Re: Subject: Auditory Illusions ("Beerends, J.G. (John)" )


Subject: Re: Subject: Auditory Illusions
From:    "Beerends, J.G. (John)"  <john.beerends@xxxxxxxx>
Date:    Wed, 3 Aug 2011 08:01:27 +0000
List-Archive:<http://lists.mcgill.ca/scripts/wa.exe?LIST=AUDITORY>

--_000_E64149642826D74489BC7362384E1BFC63BA8AEXCMBX01tsntnonl_ Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable Dear Imran, Nedra, All, Based on the ongoing discussion we may want to introduce three levels of au= ditory (visual) illusions: 1) Illusions that can be explained by peripheral processing 2) Illusions that can be explained by central processing 3) Illusions that can be explained by cognition Hearing words that are not there (one of the Deutsch illusions) is a level = 3 auditory illusion, Mc Gurk also. Pitch perception (missing fundamental) a level 2 auditory illusion. Parallel lines that look non-parallel is a level 1 visual illusion. The old lady / young woman illusion is a level 3 visual illusion. Like testing the Mc Gurk effect in a noisy environment may be a useful way = to understand auditory visual speech perception in HA users it would be interesting to know if HA users show a decrease in cen= tral pitch processing capabilities by running the Houtsma left right integr= ation experiment. If we are also discussing what would benefit HA users most I would say that= optimizing speech intelligibilty is too limited, subjects are mostly bothered by poor speech quality (quality and intelligib= ility are definitely not the same), poor background noise quality (daily so= unds like birds, cars, slamming doors etc.) and poor music quality. John Beerends TNO The Netherlands ________________________________ From: AUDITORY - Research in Auditory Perception [mailto:AUDITORY@xxxxxxxx= ILL.CA] On Behalf Of Imran Dhamani Sent: woensdag 3 augustus 2011 7:05 To: AUDITORY@xxxxxxxx Subject: Subject: Auditory Illusions Dear Nedra, I have a slightly different understanding in the regard of auditory illusio= ns such as the Mc Gurk effect not contributing to hearing aid design. I thi= nk these illusions may have more to contribute to not only hearing aid desi= gn but also to psycho-acoustics. One of the key uses of the Mc Gurk effect = was in terms of studying the re-organisation/remapping process in the audit= ory and visual cortex which has been recently reported due to auditory depr= ivation (deafness) in hearing impaired children and the changes to the same= with cochlear implantation and I guess the same would apply to hearing aid= s as well (especially the ones with severe hearing loss). One of the other = interesting illusions (although not purely an auditory illusion - or maybe = with the current discussion that I have noticed in the list may be not even= an illusion) I remember working with hearing aid and cochlear implanted us= ers is the Kiki-Bouba effect which helps us study the sound-shape correspon= dences in the brain. One of the key missing link I think in the current wav= e of hearing aid technology which is mainly driven via psycho-acoustical di= fficulties that a hearing impaired user faces is the overall auditory proce= ssing (central auditory processing as well as Auditory- visual processing) = difficulties that a hearing loss impinges along-with the lower level psycho= -acoustic deficits like poor frequency selectivity, temporal resolution etc= . Maybe broadening the definition of psycho-acoustics a bit these aspects m= ay also be well under its domain. The subtle but existing difference betwee= n two hearing aid users with the same type of hearing loss characteristics = may well be in terms of their central auditory visual processing abilities = which might have been differently affected by the loss (although there may = be alternative aspects such as difference in some relatively less explored = low level psycho-acoustic abilities adding to it). In a nutshell what I wan= t to express is that studying these illusions is to somehow ensure that we = reach limitations which are imposed by an impaired ear and not those impose= d by inadequate amplification systems and may be try to find out and compen= sate for the additional auditory-visual processing/integration limitations = that the hearing loss has added using appropriate rehabilitative/training m= easures. I don't know why there were those strange characters visible in my previous= post for this topic so I thought of attaching that previous message at the= end as a paragraph for better readability. Apologies to all for any inconv= enience caused. Although I am not an expert in this area, but I have done some preliminary = (unpublished) work on the use of the Mc Gurk effect for the evaluation of c= urrent day multichannel digital hearing aids. Few of the basic assumptions = for the rationale and implications were as follows. One of the main benefit= s of audio visual integration is in difficult listening situations when eit= her the speech is degraded or there is background noise. The visual cue in = an auditory visual integration task is relatively unaffected by noise. Hear= ing impaired listeners especially those with relatively larger auditory dep= rivation periods rely more on visual cues (the AV balance is slightly tilte= d towards visual dominance) in speech perception than normal hearing listen= ers, thus making them better speech readers and also relatively poor AV int= egrators. The psycho-acoustical aspects of sensory-neural hearing loss sugg= est a reduction in the possibility of perceiving certain classes of speech = sounds especially in presence of noise. Moreover some of the speech sounds = may not be heard in noisy environments no matter how intensely they are amp= lified by the hearing aids. The Digital multichannel hearing aids may also = have some amount of internal distortion and delay due to digital processing= and filtering. In such scenarios the hearing aid user may benefit from the= visual cues provided by a speakers facial and lip movements which should r= elatively be unaffected by noise and thus boost speech perception for energ= etic masking at poor signal to noise ratios as well as in informational mas= king. We presented hearing aid users (relatively homogenous group in terms = of hearing aid used) the Mc Gurk stimuli in both congruent and in-congruent= conditions in quiet and in noise (three different SNRs) at comfortable lev= el at which the subjects scored > 70% on a screening test using PB words. A= criterion of 3/4th fusion responses was kept to determine the presence of = Mc Gurk effect. The results indicated that subjects with normal hearing per= formed better than the subjects using HA's in all conditions. The presence = of auditory and visual information simultaneously in the congruent conditio= n was beneficial for speech perception in quiet and in noise. Noise reduced= the subjects ability to perceive speech at poor SNR and had a more severe = impact on the performance of the HA users than in normal subjects. The Mc G= urk effect was absent in the HA users at the poorer SNRs compared to normal= listeners and in these conditions the HA users responses to the Mc Gurk st= imuli were mainly visually dominated. Background noise and increased listen= ing effort are significant factors influencing hearing-aid satisfaction and= one of the major reason for rejection of HA's. Testing Mc Gurk effect in a= noisy environment may be a useful way to understand auditory visual speech= perception in HA users and verify the benefits of aided AV speech percepti= on in noise in HA users. Some of the implications in terms of hearing aid r= ehabilitation for the same may be in terms of enhancing optimal HA fitting = to achieve not only good auditory perception in noise but also optimum audi= tory visual perception in noise and the emphasis on auditory training and t= he use of speech reading skills. The study that I have mentioned above was = in no way devoid of limitations like lesser sample size and the hearing aid= s were used at the same programs that the HA users were using in their ever= yday listening environments and thus all the user had different program set= tings. May be the rationale and implications might be of some interest to y= ou though. Regards, Imran Dhamani This e-mail and its contents are subject to the DISCLAIMER at http://www.tn= o.nl/emaildisclaimer --_000_E64149642826D74489BC7362384E1BFC63BA8AEXCMBX01tsntnonl_ Content-Type: text/html; charset="us-ascii" Content-Transfer-Encoding: quoted-printable <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN"> <html> <head> <meta http-equiv=3D"Content-Type" content=3D"text/html; charset=3Dus-ascii"> <meta name=3D"GENERATOR" content=3D"MSHTML 8.00.6001.19046"> </head> <body> <div dir=3D"ltr" align=3D"left"><font color=3D"#0000ff" size=3D"2" face=3D"= Arial">Dear&nbsp;<span class=3D"206515907-03082011">Imran, Nedra, </span>All,</font></div> <div><font color=3D"#0000ff" size=3D"2" face=3D"Arial"></font>&nbsp;</div> <div dir=3D"ltr" align=3D"left"><font color=3D"#0000ff" size=3D"2" face=3D"= Arial">Based on the ongoing discussion we may want to introduce three level= s of auditory (visual) illusions:<br> 1) Illusions that can be explained by peripheral processing <br> 2) Illusions that can be explained by central processing<br> 3) Illusions that can be explained by cognition </font></div> <div><font color=3D"#0000ff" size=3D"2" face=3D"Arial"></font>&nbsp;</div> <div dir=3D"ltr" align=3D"left"><font color=3D"#0000ff" size=3D"2" face=3D"= Arial">Hearing words that are not there (one of the Deutsch illusions) is a= level 3 auditory illusion, Mc Gurk also.<br> Pitch perception (missing fundamental) a level 2 auditory illusion.<br> Parallel lines that look non-parallel is a level 1 visual illusion.<br> The old lady / young woman illusion is a level 3 visual illusion.</font></d= iv> <div><font color=3D"#0000ff" size=3D"2" face=3D"Arial"></font>&nbsp;</div> <div dir=3D"ltr" align=3D"left"><font color=3D"#0000ff" size=3D"2" face=3D"= Arial">Like testing the Mc Gurk effect in a noisy environment may be a usef= ul way to understand auditory visual speech perception in <br> HA users it would be interesting to know if HA users show a decrease in cen= tral pitch processing capabilities by running the Houtsma left right integr= ation experiment.</font></div> <div><font color=3D"#0000ff" size=3D"2" face=3D"Arial"></font>&nbsp;</div> <div dir=3D"ltr" align=3D"left"><font color=3D"#0000ff" size=3D"2" face=3D"= Arial">If we are also discussing what would benefit HA users most I would s= ay that optimizing speech intelligibilty is too limited, <br> subjects are mostly bothered by poor speech quality (quality and intelligib= ility are definitely not the same), poor background noise quality (daily so= unds like birds, cars, slamming doors etc.) and poor music quality.</font><= /div> <div><font color=3D"#0000ff" size=3D"2" face=3D"Arial"></font>&nbsp;</div> <div dir=3D"ltr" align=3D"left"><font color=3D"#0000ff" size=3D"2" face=3D"= Arial">John Beerends<br> TNO</font></div> <div dir=3D"ltr" align=3D"left"><span class=3D"206515907-03082011"></span><= font color=3D"#0000ff" size=3D"2" face=3D"Arial">T<span class=3D"206515907-= 03082011">he Netherlands</span><br> </font></div> <br> <div dir=3D"ltr" lang=3D"en-us" class=3D"OutlookMessageHeader" align=3D"lef= t"> <hr tabindex=3D"-1"> <font size=3D"2" face=3D"Tahoma"><b>From:</b> AUDITORY - Research in Audito= ry Perception [mailto:AUDITORY@xxxxxxxx <b>On Behalf Of </b>Imran Dhamani<br> <b>Sent:</b> woensdag 3 augustus 2011 7:05<br> <b>To:</b> AUDITORY@xxxxxxxx<br> <b>Subject:</b> Subject: Auditory Illusions<br> </font><br> </div> <div></div> <div style=3D"BACKGROUND-COLOR: #fff; FONT-FAMILY: times new roman, new yor= k, times, serif; COLOR: #000; FONT-SIZE: 12pt"> <div style=3D"TEXT-ALIGN: justify" class=3D"yiv580931635MsoNormal">Dear Ned= ra,</div> <div style=3D"TEXT-ALIGN: justify" class=3D"yiv580931635MsoNormal"><span></= span>I have a slightly different understanding in the regard of auditory il= lusions such as the Mc Gurk effect not contributing to hearing aid design. = I think these illusions may have more to contribute to not only hearing aid design but also to psycho-acoustics.= One of the key uses of the Mc Gurk effect was in terms of studying the re-= organisation/remapping process in the auditory and visual cortex which has = been recently reported due to auditory deprivation (deafness) in hearing impaired children and the changes to the= same with cochlear implantation and I guess the same would apply to hearin= g aids as well (especially the ones with severe hearing loss). One of the o= ther interesting illusions (although not purely an auditory illusion &#8211; or maybe with the current discussi= on that I have noticed in the list may be not even an illusion) I remember = working with hearing aid and cochlear implanted users is the Kiki-Bouba eff= ect which helps us study the sound-shape correspondences in the brain. One of the key missing link I think in the c= urrent wave of hearing aid technology which is mainly driven via psycho-aco= ustical difficulties that a hearing impaired user faces is the overall audi= tory processing (central auditory processing as well as Auditory- visual processing) difficulties that a hea= ring loss impinges along-with the lower level psycho-acoustic deficits like= poor frequency selectivity, temporal resolution etc. Maybe broadening the = definition of psycho-acoustics a bit these aspects may also be well under its domain. The subtle but existi= ng difference between two hearing aid users with the same type of hearing l= oss characteristics may well be in terms of their central auditory visual p= rocessing abilities which might have been differently affected by the loss (although there may be alternat= ive aspects such as difference in some relatively less explored low level p= sycho-acoustic abilities adding to it). In a nutshell what I want to expres= s is that studying these illusions is to somehow ensure that we reach limitations which are imposed by an imp= aired ear and not those imposed by inadequate amplification systems and may= be try to find out and compensate for the additional auditory-visual proce= ssing/integration limitations that the hearing loss has added using appropriate rehabilitative/training measu= res. </div> <div style=3D"TEXT-ALIGN: justify" class=3D"yiv580931635MsoNormal"><br> I don&#8217;t know why there were those strange characters visible in my pr= evious post for this topic so I thought of attaching that previous message = at the end as a paragraph for better readability. Apologies to all for any = inconvenience caused. </div> <div style=3D"TEXT-ALIGN: justify" class=3D"yiv580931635MsoNormal">&nbsp;</= div> <div style=3D"TEXT-ALIGN: justify" class=3D"yiv580931635MsoNormal">Although= I am not an expert in this area, but I have done some preliminary (unpubli= shed) work on the use of the Mc Gurk effect for the evaluation of current d= ay multichannel digital hearing aids. Few of the basic assumptions for the rationale and implications were as fo= llows. One of the main benefits of audio visual integration is in difficult= listening situations when either the speech is degraded or there is backgr= ound noise. The visual cue in an auditory visual integration task is relatively unaffected by noise. Hearin= g impaired listeners especially those with relatively larger auditory depri= vation periods rely more on visual cues (the AV balance is slightly tilted = towards visual dominance) in speech perception than normal hearing listeners, thus making them better speech r= eaders and also relatively poor AV integrators. The psycho-acoustical aspec= ts of sensory-neural hearing loss suggest a reduction in the possibility of= perceiving certain classes of speech sounds especially in presence of noise. Moreover some of the speech sounds= may not be heard in noisy environments no matter how intensely they are am= plified by the hearing aids. The Digital multichannel hearing aids may also= have some amount of internal distortion and delay due to digital processing and filtering. In such scenarios the h= earing aid user may benefit from the visual cues provided by a speakers fac= ial and lip movements which should relatively be unaffected by noise and th= us boost speech perception for energetic masking at poor signal to noise ratios as well as in informational masking= . We presented hearing aid users (relatively homogenous group in terms of h= earing aid used) the Mc Gurk stimuli in both congruent and in-congruent con= ditions in quiet and in noise (three different SNRs) at comfortable level at which the subjects scored &gt; 70%= on a screening test using PB words. A criterion of 3/4th fusion responses = was kept to determine the presence of Mc Gurk effect. The results indicated= that subjects with normal hearing performed better than the subjects using HA&#8217;s in all conditions. The= presence of auditory and visual information simultaneously in the congruen= t condition was beneficial for speech perception in quiet and in noise. Noi= se reduced the subjects ability to perceive speech at poor SNR and had a more severe impact on the performance of the = HA users than in normal subjects. The Mc Gurk effect was absent in the HA u= sers at the poorer SNRs compared to normal listeners and in these condition= s the HA users responses to the Mc Gurk stimuli were mainly visually dominated. Background noise and incre= ased listening effort are significant factors influencing hearing-aid satis= faction and one of the major reason for rejection of HA&#8217;s. Testing Mc= Gurk effect in a noisy environment may be a useful way to understand auditory visual speech perception in HA user= s and verify the benefits of aided AV speech perception in noise in HA user= s. Some of the implications in terms of hearing aid rehabilitation for the = same may be in terms of enhancing optimal HA fitting to achieve not only good auditory perception in noise b= ut also optimum auditory visual perception in noise and the emphasis on aud= itory training and the use of speech reading skills. The study that I have = mentioned above was in no way devoid of limitations like lesser sample size and the hearing aids were used at t= he same programs that the HA users were using in their everyday listening e= nvironments and thus all the user had different program settings. May be th= e rationale and implications might be of some interest to you though.</div> <div style=3D"TEXT-ALIGN: justify" class=3D"yiv580931635MsoNormal">&nbsp;</= div> <div style=3D"TEXT-ALIGN: justify" class=3D"yiv580931635MsoNormal">Regards,= </div> <div style=3D"TEXT-ALIGN: justify" class=3D"yiv580931635MsoNormal">Imran Dh= amani</div> <div>&nbsp;</div> <div><br> </div> </div> <font face=3D"monospace">This e-mail and its contents are subject to the DI= SCLAIMER at http://www.tno.nl/emaildisclaimer</font></body> </html> --_000_E64149642826D74489BC7362384E1BFC63BA8AEXCMBX01tsntnonl_--


This message came from the mail archive
/var/www/postings/2011/
maintained by:
DAn Ellis <dpwe@ee.columbia.edu>
Electrical Engineering Dept., Columbia University