1. Testing, testing – can you hear me, understand
me, remember me?
Mai-Britt Beldam
Saint-Gobain Ecophon AS, Hammerholmen 18 E, 2650 Hvidovre, Denmark.
Summary
The primary indicator of room acoustic response still is Reverberation Time (RT) and this acoustic
descriptor is also known be the only demand/descriptor in building regulations for room acoustics
in schools and day care centres in many countries. In comparison, research for many years and
especially many theoretical studies have shown that RT alone can be insufficient to describe the
acoustic conditions in non-diffuse environments, mainly in classrooms and learning environments
where the majority of absorption is on one surface. The typical solution is a suspended absorbing
ceiling and this exact placement of the absorption material normally leads to a non-diffuse sound
decay which is one of the reasons why the descriptor RT alone can be insufficient to evaluate the
room. Previous work has shown that Speech Clarity (C50)/ Deutlichkeit (D50) and sound strength
in combination with RT are very relevant descriptors but we still know very little about how
children’s performance practically is affected when C/D50 and sound strength are changed and RT
stays the same. This pilot study investigates 8 Danish children’s performance in two rooms with
same RT but different values on C50 and sound strength. The children are tested in working
memory and listening comprehension with/without background noise and it is concluded that there
is a trend that RT alone is not sufficient: The children perform differently in the two rooms and
they are much more negatively affected by background noise, when C50 drops and SPL increases.
PACS no. 43.55.Br, 43.55.Hy, 43.55.Gx
1 Introduction1
Teaching has changed and so has the attitude of
our children. The school is no longer the place for
endless teacher monologues and our children are
expected to participate in discussions, in group
work and in cross discipline projects. Traditional
teaching today is not the same as it was only 10
years ago and the acoustical demands for schools
are very often inadequate. The common way to
describe classroom acoustics is to use the acoustic
descriptor reverberation time (RT) and this
parameter is the preferable descriptor in building
regulations for learning environments. Rasmussen
et al [1] give an overview of the trend in RT
criteria for classrooms in the Nordic countries and
they describe the trend to be clear: Shorter RT is
assumed to be better. In Denmark RT in
classrooms for traditional teaching has to reach ≤
0.6 sec. and group rooms and rooms for students
with special needs have to reach ≤ 0.4 sec.
1
(c) European Acoustics Association
according to the Danish Building Regulations [2].
Despite these relatively short RTs the end users –
both children and teachers - still complain about
high noise levels and poor sound environments [3,
4]. In 2012 the ISO standard for open offices [5]
was published and RT was not part of the
guideline. In Denmark, we still see a lot of open
offices that are acoustically analyzed by using RT
but more and more the demands are set according
to the new standard because of the changing of the
use of the office and acoustical awareness growing
in the market. Like the office has changed so have
the learning environments and it might be relevant
to include more acoustic descriptors to secure low
sound pressure levels (SPL) and great speech
intelligibility in classrooms weather we have one
speaker at the time or several.
The discussion about the relation between early
and late reflections is not new and ever since the
investigations of Joseph Henry, [6, 7] acousticians
have discussed the importance. An argument to
continue this discussion could be to accept that
teaching and learning styles have developed in the
same time as children with special needs get

2. FORUM ACUSTICUM 2014 M. Beldam: Testing, testing - can you hear me, understand me remember me?
7–12 September, Krakow
included in the traditional classroom. To secure a
suitable sound environment for everybody it seems
as if we have to consider both the patterns of the
reflections and the sound levels to be just as
important as RT. The cognitive demands on
children have never been higher and the children
of today need to manage multiple tasks such as
writing, reading, controlling modern technology,
engaging in discussions with classmates and
listening to the teachers voice while
simultaneously the children need to ignore
competing background noise from the classroom
and the hallway. This means a significant load on
their cognitive resources [8]. Bradley and Sato [9]
commented more than a decade ago that ‘A lack of
appreciation of the importance of early reflections
is no doubt responsible for some recommendations
for very short reverberation times for rooms for
speech’ and research has shown some cases of
problems with speakers comfort when
reverberation time gets too short and it would be
interesting to learn if a more detailed overview of
several descriptors would secure optimum room
acoustics in schools.
A lot of theoretical work has been done to find out
what acoustical descriptors are most beneficial but
practical research with small children is limited.
This paper seeks to investigate if other descriptors
should be included in the evaluation of the
acoustics of a room and practically how the
performance of 7-year olds is affected in situ when
RT stays the same but values for Clarity of Speech
(C50) and SPL are changed.
2 Background
Despite a long tradition of research showing that
other acoustic descriptors are necessary to secure
good classroom acoustics that supports teaching
and learning in regards to speech intelligibility and
sufficient signal-to-noise ratios, calculations and
measurements of RT are still used to evaluate
classrooms, group rooms etc. [10, 11, 12]. RT was
developed by Sabine in the 1890s and still remains
the preferred descriptor even though most
traditional classrooms cannot be described as a
diffused sound field since most of the absorption
material is on one surface; the ceiling. The diffuse
field is only a theoretical condition impossible to
obtain practically in reality. The decay will not
follow a straight line according to the theory but
will be split in an early part correlating more or
less to the theory and a late part with a longer
reverberation time. This is thoroughly discussed
and described by E. Nilsson [13, 14]. RT is
defined in ISO 3382-1 [15] as the time it takes for
sound source to decrease in level by 60 dB after
the source emission has stopped. RT is more
commonly measured over a 20 or 30 dB range
(T20 and T30) starting 5 dB below the initial level
and extrapolated to the full 60 dB range. The rate
of the first 10 dB of the decay is more related to
the perceived reverberance and can be measured
by the EDT [13, 14]. Starting 5 dB below the
initial level can be problematic since this part of
the decay contains a lot of information – both
direct sound and early reflections – important for
the perception of sound and speech clarity. The
human ear analyses so much more than the defined
RT but the simplicity of the Sabine formula and
the fact that only one single number is ‘enough’ to
describe the acoustic quality of a room might be
the reason why RT – calculated (and/or measured)
– has been preferred for many years [16].
Considering how the teaching methods have
changed together with the attitude of children and
together with the development of acoustic
knowledge, room acoustic quality today needs to
be described with more than just RT. A
supplement to RT is suggested to be both Strength
(G) and C50 according to ISO 3382-1 [15]. C50
evaluates the effect of the room’s early reflections
whereas G measures the room’s contribution to the
noise level from a source and it is known that these
descriptors can be different from room to room
even if the late RT is the same [17]. Even though
the purpose of the mentioned ISO standard is
related to performance spaces rather than ordinary
rooms for speech, research has shown that both G
and C50 are good indicators for room acoustic
quality together with RT [18, 19]. Still we need
practical research discussing how children’s
performance is affected when RT stays the same
and G and C50 are changed and the purpose of this
paper and the underlying pilot project is to see if
the performance of 7-year olds in two rooms with
same RTs but different values on C50 and SPL is
affected or not. G is not measured since the rooms
are relatively small and since it is not possible to
measure G according to ISO 3382-1 [15] – but
SPL is measured instead.
In this study there is used sound absorption
material on one surface only, to illustrate a typical
solution in a Danish classroom or group room.
This is not known to be the ideal solution for good
room acoustics – but this was necessary to secure

3. FORUM ACUSTICUM 2014 M. Beldam: Testing, testing - can you hear me, understand me remember me?
7–12 September, Krakow
same RT in the two rooms. Furthermore, this is the
preferred solution in Danish schools to reach 0.6
sec. and many products are known to have the
possibility to reach this demand. Therefore, this
study does not discuss placement of absorption as
such.
3 Design of the study
3.1 The rooms
The rooms are situated in Hyllinge, Sweden –
Ecophon premises. The rooms are not as big as
ordinary classrooms and relate more to group
rooms suitable for fewer children. The sizes of
rooms are (length x width x height) = 3.2 m x 3.7
m x 2.06 m. The rooms have identical furniture
and surfaces except for the ceiling: Both ceilings
have suspended ceilings that lay in a grid system,
system edge A (same overall depth of system), but
the tiles have different sound absorption
performance. In Room 1 the absorber is a 20 mm.
‘Class A’ glass wool absorber and in Room 2 the
absorber is a 14 mm mineral wool ‘Class E’
absorber. The classification of the absorbers is
done according to ISO 11654 [20]. The acoustic
descriptors RT, T20, and C50 are measured
according to ISO 3382-1 [15]. The values of the
descriptors are illustrated in figure 1 and 2.
Figure 1. Measured T20 in the two rooms.
Figure 2. Measured C50 in the two rooms.
3.2 Testing material and procedure
The testing material was developed to secure no
need for student reading and was developed to
reflect some of the abilities needed in 1st
grade in a
Danish school: 1) Working memory test and 2)
Simple instructions to be followed. A working
memory test was developed by the EDUnet of
Ecophon and the task is to remember the order of
pictograms in two conditions: With and without
background noise. The reason for these different
configurations with/without background noise is
that for many years research has shown that
cognitive processing is known to be easily
disturbed by incompatible environmental
stimulation [21, 22]. At the same time it is
important to mention that recent research has
shown that exposure to background noise can
improve performance of some inattentive children
– for example children with ADHD [23, 24].
Nine pictograms are in each box (total is 12 boxes
in the test) and 4 of them are played from a
recorded speech material. An example could sound
like: House (pause 1 sec.). Moon (pause 1 sec.).
Cup (pause 1 sec.). Chair (pause 1 sec.). When the
recording is played the papers face down and are
not turned until the four words/pictograms are all
mentioned. The subjects then have to write the
order (with numbers next to the pictogram) after
the recording is played. The first 6 boxes are
played with the signal only and the last 6 boxes are
played with added background noise. The signal is
set to be 55 dB(A) and the background noise is set
to be 56 dB(A) in Room 1. The task is done in
both rooms – and the effect from the loudspeakers
is unchanged from Room 1 to Room 2. In Room 2
the room acoustical conditions affect the SPL even
though the effect stays the same. The signal in
Room 2 is measured to be 59 dB(A) and the
background noise is measured to be 61 dB(A). The
simple instruction test developed by audiologist
Anna K. Lejon and the author, seeks to reflect the
simple instructions given by the teacher to the
pupils in a normal teaching situation. The subjects
have 4 pens in red, blue, green and yellow colour
and they are given a simple instruction to perform.
The instruction contains the choice of colour,
form, placement and number of objects to drawn.
A potential instruction could be: Take the blue pen
and draw a circle near the bottom of the box. Five
instructions are given per box. There are 4 boxes
to be done in total. The simple instruction test is
also done in both rooms and the effect of the

4. FORUM ACUSTICUM 2014 M. Beldam: Testing, testing - can you hear me, understand me remember me?
7–12 September, Krakow
loudspeakers and the SPL in the rooms are the
same as with the working memory test. The
procedure is also the same: The first half of the test
(= two boxes) is to be done only listening to the
signal and the second half of the test (= two boxes)
the background noise is added. The sound is
measured with a sound level meter: Integrating
Sound Level Meter Type 2239A. The loud speaker
used for the signal is a Studio monitor Genelec
8030A. The loudspeaker used for background
noise is a Logitech Laptop Speaker Z305. Two
computers are used to play the sound files – both
Lenovo X230. The background noise is a sound
file recorded directly in a classroom during a lunch
break and it is not possible to understand words or
sentences. The signal is recorded in relatively dry
conditions and has unnoticeable RT. The testing
periods were filmed with a Go Pro Hero III Black
Edition. The subjects are divided in two teams.
Team A begins the testing in Room 2 whereas
Team B begins in Room 1. This is to see whether
the order/experience of the testing could influence
the results or not.
3.3 The subjects
The eight subjects chosen are 7-year olds attending
1st
grade in a traditional Danish public school. This
age is chosen to this study to see if so called young
sensitive listeners are affected by different values
of C50 and SPL when RT stays the same. Even
teenagers’ brains are still immature with respect to
the number and strength of neural connections
underlying the cognitive inhibitory control and
together with the sensitivity of the hearing system
of even younger children details in the room
acoustic quality can be a challenge [22]. M. Klatte
et al. [25] discus how younger children under 9
years old show disruption of speech perception in
noise and also comment on how the speech
perception decreases as RT increases which leads
to the understanding that classrooms suitable for
children under 9 years old should reach higher
standards in general in regards to RT. The 7-year
olds in the study are not in the same school but
live in different places and regions of Denmark
(and Sweden). None of them are known to have
hearing disabilities but no formal listening tests are
done prior to the study. One is bilingual with
Swedish as his mother tongue and Danish as his
second language (he lives in Sweden and has a
Danish father and Swedish mother. Both languages
are spoken in their home). The testing is done over
two weekends – four children each weekend to
secure short waiting times and to avoid fatigue
during the testing.
The subjects were placed next to each other and
were instructed not to look at their neighbor’s
paper. Unfortunately the filming showed that
subject no. 5 couldn’t help looking at his
neighbor’s papers during both tests which is the
reason why his results are not analyzed and
discussed in this paper. The subjects were placed
approx. 2.5 m. from both speakers.
The subjects received only a short instruction prior
to the test – and only to secure that they
understood the testing material. They knew that is
was a study about ‘sound’ but neither the kids nor
the parents new any details of the room conditions
and what the purpose of the study was. To secure
anonymity the subjects were given a number (1-8)
which they should use in every test.
4 Results
4.1 Working memory test
The working memory test was scored as follows: If
the picture is right but the order is wrong – it gives
1 point. If both the picture and the order are
correct – it gives 2 points. Maximum score per
configuration is 48. The results of subject no. 5 are
not analyzed as mentioned earlier. Abbreviations
used are: R1 (Room 1), R2 (Room 2), R1-S (Room
1 with signal only), R1-N (Room 1 with added
background noise), R2-S (Room 2 with signal
only), R2-N (Room 2 with added background
noise). Figure 3 shows the results divided in rooms
and different conditions.
Figure 3. Working memory results.
Statistical significance has not been investigated
but there is a trend and an indication that the

5. FORUM ACUSTICUM 2014 M. Beldam: Testing, testing - can you hear me, understand me remember me?
7–12 September, Krakow
subjects perform differently in the two rooms
despite same RT, both when they only listen to the
signal and when the background noise is added.
The trend is clearly that the subjects on average
perform better in R1. The best condition in general
is in R1-S. Only subject no. 4 and 6 have their best
performance in R2 – and they perform the best in
R2-N. The average performance shows that the
subjects in general perform better in R1 and
furthermore, subject no. 1 and 3 perform better in
R1-N compared to R2-S. The bilingual subject is
no. 3 and it shows that he is the top performer of
the total group in both conditions in R1 and in R2-
S. Only no. 6 performs better than him in R2-N in
noise. There is a general trend which shows that
Team B (subject no. 6-8) perform better in R2 than
Team A (subject no. 1-4) and especially no. 7
shows very little difference in performance from
R1 to R2 within the same condition yet still he
performs best in R1 when the conditions
noise/silence are the same. The biggest differences
in performance are between R1-S and R2-N and
subject no. 1 actually only perform less than half
as well in R2-N. Together with no. 8 subject no. 1
has the biggest gap in performance between R1-S
and R2-S. There is a significant trend in
performance from silence to noise in both rooms
(except the earlier mentioned performance of no. 4
and 6). It shows that the drop of performance is
nearly the same in the two rooms when going from
silence to noise. Only subject no. 8 shows a
‘unique’ gap from R1-S in comparison to all the
other configurations and the performance in R1-N
is also better than R2-S.
4.2 Simple instructions test
The simple instructions test is scored as follows: To
get 1 point everything in the instruction has to be
followed; colour, number, form and placement of
the figure. If anyone is wrong there will be no point
awarded. There are 5 instructions per. box, 2 boxes
per. condition. Maximum score per. configuration is
10. Figure 4 shows the results divided in rooms and
different conditions.
A similar trend is seen in this test; there is a
difference in performance in the two rooms in
general. Almost all the subjects score maximum
points in R1-S whereas the score in general drops
in all the other configurations. Two of the subjects
(1 and 3) score maximum points in both R1-S and
R2-S but their performance drops when noise is
added and especially in R2-N. All the subjects
have their best score in R1-S except no. 4 who has
the same score in R1-S and R1-N. The general
performance in R2-N is lowest for every subject
except for no. 6 and the difference from the other
configurations seems to be bigger than what is
seen in the working memory test. Subject no. 3
who was the top performer in the previous test has
together with no 2 and 4 the lowest score of the
subjects in R2-N. Again Team B performs better
than Team A in R2 in general and they all have
better scores in R2-N than the other group. Both
subject no. 7 and 8 have the same score in R1-N
and R2-S and they also have the same size of gap
down to R2-N. Only subject no. 6 performs better
in R2-N in comparison to R1-N. Like the working
memory test there is a significant trend in
performance from silence to noise in both rooms -
where the score is best in silence – except for no. 6
who has the same score in R2-S and R2-N (and
who also scored best in R2-N in the working
memory test). In average there is much a bigger
drop in performance between R2-S to R2-N than is
seen between R1-S to R1-N.
Figure 4. Simple instructions test.
5 Discussion and conclusion
So how do we secure that younger subjects hear,
understand and remember what is said in the
classroom weather the teaching method allows
background noise or not? The aim of this pilot
project and this paper was to investigate if other
acoustic descriptors besides RT are needed to
describe a sound environment suited for verbal
communication and teaching of sensitive listeners /
7-year olds. The aim was also to learn if the theory
and known research about the necessity of using
C50 and SPL would be supported by working

6. FORUM ACUSTICUM 2014 M. Beldam: Testing, testing - can you hear me, understand me remember me?
7–12 September, Krakow
memory testing and simple instructions testing of
subjects in situ.
Despite the fact that statistical significance has not
been investigated it is clear within this project
that the subjects perform differently in the two
rooms and there is a strong indication of the
importance of looking at the early and late energy
ratios and sound levels together with RT. The
subjects perform better in R1 in general where
C50 is higher and the SPL lower – and it is
remarkable how negatively the subjects’
performance is affected when noise is added in
especially R2. It seems as if background noise is
more annoying when C50 is lower and SPL is
higher. There are only two cases in total where
subjects perform better in R2-N (working memory
test, subject no. 4 and 6) which need more research
to explain. It can be ‘just’ an individual preference.
Maybe there would be no trend at all if the testing
was done twice in both rooms or maybe this is
exactly what the trend would be if the project
included 100 subjects. If both tests in all
configurations are compared it though seems as if
the last explanation is most unlikely to happen.
The testing material gives an idea of how the
subjects would perform in different tasks in school
and it seems as if they are most vulnerable to noise
in general in R2 while solving the simple
instructions test. This task reflects a normal
situation that happens all the time in a classroom
every day; the teacher gives an instruction to be
followed. In a real classroom the instructions
might not be as simple as the one in this project
and it would be interesting to see what happens
when the simple instructions are carried out in a
real situation: Would the subjects be more affected
by the acoustics or would it be the same? The
bilingual subject does not perform as ‘bad’ as
expected since he is the top performer in total of
both teams. Again there is not enough material to
be statistically significant but generally second
language learners – together with hearing impaired
learners, are more sensitive to acoustics [26, 27].
His results maybe indicate a clever kid who works
with great attention or maybe it is just a
coincidence? In general Team B has better results
in R2 than Team A in both tests. They still perform
best in R1 but it could be possible that the
knowledge and experience learned in R1 where
they started the testing help them in R2 which
could lead to a discussion whether details in
acoustics are that important or not, if the task is
well known? It would be interesting to see what
happens if the testing was done over and over
again to see if there is a maximum of ‘experience’
to affect the results positively or not. Maybe this
project already shows the maximum?
This project has not looked on different RTs and it
would be highly relevant to see if other descriptors
are necessary if RT is short ‘enough’. Would there
be any difference in performance in two rooms
with RT = 0.4 sec. if C50 and SPL are different? Or
would the relatively short RT ‘secure’ an
optimum? It would also be interesting to see if the
same trend shown in this project will be seen if the
rooms get bigger and if there are more subjects in
the rooms – would the results be even more
different/ exaggerated in a larger but more typical
classroom? Finally it would be most interesting to
explore what background noise means in a real
classroom when some of the noise is
understandable words (as in a real classroom
situation). Will it then be more or less necessary to
include other acoustic descriptors? And how is the
Lombard-Effect related to this?
The results of this pilot project give an image of
the necessity of using C50 and sound strength
together with reverberation time and it would be
relevant to work with more practical research in
this direction - and not only with 7-year olds but in
schools and learning environments in general.
References
[1] B. Rasmussen, J. Brunskog, D. Hoffmeyer:
Reverberation time in class rooms – Comparison of
regulations and classifications criteria in the Nordic
countries. Proc. 2012 BNAM, Odense, Denmark,
[2] Building Regulations 2010, Danish Enterprise and
Construction Authority, Danish Ministry of
Economic and Business Affairs.
http://erhvervsstyrelsen.dk/file/155699/BR10_englis
h.pdf
[3] The National Research Centre for the Working
Environment, Arbejdsmiljø og helbred i Danmark
2012,
http://www.arbejdsmiljoforskning.dk/~/media/Boege
r-og-rapporter/Arbejdsmiljoe-og-helbred-i-
Danmark2012-Netversion-Juni2013.pdf
[4] Danish Centre of Educational Environment, DCUM,
Indeklima og fagligt udbytte, 2013,
http://dcum.dk/indeklima-og-fagligt-udbytte-rapport

7. FORUM ACUSTICUM 2014 M. Beldam: Testing, testing - can you hear me, understand me remember me?
7–12 September, Krakow
[5] ISO 3382-3:2012 - Acoustics -- Measurement of
room acoustic parameters -- Part 3: Open plan
offices.
[6] Wilbert F. Snyder: Acoustical Investigations of
Joseph Henry as Viewed in 1940. J. Acoust. Soc.
Am. 12 (1941) 464 ff.
[7] Emily Ann Thompson: The Soundscape Of
Modernity: Architectural Acoustics And The Culture
Of Listening In America, 2002
[8] Howard et al: Listening effort at signal-to-noise
ratios that are typical of the school classroom. Int. J.
Audiol. 48(12), (2010) 928ff
[9] J. S. Bradley and H. Sato: On the importance of
early reflections for speech in rooms. J. Acoust. Soc.
Am. 113 (2003) 3233-3244
[10] J. S. Bradley et al: A just noticeable difference in
C50 for speech, Applied Acoustics 58 (1998), 99ff,
[11] W. Yang, J. S. Bradley: Effects of room acoustics
on the intelligibility of speech in classrooms for
young children. J. Acoust. Soc. Am. 125 (2) (2009)
[12] J. S. Bradley et al: On the combined effects of
signal-to-noise ratio and room acoustics on speech
intelligibility. J. Acoust. Soc. Am. 106 (4) (1999)
1820ff.
[13] E. Nilsson: Decay Processes in Rooms with Non-
Diffuse Sound Fields Part I: Ceiling Treatment with
Absorbing Material. Building Acoustics 11(2004)
39ff.
[14]E. Nilsson: Decay Processes in Rooms with Non-
Diffuse Sound Fields Part II: Effect of irregularities.
Building Acoustics 11(2004) 133ff.
[15] ISO 3382-1:2009, Acoustics – Measurements of
room acoustic parameters. Part 1, Performance
Spaces
[16] J. P. A. Lochner and J. F. Burger: The influence of
reflections on auditorium acoustics. J. Sound Vib.
(1) (1964) 426ff.
[17] E. Nilsson: Room Acoustic Measures for
Classrooms, Proc. 2010 Internoise, Lisbon, Portugal
[18] J. Bradley: Review of objective room acoustic
measures and future needs. Applied Acoustics 72
(2011) 713ff.
[19] J. Harvie-Clark and N. Dobinson: The Practical
application of G and C50 in classrooms. Proc. 2013
Internoise, Innsbruck, Austria,
[20]ISO 11654:1997, Acoustics - Sound absorbers for
use in buildings - Rating of sound absorption
[21]D. E. Broadbent: The effects of noise on behavior.
Elmsford, NY, US, Pergamon Pres, 1958
[22] H-J. Hwang et al: Strengthening of top-down frontal
cognitive control networks underlying the
development of inhibitory control: a functional
magnetic resonance imaging effective connectivity
study. The Journal of Neuroscience 30(46) (2010)
15535ff.
[23]G. B. W. Söderlund et al: Listen to the noise: Noise
is beneficial for cognitive performance in ADHD. J.
Child Psychiatry 48(2007) 840ff,
[24]G. B. W. Söderlund et al: The effects of background
white noise on memory performance in inattentive
school children, Behavioral and Braind Functions 6
(2010) 55ff
[25] M. Klatte et al: Effects of noise and reverberation
on speech perception and listening comprehension of
children and adults in a classroom-like setting. Noise
Health, 12 (49) (2010) 270 ff.
[26]Nelson, P. et al: Acoustical barriers to learning:
Understanding the need for a classroom acoustics
standard. Acoustical Society of America (2002).
[27]Su-Hyun Jin et al: Speech perception in gated noise:
The effects of temporal resolution. Journal of the
Acoustical Society of America 119 (2006) 3097 ff