0" src=3D"https://sesame.univ-amu.fr/Logos/logo_amu.png" height=3D"118" hspace=3D"0" align=3D"left" width=3D"227"> <div style=3D"line-height:22px">Danie= le Schon -=A0Chercheur - INS INSTITUT DE NEUROSCIENCES DES SYSTEMES</div> <div style=3D"line-height:22px">INSERM & Aix-Marseille Universit=E9 - Timone - 27 BVD Jean Moulin - 13005 Marseille</div> <div style=3D"line-height:22px">T=E9l: +33(0)4 91 38 46 32 / 55 77</div> <div style=3D"line-height:22px">Site=A0: <a moz-do-not-send=3D"true" id=3D"NoLP" href=3D"https://sites.google.com/site/danieleschon/home" title=3D"https://sites.google.com/site/danieleschon/home" style=3D"line-height:20px; color:#22bbea; font-size:12px; font-family:verdana"> https://sites.google.com/site/danieleschon/home</a> - Email=A0: <a moz-do-not-send=3D"true" id=3D"NoLP" href=3D"mailto:daniele.schon@xxxxxxxx" style=3D"line-height:20px; color:#22bbea; font-size:12px; font-family:verdana"> daniele.schon@xxxxxxxx</a></div> <div style=3D"line-height:22px">Afin de respecter l'environnement, merci de n'imprimer cet email que si n=E9cessaire.</div> </div> </div> </div> <div style=3D"word-wrap:break-word"> </div> </blockquote> <pre class=3D"moz-signature" cols=3D"72">--=20 Roy D. Patterson Department of Physiology, Development and Neuroscience University of Cambridge, Downing Street, Cambridge, CB2 3EG email: <a class=3D"moz-txt-link-abbreviated" href=3D"mailto:rdp1@xxxxxxxx= k">rdp1@xxxxxxxx</a> phone +44 (1223) 333819=20=20=20 <a class=3D"moz-txt-link-freetext" href=3D"http://www.pdn.cam.ac.uk/direc= tory/roy-patterson">http://www.pdn.cam.ac.uk/directory/roy-patterson</a> <a class=3D"moz-txt-link-freetext" href=3D"http://www.pdn.cam.ac.uk/group= s/cnbh/">http://www.pdn.cam.ac.uk/groups/cnbh/</a> <a class=3D"moz-txt-link-freetext" href=3D"http://www.AcousticScale.org">= http://www.AcousticScale.org</a></pre> </body> </html> --------------FE2D47AB8CFD0310EB7078C0--

This message came from the mail archive
/var/www/html/postings/2016/
maintained by:

DAn Ellis <dpwe@ee.columbia.edu>
Electrical Engineering Dept., Columbia University