<I>

  <&>Wellington Corpus of Spoken New Zealand English Version One</&>
  <&>Copyright 1998 School of Linguistics & Applied Language Studies</&>
  <&>Victoria University of Wellington</&>

  <&>side one</&>
  <&>2:32</&>
  

  <WSC#MUL013:0005:RR>
      right <.>now</.> <,> last week we finished the lecture with a
      discussion of the um lincoln centre hall philharmonic hall in
      new york built by leo <?>beranic</?> and <,> the book that
      documents that

  <WSC#MUL013:0010:RR>
      we've um had a look at some of <.>the</.> the halls there and
      the design criteria and we know there's failure and in <.>fact</.>
      but <.>in</.> about ten <&>3:00</&> years later they completely
      gutted it and um it went right down to the brick work and
      rebuilt new sides simply because they had tried time and again
      to adjust the acoustics

  <WSC#MUL013:0015:RR>
      now if everything <,> um <,> in beranic's design criteria had
      been right then um it would have been a perfect hall but there
      was one very important failing er which is the subject of the
      <.>lec</.> lecture today um <,,><&>4</&>

  <WSC#MUL013:0020:RR>
      if you recall the er the design of that lincoln centre concert
      hall <,,><&>12</&> very approximately the seats stage you can
      see pictures of the <,> special <.>ceiling</.> um <&>4:00</&>
      reflecting panels which were built into the hall so that if
      somebody was on stage here er the direct sound proceeded there
      and the first major reflection came in after an initial time
      delay which was fairly short because the path distance wasn't
      too long this was lower <.>th</.> obviously than the basic er
      structure of the hall and this meant to er produce <.>i</.>
      intimacy which could be rated <.>a</.> as forty percent of the
      total hundred percent value for a good hall

  <WSC#MUL013:0025:RR>
      now <O>clears throat</O> there are two things wrong with this um
      reflecting system that had been added to the hall

  <WSC#MUL013:0030:RR>
      one of them was that all the um reflecting bits they weren't
      exactly that shape were exactly the same size and they had gaps
      between them

  <WSC#MUL013:0035:RR>
      they represented a grid with <.>sp</.> air space in between so
      that produced the effect of um a grill which could absorb er and
      <&>5:00</&> destructively interfere with certain frequencies <.>wh</.>
      so that meant there was a <.>t</.> total distortion from this
      object um

  <WSC#MUL013:0040:RR>
      <.>a</.> as well or perhaps related to that sound got behind it
      and seemed to get trapped so that you have er certain sound
      waves able to pass specially low frequencies pass the um
      reflection but then not <.>s</.> adequately get back into the
      main space

  <WSC#MUL013:0045:RR>
      that was because of the the shape um of that constructed ceiling

  <WSC#MUL013:0050:RR>
      the other major thing that turned out to be wrong with it was
      something that was only discovered with er subsequent research
      and um howard marshall was strongly involved in that
      developmental work in the <.>s</.> <.>seven</.> sixties and into
      seventies <,,>

  <WSC#MUL013:0055:RR>
      now this <&>6:00</&> developmental work involved laboratory
      experiments and <.>i</.> <.>w</.> there are papers on this

  <WSC#MUL013:0060:RR>
      if you want more detail i can give the um specific references

  <WSC#MUL013:0065:RR>
      they're um in journals um journal of sound vibration and um um
      and i can't think of the other one at the moment um but they did
      things like this

  <WSC#MUL013:0070:RR>
      what they wanted to find out in a laboratory so that <quietly>oh
      there's a pen <unclear>word</unclear></quietly> in a fairly
      small enclosed and probably anechoic chamber without any echo um
      they've very absorbing walls they would have a subject sit er in
      a position here with their ears <,> like that and they would
      play sounds from a loud speaker in the laboratory and then they
      would play delayed sounds <,> on another loud speaker somewhere
      else in the space and <&>7:00</&> there'd be a delay time here
      which would correspond to the initial time delay um in a concert
      hall in other words it would be delayed by an amount which would
      be equivalent to er a wall thirty feet or thirty metres whatever
      away okay and this sound would come here and the first major
      reflection in this laboratory would come in a certain time later
      <,>

  <WSC#MUL013:0075:RR>
      now what they discovered from these experiments <,> er was that
      there were certain places where this loud speaker was very
      satisfactory for the listener and certain places where it was
      very unsatisfactory and one of the places where it was least
      satisfactory was when it was directly over the head of the
      person um sitting in this space <?>and</?> i can't draw it of
      that diagram but just remember that fact

  <WSC#MUL013:0080:RR>
      um the direct sound comes from there from that loud speaker and
      another loud speaker is placed up there and no matter what the
      delay times or the <&>8:00</&> loudness of the reflections er
      that is unsatisfactory for the listener

  <WSC#MUL013:0085:RR>
      what the listeners turned out to really like um are er
      reflections which come in from an angle

  <WSC#MUL013:0090:RR>
      if you sort of are sitting here with your head loud speakers
      there and we're talking about where it would be in the just by
      waving my arms rather than talking in angles it's somewhere up
      there or up there on and those out of the um direct line there
      above um and coming in certainly well let's say <O>voc</O> yeah
      certainly coming in from a different angle not with your head
      straight ahead and two ears pointed towards the direct sound er

  <WSC#MUL013:0095:RR>
      it was much better if the first major reflection came <.>s</.>
      from a different position <.>le</.> <O>voc</O> was not in the
      same plane as that direct sound and that was of course binaural
      dissimilarity which <&>9:00</&> you all know about okay so
      laboratory experiments proved that um it was much better if the
      concert <.>hall</.> er well in the laboratory here er that the
      other speaker <.>came</.> had sound which came from er a
      different angle and therefore spread the image okay but <.>it's</.>
      it's more complicated than that even <,,>

  <WSC#MUL013:0100:RR>
      they also measured how much energy you'd like to have coming
      from the side reflections so um <.>th</.> this was very
      interesting in terms of subsequent design of the christchurch
      town hall and the michael fowler centre

  <WSC#MUL013:0105:RR>
      what kind of energy would would you like to have coming in as a
      major reflection and this is where i think the experiments went
      wrong and i'll just explain briefly why i think <.>that's</.>

  <WSC#MUL013:0110:RR>
      if you can put yourself mentally in this position it's something
      like listening to a stereo isn't it er with one speaker there
      and the other one <&>10:00</&> somewhere else

  <WSC#MUL013:0115:RR>
      now everybody knows that you're not going to have <.>th</.> a
      stereo with the speakers sitting on top of each other and then
      sit in front of it and expect to hear a stereo image

  <WSC#MUL013:0120:RR>
      you want to have binaural dissimilarity all the sounds coming
      from different angles

  <WSC#MUL013:0125:RR>
      now if you take your average <O>clears throat</O> laboratory um
      guinea pig in these experiments in the sixties

  <WSC#MUL013:0130:RR>
      they weren't musicians necessarily

  <WSC#MUL013:0135:RR>
      they were likely to have been colleagues um working in sound and
      scientists <O>coughs</O> put them in a situation like that and
      feed them two loud speakers and delay one and move it round what
      they're likely to enjoy most of all is that something that
      spreads the sound and sounds a bit like a stereo that <.>gives</.>
      that means almost completely equal energy each side of the
      reflector <,,> okay

  <WSC#MUL013:0140:RR>
      the direct sound and the <unclear>word</unclear> from these two
      loud speakers they're pin points of sound and what would
      probably be ideal in the laboratory situation is almost as much
      energy going to the second loud speaker

  <WSC#MUL013:0145:RR>
      remember it's delayed in time by some er factor twenty thirty
      forty fifty <&>11:00</&> sixty milliseconds <,> okay so it could
      it could start to sound like an echo if it comes too um loud too
      late but basically it's going to be er an image spreading um
      device <,,><&>3</&>

  <WSC#MUL013:0150:RR>
      now if we have a look at the situation here on the transparency
      just a minute <,,><&>11</&> <?>here we go sorry oh here it is
      it's on the piano</?> <O>clears throat</O> <,,><&>5</&>

  <WSC#MUL013:0155:RR>
      this lower graph here at the moment is the one which you should
      concentrate on <,,><&>4</&> <&>12:00</&>

  <WSC#MUL013:0160:RR>
      <.>it's</.> it's to do with the reflection level <,> that is the
      delayed speaker relative to the direct sound in decibels
      <,,><&>5</&> and this is time okay so um decibel level relative
      to direct sound um and the time after which it can come from the
      second speaker for a major reflection in a concert hall <,,>

  <WSC#MUL013:0165:RR>
      okay so you got um a situation here where they what they
      described as a preference range for the reflected loud speaker
      <,> and you can see that that's corresponds to or is even
      slightly above the level of the original sound that's that's
      where the sort of unity value goes across to zero um the
      reflected sound is the same loudness as the direct sound in that
      case <,,>

  <WSC#MUL013:0170:RR>
      if it's less <&>13:00</&> down ten decibels um <,,> it gets <,,>
      well it's not graded in the preference range and in fact there
      are two concert hall types mentioned down here small fan shape
      and a large fan shape hall

  <WSC#MUL013:0175:RR>
      those are the halls where reflections can't come back off the
      side walls

  <WSC#MUL013:0180:RR>
      do you remember um the models that we looked at last week

  <WSC#MUL013:0185:RR>
      they were usually very wide flat halls um <,> with er side
      reflections <,> um <,> were going to simply be er not <.>a</.> a
      factor

  <WSC#MUL013:0190:RR>
      they would come in too late or too weak in relationship to the
      direct sound <,>

  <WSC#MUL013:0195:RR>
      so what beranic had originally calculated <,> as to be very
      important the initial time delay between direct sound and the
      first major reflection er <&>14:00</&> was also dependent on
      where it came from in terms of the um the er energy but in this
      case here what we are talking about is the intensity and time
      delay

  <WSC#MUL013:0200:RR>
      if it's too quiet if the major reflection comes in too late then
      it's simply masked by the direct sound it has no influence on
      the hall um and if it comes in er latish and too loud then
      there's an image shift which you couldn't get in a natural pen a
      natural acoustic and that is that you actually start to think
      the sound's coming from the other speaker the one with the
      reflection

  <WSC#MUL013:0205:RR>
      if it gets too loud then the direct sound's no longer the source
      point er <,>

  <WSC#MUL013:0210:RR>
      there's an effect there called the haas effect <,,> um which
      tells us <.>that</.> <,,> at least <.>to</.> to some extent we
      will always isolate as the <.>ori</.> <&>15:00</&> original
      source er the sound that comes first but that's only up to a
      certain point and obviously when you get to plus ten decibels um
      direct sound and then ten decibels louder your reflections then
      your attention shifts and you start to take the delayed sound as
      par that's where the orchestra is

  <WSC#MUL013:0215:RR>
      it's no longer there on the stage it's actually coming from
      something near to you

  <WSC#MUL013:0220:RR>
      if you've ever been in um <,,> <.>a</.> a space where there is
      artificial assistance for a speaker <.>or</.> or something and
      it's a crude system where they have a speaker with a microphone
      on a lectern or something and then you're sitting at the back of
      the hall and there's a speaker behind you you see the person
      talking um but you hear the sound coming from behind you

  <WSC#MUL013:0225:RR>
      do you know the sort of thing i mean

  <WSC#MUL013:0230:RR>
      i mean it's <.>s</.> simply that the source is not that person
      any more

  <WSC#MUL013:0235:RR>
      you might <&>16:00</&> be able to hear that person er but it's
      delayed by the distance whereas the the air sound travelling in
      air whereas the er er electrical circuits are travelling at the
      speed of light is coming through much faster to that speaker and
      then coming to your ears therefore you take that um as the sound
      the main source of the sound and what should happen in those
      situations in a church where there's very echoey um kind of long
      <.>rev</.> reverberation and they're trying to have the speech
      heard is that they will have a system like this but they put in
      a delay so that the first thing you hear is the preacher and
      then slightly later comes the sound from the speaker behind you
      and so long as that speaker isn't too loud er you'll actually
      take the first thing that you hear as the source so this is um
      related to that and in fact what it's saying is that if it gets
      too loud um then the image will shift and you'll start to think
      that the person talking to you <&>17:00</&> is in fact the voice
      of god from above and behind that kind of thing <,,><&>4</&>

  <WSC#MUL013:0240:RR>
      okay now we'll just go back and show you why <,,><&>6</&> the
      energy for a fan shaped hall cannot conform to this <,>

  <WSC#MUL013:0245:RR>
      it cannot give you a lot of energy from the side or above <,,>
      and in fact the lincoln <.>ph</.> the philharmonic hall er had
      this fan shaped component as well it wasn't very strongly fan
      shaped slightly fan shaped um but on the stage the energy was
      direct and then it comes off and hits the sides of the fan and
      it then continues as what they describe in the field as near
      frontal <,> lateral energy so <&>18:00</&> it's slightly off
      centre but it hits the side wall and then <.>go</.> proceeds to
      go more or less parallel to the direct sound waves because of
      the fan by hitting the fan and then going straight ahead okay so
      you don't get lateral energy from a fan shaped hall and it was
      this lateral energy that they discovered was really interesting
      for a sense of <?>envelopment</?> in a concert hall and if you
      take your music <?>beranic style</?> um <?>coffin</?> which is
      actually longer and and <.>th</.> thinner than that isn't it
      it's an extraordinary long thin building there's the stage and
      here's the audience out here direct sound and lots of
      reflections off the side walls cause the side walls are really
      close in those <?>coffin shaped</?> halls aren't they the shoe
      box effect so they have good lateral energy in fact quite
      accidental because they were just the design that was popular at
      the turn of the century or in the <&>19:00</&> nineteenth
      century <&>19:03</&>
</I>
