There are 3 messages totalling 251 lines in this issue.
Topics of the day:
1. The Auditory Continuity Illusion/Temporal Induction: Expanding the
2. PEAQ Advanced model
3. R-SPIN transcript
Date: Fri, 16 Dec 2005 11:25:13 +0100
From: Christopher Petkov
Subject: The Auditory Continuity Illusion/Temporal Induction: Expanding the Discussion
It's been great to see the recent discussions on the continuity illusion
and temporal induction from everyone. I figured since we got called by name
it was probably time to at least introduce ourselves to the list. The rest
are just some of my own thoughts on the previous discussions.
As Eli (Nelken) wrote, Mitch Sutter, myself and Kevin O'Connor have been
working on understanding a neurophysiological basis for the continuity
illusion in A1 in awake macaques. In our J. Neurosci 2003 paper we found
behavioral support for the illusion occurring with macaques which gave us
the green flag (so to say) to pursue it neurophysiologically. (In hindsight
I think we recorded from A1 because we thought it was named after Al
Bregman.) Anyhow, we're hoping you'll have something to read on this soon
enough, so I won't go into details here. But the enthusiasm is nice to see.
In any case we had a poster at APAN this year in D.C. tha t gave a glimpse of
some of the findings.
My impression is that the neurophysiological bases behind many processes
of auditory scene analysis are not that well known, or are at least
beginning to be understood. Behavior provides a great basis to guide these
approaches and Dick (Warren) and Al (Bregman) along with many others (Bob
Carlyon, etc.) have an incredible body of work on this. But as for the
other methods, Mitch Steinschneider and his group (neurophysiology in
macaques) and Christophe Micheyl and collaborators (MEG in humans and
neurophysiology in macaques) have been tackling perceptual streaming in
primates. Christophe Micheyl and Bob Carlyon also have a very nice paper on
the continuity illusion using EEG and the mismatch-negativity in humans.
Eli and his group also neurophysiologically address aspects of segregating
sound 'objects', as he noted. And others (too many to mention) have been
addressing aspects of these or r elated scene analysis issues using fMRI, EEG
and MEG in humans and behavior or electrophysiology in various species.
Certainly this work could be more extensive, but clearly many of the current
neuroscientific techniques are being used.
We'll, of course, in the future go a long way toward addressing some of
the issues that were brought up by Dick and others. Simply, we need more
detail from the various techniques on how all levels of the auditory system
contribute toward segregating sound mixtures and how perceptions are shaped
during different processes (illusory or not). The discussion so far has
centered specifically on auditory continuity, but streaming and continuity
are just two models or descriptions of natural abilities of a working (dare
we say 'normal'?) auditory system. Even Al himself might tell you that
there's a relationship between streaming and continuity (he's got work on
this). Thus many of the questions and issue s that were brought up are
certainly more generally applicable to scene analysis.
Additionally, the better that we know how the typical auditory system
solves these problems, the better in position we'll be to understand how
perceptions differ for impaired listeners. In this direction there's some
behavioral literature on scene analysis, including our work on dyslexics
(Sutter et al., 2000, Petkov et al., 2005). In those studies, we used a
modified perceptual streaming paradigm to approximate the source of
dyslexics' perceptual grouping impairments. Here, saying the impairment is
all over the periphery and brain is not so useful since even if everything
is affected, different areas are likely differently functionally affected.
Behavioral results can address this to some extent but then other methods
will have to step in.
It will be nice to see how groups come together on these issues since
each technique provides its own description (and inherent bias) of what is
going on. Each method (including behavioral work) has a different scope on
what is going on in the brain, with its own advantages (see Chris Stecker's
and subsequent discussion on this for fMRI) and limitations. From the
perspective of electrophysiology, however, considering how long physiology
from one auditory area takes I'm hoping (and gambling) that something like
fMRI can help guide the approach for us or at least provide a more direct
comparison to human fMRI data. Thus I'm excited about the modeling of
auditory continuity by Fatima Husain, Barry Horwitz and their group. I do
see Dick's point about how subcortical auditory areas also need to be
considered in the modeling. But in regards to modeling for guiding human
fMRI (I think a main objective of their work), imaging subcortically is a
hurdle fMRI has yet to overcome.
There's of course much to be done. Yet if enthusiasm is a gauge of
things to come, then we will undoubtedly see further work (using everyone's
favorite technique) on many issues of auditory scene analysis in general,
including, of course, further discussion of what each method contributes. I
look forward to this.
Best wishes to everyone and happy holidays,
Christopher I. Petkov
Max Planck Institute for Biological Cybernetics
72076 Tuebingen, Germany
> Date: Wed, 14 Dec 2005 08:28:35 +0200
> From: Israel Nelken
> Subject: Re: The Auditory Continuity Illusion/Temporal Induction
> Dear all,
> There's some electrophysiological work in animals that has bearing
> on the issue of continuity. Mitch Sutter has strong evidence that the
> illusion is operative in macaques, and he has some accompanying
> electrophysiology (that has not been published yet to the best of my
> knowledge) showing correlates of induction in primary auditory cortex.
> We (Las et al. J. Neurosci. 2005) published data related to the coding
> of a pure tone in fluctuating masker. Although our main emphasis was on
> comodulation masking release, the results can be interpreted in terms of
> continuity. In short, the responses of neurons in A1 of cats to the
> interrupted noise were very strong and locked to the noise envelope.
> Adding a low-level tone close to the BF of the neurons suppressed the
> envelope locking, resulting in responses that were similar to those
> evoked by tones in silence. Thus, these neurons seem to reflect the
> perceived continuity of the tone, ignoring the noise. We have further
> demonstrated that neurons with these responses a re present in the
> auditory thalamus but not in the inferior colliculus. All of this would
> suggest that activity that reflects the continuity of the tone is
> already present in thalamus/primary auditory cortex (although
> anesthetized cats are certainly not awake humans). We don't know however
> whether this activity is generated there or whether we see a reflection
> of processing at higher brain areas.
> Israel Nelken
> Dept. of Neurobiology
> The Alexander Silberman Institute of Life Sciences
> Edmond Safra Campus, Givat Ram | Tel: Int-972-2-6584229
> Hebrew University | Fax: Int-972-2-6586077
> Jerusalem 91904, ISRAEL | Email: israel@xxxxxxxxxxxxx
Date: Fri, 16 Dec 2005 18:47: 06 +0100
From: Goran Bozidar Markovic
Subject: PEAQ Advanced model
Hello to all who are reading this.
I have implemented advanced model of PEAQ (ITU BS.1387-1) as part of my
master thesis and checked it many times. But I am unable to get ODG for
conformance test as it is given in table 23 of BS.1387.
Values of MOVs from FFT ear model match very closely, but values of MOVs
from filterbank differ a lot, especially AvgLinDistA.
In order to test my code, I have also implemented some parts of Basic
model - AvgModDiff1B and RmsNoiseLoudB. Values of those 2 MOVs differ
significantly from reference values in conformance test, but are almost
identical to values from EAQUAL and PQevalAudio.
Please help me to identify the problem.
Date: Fri, 16 Dec 2005 16:31:24 -0800
Subject: R-SPIN transcript
This is a multi-part message in MIME format.
Does anyone have a text file with the transcribed sentences of the R-SPIN
test they would be willing to share?
Does anyone have a text file =
with the transcribed sentences of the R-SPIN test they would be willing =
End of AUDITORY Digest - 15 Dec 2005 to 16 Dec 2005 (#2005-254)
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around