ICSI_original_transcripts/transcripts/Bmr028.mrt

<?xml version="1.0" encoding="iso-8859-1"?>
<!DOCTYPE Meeting SYSTEM "http://www.icsi.berkeley.edu/Speech/mr/Meeting1.dtd">

<Meeting Session="Bmr028" DateTimeStamp="2001-06-21-1445" Version="2" VersionDate="Jan 7 2004">

  <Preamble>
    <Notes>
      Digits were read simultaneously.
   
   fe016 wrote "I didn't have mic on right initially.
   Doesn't fit very comfortably on my ear".
   This situation allowed an exceptional amount of ambient
   noise.  At times, therefore, her speech was nearly or
   completely indecipherable.  Sometimes the noise made
   it difficult to determine whether she was speaking at
   all.
   There is a lot of mike noise on this channel (A) as she
   works to get the mike reasonably positioned, but this
   settles down a lot at 5:33.  Thereafter, there are only
   occasional stretches of bad mike noise (e.g. around 22:20-40).
   
   me001 (channel 4) fiddles with his mike frequently, so the
   gain varies considerably, resulting in some stretches of
   very pronounced breathing and laughs saturating the line.
    </Notes>
    <Channels>
      <Channel Name="chan0" Mic="c1" Transmission="w1" Gain="11" AudioFile="chan0.sph"/>
      <Channel Name="chan1" Mic="s1" Transmission="w1" Gain="9" AudioFile="chan1.sph"/>
      <Channel Name="chan2" Mic="s1" Transmission="w1" Gain="9" AudioFile="chan2.sph"/>
      <Channel Name="chan3" Mic="s1" Transmission="w1" Gain="9" AudioFile="chan3.sph"/>
      <Channel Name="chan4" Mic="s1" Transmission="w1" Gain="9" AudioFile="chan4.sph"/>
      <Channel Name="chan5" Mic="c1" Transmission="w1" Gain="11" AudioFile="chan5.sph"/>
      <Channel Name="chan6" Mic="c2" Transmission="u1" Gain="9" AudioFile="chan6.sph"/>
      <Channel Name="chan7" Mic="c2" Transmission="u1" Gain="9" AudioFile="chan7.sph"/>
      <Channel Name="chan9" Mic="p1" Transmission="j1" Gain="7" AudioFile="chan9.sph"/>
      <Channel Name="chanA" Mic="u2" Transmission="j1" Gain="9" AudioFile="chanA.sph"/>
      <Channel Name="chanB" Mic="u1" Transmission="j1" Gain="12" AudioFile="chanB.sph"/>
      <Channel Name="chanC" Mic="u3" Transmission="j1" Gain="12" AudioFile="chanC.sph"/>
      <Channel Name="chanD" Mic="u3" Transmission="j1" Gain="12" AudioFile="chanD.sph"/>
      <Channel Name="chanE" Mic="c2" Transmission="u1" Gain="9" AudioFile="chanE.sph"/>
      <Channel Name="chanF" Mic="c2" Transmission="u1" Gain="9" AudioFile="chanF.sph"/>
    </Channels>
    <Participants DB="icsi1.spk">
      <Participant Name="me011" Channel="chan0" Seat="9"/>
      <Participant Name="me018" Channel="chan1" Seat="5"/>
      <Participant Name="me013" Channel="chan2" Seat="7"/>
      <Participant Name="mn014" Channel="chan3" Seat="6"/>
      <Participant Name="me001" Channel="chan4" Seat="2"/>
      <Participant Name="fe008" Channel="chan5" Seat="8"/>
      <Participant Name="me051" Channel="chan9" Seat="4"/>
      <Participant Name="fe016" Channel="chanA" Seat="3"/>
      <Participant Name="mn017" Channel="chanB" Seat="1"/>
    </Participants>
  </Preamble>


  <Transcript StartTime="0.0" EndTime="3312.331">

    <Segment StartTime="0.319" EndTime="1.395" Participant="me011">
      And we're going.
    </Segment>
    <Segment StartTime="5.340" EndTime="5.619" Participant="me011">
      <VocalSound Description="mouth"/> <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="5.619" EndTime="8.866" Participant="me011">
      So the only status ite- well, <Emphasis> first </Emphasis> of all we h- haven't decided whether we're
    </Segment>
    <Segment StartTime="8.866" EndTime="9.723" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="9.723" EndTime="11.773" Participant="me011">
      Meeting Recorder data issues, or
    </Segment>
    <Segment StartTime="12.572" EndTime="13.777" Participant="me011">
      recognition this week.
    </Segment>
    <Segment StartTime="14.316" EndTime="15.623" Participant="me011">
      I <Emphasis> think </Emphasis> we were recognition.
    </Segment>
    <Segment StartTime="15.996" EndTime="16.649" Participant="me013">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="16.878" EndTime="21.279" Participant="me013">
      Wha- what was on the list? Th- the - I mean, I sent you a couple things, although I don't remember them.
    </Segment>
    <Segment StartTime="20.524" EndTime="22.561" Participant="me011">
      You only sent me one thing, which was demo status.
    </Segment>
    <Segment StartTime="24.711" EndTime="25.424" Participant="me013">
      Right.
    </Segment>
    <Segment StartTime="24.843" EndTime="31.054" Participant="me011">
      And asking which one we were on this week. <Comment Description="last two words while laughing"/> <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="26.730" EndTime="30.180" Participant="me013">
      Ah. That was the second thing. Right. <VocalSound Description="laugh"/> Right.
    </Segment>
    <Segment StartTime="28.028" EndTime="28.932" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="31.054" EndTime="33.405" Participant="me011">
      So should we simply assert that this week we are
    </Segment>
    <Segment StartTime="33.732" EndTime="36.290" Participant="me011">
      recognition, and next week data issues, and - ?
    </Segment>
    <Segment StartTime="34.835" EndTime="35.177" Participant="fe008">
      Hmm.
    </Segment>
    <Segment StartTime="35.780" EndTime="36.701" Participant="me018">
      I think that's correct.
    </Segment>
    <Segment StartTime="36.613" EndTime="38.699" Participant="fe008">
      Yeah. I think so, too. Yeah.
    </Segment>
    <Segment StartTime="37.505" EndTime="37.928" Participant="me018">
      Yeah.
    </Segment>
    <Segment StartTime="37.617" EndTime="38.362" Participant="me011">
      And, uh -
    </Segment>
    <Segment StartTime="38.362" EndTime="38.842" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="38.842" EndTime="40.919" Participant="me011">
      So, I think what we should <Emphasis> probably </Emphasis> do is any
    </Segment>
    <Segment StartTime="40.982" EndTime="43.209" Participant="me011">
      quick, small stuff we can do every week.
    </Segment>
    <Segment StartTime="43.810" EndTime="47.633" Participant="me011">
      So - like Morgan asked about the demo status. We can go ahead and talk about that a little bit.
    </Segment>
    <Segment StartTime="47.633" EndTime="48.178" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="48.178" EndTime="48.942" Participant="me011">
      And then do -
    </Segment>
    <Segment StartTime="49.175" EndTime="51.323" Participant="me011">
      then alternate in more depth. <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="52.002" EndTime="53.083" Participant="me013">
      <VocalSound Description="inbreath"/> <VocalSound Description="clears throat"/>
    </Segment>
    <Segment StartTime="53.083" EndTime="58.310" Participant="me013">
      By the way, I'll - I - I won't be here <Pause/> next Thursday. I'll be out of town. <VocalSound Description="mouth"/> But -
    </Segment>
    <Segment StartTime="56.889" EndTime="57.309" Participant="me011">
      <VocalSound Description="mouth"/> O_K.
    </Segment>
    <Segment StartTime="59.615" EndTime="61.636" Participant="me011">
      <Pronounce> Actually,  </Pronounce> I <Pause/> may not be here either. So <Pause/>
    </Segment>
    <Segment StartTime="62.228" EndTime="65.150" Participant="me011">
      I gotta double check the dates. But, <VocalSound Description="inbreath"/> anyway.
    </Segment>
    <Segment StartTime="65.694" EndTime="66.005" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="66.005" EndTime="71.437" Participant="me011">
      So, uh, demo status. First of all, I did a little thing for Liz with the Transcriber tool,
    </Segment>
    <Segment StartTime="71.623" EndTime="74.630" Participant="me011">
      that, um - <Emphasis> first </Emphasis> of all, it uses the forced alignments,
    </Segment>
    <Segment StartTime="74.630" EndTime="75.097" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="75.097" EndTime="76.546" Participant="me011">
      so that the words appear,
    </Segment>
    <Segment StartTime="77.744" EndTime="79.224" Participant="me011">
      uh, in their own segments,
    </Segment>
    <Segment StartTime="79.333" EndTime="80.470" Participant="me011">
      rather than in long -
    </Segment>
    <Segment StartTime="81.030" EndTime="81.594" Participant="mn017">
      Oh.
    </Segment>
    <Segment StartTime="81.529" EndTime="83.179" Participant="me011">
      in long <Pronounce> chunks.  </Pronounce> She said that that -
    </Segment>
    <Segment StartTime="83.319" EndTime="85.626" Participant="me011">
      she thought that was a much better idea for
    </Segment>
    <Segment StartTime="85.844" EndTime="87.121" Participant="me011">
      the other stuff she's working on.
    </Segment>
    <Segment StartTime="87.309" EndTime="88.892" Participant="mn017">
      Yeah. That's great.
    </Segment>
    <Segment StartTime="87.947" EndTime="88.584" Participant="me011">
      Um -
    </Segment>
    <Segment StartTime="88.974" EndTime="92.975" Participant="me011">
      And that works fine except it's even <Emphasis> slower </Emphasis> to load. It's already pretty slow to load.
    </Segment>
    <Segment StartTime="90.968" EndTime="93.680" Participant="me001">
      <VocalSound Description="inbreath"/> Yeah. <VocalSound Description="outbreath"/> It's slow.
    </Segment>
    <Segment StartTime="91.315" EndTime="93.350" Participant="mn014">
      Yeah. It's more segments? Or - ? <Comment Description="leading question"/> Yeah.
    </Segment>
    <Segment StartTime="93.350" EndTime="97.860" Participant="mn017">
      Is that because the transcripts get longer? The f- transcript file gets longer?
    </Segment>
    <Segment StartTime="93.770" EndTime="93.970" Participant="me013">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="93.862" EndTime="94.317" Participant="mn014">
      <Pronounce Pronunciation="&quot;nn-yuh&quot;"> Yeah.  </Pronounce>
    </Segment>
    <Segment StartTime="95.641" EndTime="95.953" Participant="me011">
      Yeah.
    </Segment>
    <Segment StartTime="95.811" EndTime="96.202" Participant="mn014">
      Yep.
    </Segment>
    <Segment StartTime="96.498" EndTime="96.716" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="97.528" EndTime="99.117" Participant="me011">
      And the Transcriber tool is just
    </Segment>
    <Segment StartTime="98.489" EndTime="99.445" Participant="me013">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="99.475" EndTime="101.120" Participant="me011">
      not very good n- at - <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="101.696" EndTime="105.027" Participant="mn017">
      But th- but that's - that's n- You didn't have to change
    </Segment>
    <Segment StartTime="102.030" EndTime="102.715" Participant="me011">
      at that. <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="105.027" EndTime="109.620" Participant="mn017">
      the software for that yet. Right? It's just formatting the right kind of, uh, X_M_L - ?
    </Segment>
    <Segment StartTime="106.345" EndTime="106.826" Participant="me011">
      Correct.
    </Segment>
    <Segment StartTime="108.446" EndTime="110.663" Participant="me011">
      Yeah, it's just writing conversion tools from the
    </Segment>
    <Segment StartTime="111.052" EndTime="111.535" Participant="me011">
      format
    </Segment>
    <Segment StartTime="111.784" EndTime="112.423" Participant="me011">
      that the
    </Segment>
    <Segment StartTime="112.750" EndTime="113.327" Participant="me011">
      aligner -
    </Segment>
    <Segment StartTime="113.950" EndTime="114.941" Participant="me011">
      Actually th- he did a
    </Segment>
    <Segment StartTime="115.025" EndTime="116.265" Participant="me011">
      S_R_T file for it.
    </Segment>
    <Segment StartTime="116.474" EndTime="118.000" Participant="mn017">
      Mm-hmm. O_K.
    </Segment>
    <Segment StartTime="117.233" EndTime="119.795" Participant="me011">
      And then just back into Transcriber - <Pronounce Pronunciation="&quot;trantriber&quot;"> Transcriber  </Pronounce> format.
    </Segment>
    <Segment StartTime="118.740" EndTime="120.732" Participant="mn017">
      Oh, good. That's very good. O_K.
    </Segment>
    <Segment StartTime="120.230" EndTime="123.880" Participant="me011">
      <VocalSound Description="inbreath"/> Yeah. So my - my decision was, for the <Emphasis> first </Emphasis> pass for this demo that
    </Segment>
    <Segment StartTime="123.994" EndTime="126.972" Participant="me011">
      Liz was talking about <Pause/> I decided that I would do,
    </Segment>
    <Segment StartTime="124.744" EndTime="125.211" Participant="mn017">
      Mm-hmm.
    </Segment>
    <Segment StartTime="128.320" EndTime="131.696" Participant="me011">
      um, only enough to get it <Emphasis> working, </Emphasis> as opposed to any <Emphasis> coding. </Emphasis>
    </Segment>
    <Segment StartTime="131.648" EndTime="132.104" Participant="mn017">
      Right.
    </Segment>
    <Segment StartTime="132.251" EndTime="136.443" Participant="me011">
      And so the other thing I- sh- she wanted to display the stylized F_ zeroes, I think they're called?
    </Segment>
    <Segment StartTime="136.990" EndTime="137.388" Participant="me011">
      Is that right?
    </Segment>
    <Segment StartTime="137.083" EndTime="139.780" Participant="me001">
      Yeah. The <Pause/> linear fit.
    </Segment>
    <Segment StartTime="137.301" EndTime="137.968" Participant="mn017">
      Mm-hmm.
    </Segment>
    <Segment StartTime="138.524" EndTime="139.267" Participant="me011">
      And, uh -
    </Segment>
    <Segment StartTime="139.696" EndTime="144.162" Participant="me011">
      So what I did is I just took the file with those in it, converted it so that it looks like an <Emphasis> audio </Emphasis> file.
    </Segment>
    <Segment StartTime="144.030" EndTime="144.780" Participant="mn017">
      Right.
    </Segment>
    <Segment StartTime="144.162" EndTime="144.743" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="144.743" EndTime="147.191" Participant="me011">
      And, so you s- it shows <Emphasis> that </Emphasis> instead of the wavefile.
    </Segment>
    <Segment StartTime="147.404" EndTime="149.827" Participant="me011">
      And so that - that's working, and I think it actually looks pretty good.
    </Segment>
    <Segment StartTime="147.889" EndTime="148.330" Participant="mn017">
      Cool.
    </Segment>
    <Segment StartTime="149.787" EndTime="150.407" Participant="mn017">
      Mm-hmm.
    </Segment>
    <Segment StartTime="150.000" EndTime="154.189" Participant="me011">
      <VocalSound Description="inbreath"/> Um, I'd like someone who's more familiar with it to <Emphasis> look </Emphasis> at it because when <Emphasis> I </Emphasis> was looking at it,
    </Segment>
    <Segment StartTime="154.189" EndTime="154.645" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="154.645" EndTime="155.616" Participant="me011">
      we seemed to have
    </Segment>
    <Segment StartTime="154.921" EndTime="155.435" Participant="me001">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="155.805" EndTime="157.193" Participant="me011">
      lots of stuff going on when
    </Segment>
    <Segment StartTime="157.332" EndTime="158.367" Participant="me011">
      no one's saying anything.
    </Segment>
    <Segment StartTime="158.441" EndTime="160.140" Participant="me001">
      That's just background speech.
    </Segment>
    <Segment StartTime="160.419" EndTime="160.772" Participant="me011">
      Ah.
    </Segment>
    <Segment StartTime="160.815" EndTime="161.538" Participant="me001">
      Yeah.
    </Segment>
    <Segment StartTime="160.898" EndTime="163.014" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="162.542" EndTime="164.683" Participant="me018">
      <VocalSound Description="inbreath"/> So do you have to pad that out?
    </Segment>
    <Segment StartTime="162.555" EndTime="163.821" Participant="me001">
      So.
    </Segment>
    <Segment StartTime="165.289" EndTime="168.790" Participant="me018">
      Uh, so that it looks like it's an eight kilohertz sampled thing? Or - ?
    </Segment>
    <Segment StartTime="168.063" EndTime="169.587" Participant="me011">
      No. I <Comment Description="clipping or line noise"/> - the
    </Segment>
    <Segment StartTime="170.182" EndTime="172.446" Participant="me011">
      audio file you can specify any sampling rate.
    </Segment>
    <Segment StartTime="172.534" EndTime="174.355" Participant="me011">
      And so I s- I specified - instead of,
    </Segment>
    <Segment StartTime="174.355" EndTime="177.430" Participant="me011">
      you know, sixteen thousand or eight thousand, I specified a hundred.
    </Segment>
    <Segment StartTime="177.295" EndTime="177.993" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="177.442" EndTime="178.460" Participant="me018">
      Mmm.
    </Segment>
    <Segment StartTime="178.210" EndTime="178.589" Participant="me011">
      Um,
    </Segment>
    <Segment StartTime="178.639" EndTime="181.478" Participant="me011">
      and, the only <Emphasis> problem </Emphasis> with that is that there's a bug in
    </Segment>
    <Segment StartTime="181.604" EndTime="182.424" Participant="me011">
      Transcriber,
    </Segment>
    <Segment StartTime="183.005" EndTime="186.535" Participant="me011">
      that if the sample rate is too low, when it tries to compute the shape file,
    </Segment>
    <Segment StartTime="186.749" EndTime="187.532" Participant="me011">
      it fails.
    </Segment>
    <Segment StartTime="188.250" EndTime="188.822" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="188.667" EndTime="189.730" Participant="me011">
      Um, and crashes.
    </Segment>
    <Segment StartTime="189.777" EndTime="190.134" Participant="me011">
      Um -
    </Segment>
    <Segment StartTime="190.172" EndTime="191.661" Participant="me011">
      But the solution to that is just,
    </Segment>
    <Segment StartTime="191.774" EndTime="194.121" Participant="me011">
      set the option so it doesn't compute the shape file,
    </Segment>
    <Segment StartTime="194.082" EndTime="212.225" Participant="fe016">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="194.121" EndTime="194.576" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="194.576" EndTime="197.004" Participant="me011">
      and it will <Emphasis> work </Emphasis> and the only problem with that is you can't,
    </Segment>
    <Segment StartTime="197.597" EndTime="200.272" Participant="me011">
      uh, zoom <Emphasis> out </Emphasis> on it. You can zoom in, but not out.
    </Segment>
    <Segment StartTime="200.752" EndTime="201.048" Participant="me001">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="200.840" EndTime="202.220" Participant="me013">
      What's a shape file?
    </Segment>
    <Segment StartTime="201.936" EndTime="202.365" Participant="me011">
      <VocalSound Description="mouth"/> <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="202.365" EndTime="208.359" Participant="me011">
      The shape file is - If you think about a wavefile, sixteen thousand samples per second is way too many to display on the screen.
    </Segment>
    <Segment StartTime="205.901" EndTime="206.840" Participant="me013">
      <VocalSound Description="clears throat"/>
    </Segment>
    <Segment StartTime="208.655" EndTime="209.233" Participant="me013">
      Mm-hmm.
    </Segment>
    <Segment StartTime="209.064" EndTime="213.642" Participant="me011">
      So what Transcriber does, is it computes a - <Emphasis> another </Emphasis> thing to display based on the
    </Segment>
    <Segment StartTime="211.213" EndTime="213.656" Participant="me001">
      We're talking about - We're talking about that demo. <Comment Description="whispered, to [fe016]"/>
    </Segment>
    <Segment StartTime="212.225" EndTime="213.790" Participant="fe016">
      Yeah. <Comment Description="whispered"/> <NonVocalSound Description="mike noise, adjustment in progress"/>
    </Segment>
    <Segment StartTime="213.790" EndTime="215.627" Participant="fe016">
      Tried that, and it died. <Comment Description="whispered"/> <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="214.159" EndTime="215.622" Participant="me011">
      waveform. And it displays it at -
    </Segment>
    <Segment StartTime="214.706" EndTime="216.418" Participant="me001">
      Yeah. Did it? <Comment Description="whispered to [fe016]"/>
    </Segment>
    <Segment StartTime="215.627" EndTime="221.240" Participant="fe016">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="215.866" EndTime="219.487" Participant="me011">
      And it allows you to show m- many different resolutions. So there's a little
    </Segment>
    <Segment StartTime="219.866" EndTime="222.414" Participant="me011">
      user interface component that lets you sh- <Pronounce> select  </Pronounce> the resolution.
    </Segment>
    <Segment StartTime="221.240" EndTime="222.550" Participant="fe016">
      Great! <Comment Description="whispered"/> <NonVocalSound Description="mike noise, adjustment in progress"/>
    </Segment>
    <Segment StartTime="222.550" EndTime="241.290" Participant="fe016">
      <NonVocalSound Description="mike noise, adjustment in progress"/>
    </Segment>
    <Segment StartTime="222.627" EndTime="223.291" Participant="me013">
      I see.
    </Segment>
    <Segment StartTime="223.373" EndTime="225.617" Participant="me011">
      And if you don't compute the wavefile, you can't
    </Segment>
    <Segment StartTime="226.109" EndTime="226.948" Participant="me011">
      zoom
    </Segment>
    <Segment StartTime="227.734" EndTime="230.702" Participant="me011">
      out. You can't get a larger view of it. But you can zoom in.
    </Segment>
    <Segment StartTime="231.437" EndTime="231.989" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="232.595" EndTime="233.033" Participant="me011">
      Um -
    </Segment>
    <Segment StartTime="233.352" EndTime="237.494" Participant="me011">
      And that's alright, because at - at a hundred samples that's already pretty far <Pause/> out.
    </Segment>
    <Segment StartTime="238.776" EndTime="242.144" Participant="me011">
      And, uh - so I think it looks pretty good, but I'll let Liz look at it and see what she thinks.
    </Segment>
    <Segment StartTime="240.832" EndTime="241.435" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="241.290" EndTime="243.116" Participant="fe016">
      I - I <Emphasis> got </Emphasis> the wavefile but - <Comment Description="quietly, mike adjustment in progress"/>
    </Segment>
    <Segment StartTime="243.116" EndTime="249.230" Participant="fe016">
      Sorry. I got the wavefile, but I couldn't get the words yet. But the - the wavefile part looks - looks good.
    </Segment>
    <Segment StartTime="244.331" EndTime="244.907" Participant="me013">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="248.262" EndTime="248.641" Participant="me011">
      O_K.
    </Segment>
    <Segment StartTime="249.230" EndTime="255.205" Participant="fe016">
      <NonVocalSound Description="mike noise, adjustment in progress"/>
    </Segment>
    <Segment StartTime="249.573" EndTime="251.144" Participant="mn017">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="250.792" EndTime="251.141" Participant="me011">
      We should -
    </Segment>
    <Segment StartTime="251.570" EndTime="253.174" Participant="me011">
      If you were having problems with the words, we should
    </Segment>
    <Segment StartTime="253.891" EndTime="254.724" Participant="me011">
      figure out why.
    </Segment>
    <Segment StartTime="255.205" EndTime="258.568" Participant="fe016">
      I - I'll have done - I'm probably doing something wrong.
    </Segment>
    <Segment StartTime="257.879" EndTime="258.237" Participant="me011">
      O_K. <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="258.568" EndTime="260.408" Participant="fe016">
      Sorry this microphone's moving around.
    </Segment>
    <Segment StartTime="260.408" EndTime="263.320" Participant="fe016">
      <NonVocalSound Description="mike noise"/> I can't put this over my ear. <Comment Description="whispered, away from mike"/>
    </Segment>
    <Segment StartTime="263.266" EndTime="265.120" Participant="me001">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="263.320" EndTime="266.725" Participant="fe016">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="263.398" EndTime="264.428" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="263.457" EndTime="264.308" Participant="me018">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="263.505" EndTime="264.673" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="263.613" EndTime="265.876" Participant="me013">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="263.902" EndTime="265.676" Participant="fe008">
      It could - Do you have to put it <Emphasis> in </Emphasis> your ear?
    </Segment>
    <Segment StartTime="264.843" EndTime="265.377" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="265.377" EndTime="267.703" Participant="me011">
      You c- you <Emphasis> clip </Emphasis> that part over your ear.
    </Segment>
    <Segment StartTime="266.725" EndTime="272.310" Participant="fe016">
      <NonVocalSound Description="mike noise"/> I can - I can do that but there's no orientation where the - <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="272.158" EndTime="274.877" Participant="me001">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="272.310" EndTime="273.939" Participant="fe016">
      <NonVocalSound Description="mike noise"/> <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="272.529" EndTime="274.716" Participant="me013">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="272.712" EndTime="274.542" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="273.065" EndTime="273.665" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="273.939" EndTime="274.260" Participant="fe016">
      <Uncertain> Darn! </Uncertain>
    </Segment>
    <Segment StartTime="274.260" EndTime="294.075" Participant="fe016">
      <VocalSound Description="laugh"/> <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="274.542" EndTime="278.488" Participant="me011">
      <VocalSound Description="inbreath"/> Anyway. <Comment Description="while laughing"/> We'll all watch Liz play with the mike.
    </Segment>
    <Segment StartTime="274.716" EndTime="280.089" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="275.210" EndTime="276.674" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="275.322" EndTime="276.846" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="275.631" EndTime="278.146" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="277.398" EndTime="278.440" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="278.353" EndTime="282.720" Participant="fe008">
      I- Does it really need to go <Emphasis> in </Emphasis> her ear? That - that bud? It doesn't have to go in her ear. Right?
    </Segment>
    <Segment StartTime="278.488" EndTime="279.383" Participant="me011">
      <VocalSound Description="inbreath"/> Um -
    </Segment>
    <Segment StartTime="282.462" EndTime="285.754" Participant="me011">
      Uh, no. It doesn't <Emphasis> have </Emphasis> to, but that - that's - <Emphasis> I </Emphasis> find that's the only way to wear it.
    </Segment>
    <Segment StartTime="285.821" EndTime="286.760" Participant="fe008">
      Oh, wow.
    </Segment>
    <Segment StartTime="286.371" EndTime="289.794" Participant="me011">
      Is that the bud's in the ear and that the link is over it.
    </Segment>
    <Segment StartTime="286.390" EndTime="287.200" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="290.214" EndTime="292.208" Participant="me013">
      This bud's for - No.
    </Segment>
    <Segment StartTime="291.127" EndTime="297.463" Participant="me011">
      <VocalSound Description="inbreath"/> But, so, anyway, I think <Emphasis> that </Emphasis> looks pretty good. The only - the only other thing we might wanna do with that is be able to display more than one waveform.
    </Segment>
    <Segment StartTime="294.075" EndTime="295.004" Participant="fe016">
      <NonVocalSound Description="mike noise"/> <Uncertain> Hmm. </Uncertain>
    </Segment>
    <Segment StartTime="295.004" EndTime="317.125" Participant="fe016">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="297.981" EndTime="299.933" Participant="me011">
      And that actually shouldn't be too slow,
    </Segment>
    <Segment StartTime="300.531" EndTime="303.293" Participant="me011">
      uh, because it's much lower resolution than a full waveform.
    </Segment>
    <Segment StartTime="302.180" EndTime="302.840" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="302.840" EndTime="304.488" Participant="mn014">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="303.293" EndTime="303.632" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="303.632" EndTime="305.792" Participant="me011">
      The problem with it is just it does require coding.
    </Segment>
    <Segment StartTime="305.998" EndTime="306.840" Participant="me001">
      Yeah. It migh-
    </Segment>
    <Segment StartTime="306.222" EndTime="307.644" Participant="me011">
      And so it would be much better to get,
    </Segment>
    <Segment StartTime="307.844" EndTime="311.051" Participant="me011">
      uh, Dave Gelbart to do that than me, because he's familiar with the code,
    </Segment>
    <Segment StartTime="311.131" EndTime="313.084" Participant="me011">
      and is more likely to be able to get it to work quickly.
    </Segment>
    <Segment StartTime="313.185" EndTime="317.578" Participant="me001">
      It'd be nice if we can do, like, a quick hack, just so we can play <Pause/> the audio file, too.
    </Segment>
    <Segment StartTime="317.125" EndTime="317.639" Participant="fe016">
      Right.
    </Segment>
    <Segment StartTime="317.639" EndTime="329.173" Participant="fe016">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="317.895" EndTime="323.683" Participant="me001">
      Um, with th- with the display. Like, even if we - I think that even if we didn't display the waveform, <Comment Description="rising"/>
    </Segment>
    <Segment StartTime="318.612" EndTime="319.309" Participant="me011">
      Oh, O_K.
    </Segment>
    <Segment StartTime="324.210" EndTime="327.549" Participant="me001">
      it might be better to, rather, <Emphasis> play </Emphasis> the waveform than display it.
    </Segment>
    <Segment StartTime="327.549" EndTime="331.638" Participant="me001">
      I mean, like, if we were to choose - I don't know, if I were to choose between one or the other, I'd rather have it played.
    </Segment>
    <Segment StartTime="327.659" EndTime="328.077" Participant="me011">
      Mm-hmm.
    </Segment>
    <Segment StartTime="329.173" EndTime="330.290" Participant="fe016">
      Yeah. <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="329.332" EndTime="330.247" Participant="me011">
      I understand what you mean.
    </Segment>
    <Segment StartTime="330.290" EndTime="333.060" Participant="fe016">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="332.391" EndTime="333.800" Participant="me001">
      And then displayed.
    </Segment>
    <Segment StartTime="333.557" EndTime="334.052" Participant="me011">
      <VocalSound Description="mouth"/> Right.
    </Segment>
    <Segment StartTime="334.212" EndTime="334.471" Participant="me011">
      Ps-
    </Segment>
    <Segment StartTime="335.228" EndTime="341.117" Participant="me011">
      <VocalSound Description="inbreath"/> But for the demo maybe it doesn't matter. I'm not sure whether you wanna do the demo live anyway, or just screen shots of what we have.
    </Segment>
    <Segment StartTime="341.792" EndTime="343.032" Participant="me013">
      <VocalSound Description="breath"/> <VocalSound Description="mouth"/> I don't know.
    </Segment>
    <Segment StartTime="341.811" EndTime="344.504" Participant="me011">
      The problem with doing it live is it takes <Emphasis> so </Emphasis> long to load,
    </Segment>
    <Segment StartTime="345.261" EndTime="346.120" Participant="me011">
      that, um -
    </Segment>
    <Segment StartTime="345.970" EndTime="347.170" Participant="mn017">
      So this - the -
    </Segment>
    <Segment StartTime="347.385" EndTime="355.110" Participant="mn017">
      this - uh, the sluggishness of the loading is all due to the parsing of the X_M_L format. Right?
    </Segment>
    <Segment StartTime="347.568" EndTime="350.637" Participant="fe016">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="348.195" EndTime="348.393" Participant="me013">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="353.208" EndTime="356.595" Participant="me011">
      <VocalSound Description="inbreath"/> Um, w- I was talking to Dave Gelbart about that and apparently it's <Emphasis> not </Emphasis>
    </Segment>
    <Segment StartTime="356.854" EndTime="359.066" Participant="me011">
      actually the parsing of the X_M_L <Emphasis> raw </Emphasis> -
    </Segment>
    <Segment StartTime="359.066" EndTime="359.524" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="359.524" EndTime="359.763" Participant="me011">
      that
    </Segment>
    <Segment StartTime="359.963" EndTime="363.509" Participant="me011">
      going from the X_M_L to an internal <Pause/> t- <Emphasis> tree </Emphasis> structure is pretty fast.
    </Segment>
    <Segment StartTime="363.430" EndTime="364.198" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="363.509" EndTime="363.728" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="363.728" EndTime="367.574" Participant="me011">
      But then it walks the tree to assemble its dat- internal data structures and that's slow.
    </Segment>
    <Segment StartTime="367.002" EndTime="367.565" Participant="mn017">
      Mmm.
    </Segment>
    <Segment StartTime="367.736" EndTime="371.723" Participant="me018">
      <VocalSound Description="mouth"/> Seems like you should be able to spawn that off into a background process, because <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="367.765" EndTime="369.476" Participant="me013">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="371.723" EndTime="374.339" Participant="me018">
      not everything is <Emphasis> displayed </Emphasis> in that tree at once. Right?
    </Segment>
    <Segment StartTime="374.366" EndTime="374.962" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="374.962" EndTime="380.224" Participant="me011">
      Uh, <Emphasis> no. </Emphasis> But what it does is it actually assembles all the user interface components <Emphasis> then. </Emphasis>
    </Segment>
    <Segment StartTime="377.177" EndTime="379.264" Participant="mn017">
      But - d- y-
    </Segment>
    <Segment StartTime="380.581" EndTime="381.100" Participant="me018">
      Right.
    </Segment>
    <Segment StartTime="380.981" EndTime="383.378" Participant="me011">
      And then displays all the user interface components.
    </Segment>
    <Segment StartTime="381.715" EndTime="387.520" Participant="me013">
      <VocalSound Description="inbreath"/> U- I'm - I'm confused. Uh, is - is this downloading something that happens <Emphasis> once? </Emphasis>
    </Segment>
    <Segment StartTime="382.539" EndTime="384.270" Participant="me018">
      Seems like you wanna ass-
    </Segment>
    <Segment StartTime="387.462" EndTime="387.861" Participant="me011">
      Yes.
    </Segment>
    <Segment StartTime="388.178" EndTime="390.900" Participant="me013">
      And - and then when you d- display different things, it's fine?
    </Segment>
    <Segment StartTime="392.151" EndTime="397.485" Participant="mn017">
      No. Whenever you load a new <Pause/> meeting or a new <Pause/> transcript. Right.
    </Segment>
    <Segment StartTime="392.190" EndTime="393.710" Participant="me013">
      So, in that case -
    </Segment>
    <Segment StartTime="394.093" EndTime="395.229" Participant="me011">
      A new transcript.
    </Segment>
    <Segment StartTime="395.045" EndTime="395.591" Participant="mn014">
      Yep.
    </Segment>
    <Segment StartTime="395.647" EndTime="397.645" Participant="me013">
      Well, a new meeting, a transcript. Right. But - but -
    </Segment>
    <Segment StartTime="397.645" EndTime="399.341" Participant="me013">
      i- i- for - but for -
    </Segment>
    <Segment StartTime="397.744" EndTime="398.641" Participant="me011">
      Or audio file.
    </Segment>
    <Segment StartTime="398.720" EndTime="400.513" Participant="me011">
      Well, actually the audio files are pretty fast, too.
    </Segment>
    <Segment StartTime="400.343" EndTime="400.810" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="400.451" EndTime="401.173" Participant="me001">
      Yeah.
    </Segment>
    <Segment StartTime="400.460" EndTime="402.903" Participant="me013">
      for presentation in, uh -
    </Segment>
    <Segment StartTime="403.318" EndTime="405.140" Participant="me013">
      Uh - I wouldn't be -
    </Segment>
    <Segment StartTime="403.679" EndTime="407.540" Participant="mn017">
      You just have to have the thing <Emphasis> running </Emphasis> before you open your laptop.
    </Segment>
    <Segment StartTime="404.100" EndTime="404.678" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="405.320" EndTime="406.060" Participant="me001">
      Yeah.
    </Segment>
    <Segment StartTime="405.999" EndTime="407.463" Participant="fe016">
      Yeah.
    </Segment>
    <Segment StartTime="406.052" EndTime="410.754" Participant="me011">
      Right. The only problem with that is if anything goes <Emphasis> wrong </Emphasis> <Pause/> <Emphasis> or </Emphasis> if you wanna switch from one thing to another.
    </Segment>
    <Segment StartTime="408.155" EndTime="409.274" Participant="fe016">
      Right.
    </Segment>
    <Segment StartTime="408.389" EndTime="409.978" Participant="me001">
      Yeah. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="409.553" EndTime="410.186" Participant="me013">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="410.186" EndTime="412.144" Participant="me013">
      Go <Emphasis> wrong? </Emphasis> <Comment Description="character voice of mock amazement"/> <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="410.927" EndTime="413.350" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="411.070" EndTime="412.909" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="411.083" EndTime="411.751" Participant="mn014">
      <Pronounce Pronunciation="&quot;y-huh&quot;"> Yeah.  </Pronounce>
    </Segment>
    <Segment StartTime="411.385" EndTime="412.370" Participant="me051">
      <VocalSound Description="laugh-breath"/>
    </Segment>
    <Segment StartTime="412.144" EndTime="414.740" Participant="me013">
      I see. Yeah. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="413.071" EndTime="413.902" Participant="fe016">
      Right.
    </Segment>
    <Segment StartTime="413.130" EndTime="413.940" Participant="mn014">
      <VocalSound Description="cough"/>
    </Segment>
    <Segment StartTime="414.610" EndTime="416.237" Participant="fe016">
      I guess for the demo you can always
    </Segment>
    <Segment StartTime="416.680" EndTime="422.596" Participant="fe016">
      play - just <Emphasis> store </Emphasis> the pieces that you're gonna display and play those as separate files, if we can't, you know, actually do it.
    </Segment>
    <Segment StartTime="421.810" EndTime="423.554" Participant="mn017">
      Just make shorter files.
    </Segment>
    <Segment StartTime="423.110" EndTime="425.281" Participant="fe016">
      But it's - you know, it - it's - Yeah.
    </Segment>
    <Segment StartTime="423.260" EndTime="424.763" Participant="me011">
      That's true. We could just subset it.
    </Segment>
    <Segment StartTime="423.554" EndTime="424.990" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="424.863" EndTime="427.234" Participant="me011">
      That's a good idea. That's actually probably the right thing to do.
    </Segment>
    <Segment StartTime="424.990" EndTime="425.820" Participant="mn017">
      Yeah.
    </Segment>
    <Segment StartTime="425.570" EndTime="426.192" Participant="fe016">
      And just
    </Segment>
    <Segment StartTime="426.412" EndTime="427.202" Participant="fe016">
      make it l-
    </Segment>
    <Segment StartTime="426.513" EndTime="427.033" Participant="mn014">
      Yep.
    </Segment>
    <Segment StartTime="427.947" EndTime="428.567" Participant="me011">
      You know, just take
    </Segment>
    <Segment StartTime="428.866" EndTime="431.037" Participant="me011">
      f- ten minutes instead of an hour and a half.
    </Segment>
    <Segment StartTime="429.650" EndTime="431.882" Participant="mn014">
      Yeah. Th- that's what I did for - for my talk.
    </Segment>
    <Segment StartTime="430.493" EndTime="433.120" Participant="me013">
      <VocalSound Description="inbreath"/> Oh! Oh, you're downloading a whole meeting.
    </Segment>
    <Segment StartTime="433.120" EndTime="433.628" Participant="me011">
      Yeah.
    </Segment>
    <Segment StartTime="433.718" EndTime="434.763" Participant="fe016">
      <VocalSound Description="sigh"/>
    </Segment>
    <Segment StartTime="433.876" EndTime="434.840" Participant="me013">
      Oh, yeah that <Uncertain NumSyllables="2"> @@ </Uncertain>.
    </Segment>
    <Segment StartTime="433.998" EndTime="434.508" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="435.421" EndTime="437.612" Participant="me011">
      Yeah. So that - that's actually <Pause/> the - definitely the way to do it.
    </Segment>
    <Segment StartTime="435.740" EndTime="436.582" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="437.859" EndTime="438.254" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="437.911" EndTime="438.660" Participant="me011">
      That's a good idea. <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="438.666" EndTime="439.228" Participant="me013">
      Yeah. <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="440.223" EndTime="444.139" Participant="me013">
      And then <Emphasis> still </Emphasis> do it ahead of time, but then at least you're covered if - if, uh -
    </Segment>
    <Segment StartTime="443.342" EndTime="444.644" Participant="me011">
      Yeah, if there are any problems.
    </Segment>
    <Segment StartTime="444.337" EndTime="446.240" Participant="me013">
      <VocalSound Description="inbreath"/> if there's a problem.
    </Segment>
    <Segment StartTime="446.009" EndTime="446.771" Participant="fe016">
      Right.
    </Segment>
    <Segment StartTime="446.492" EndTime="448.733" Participant="me011">
      Yeah, I mean, even <Emphasis> five </Emphasis> minutes is probably enough.
    </Segment>
    <Segment StartTime="448.147" EndTime="449.456" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="449.084" EndTime="450.728" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="449.156" EndTime="449.802" Participant="mn014">
      Huh.
    </Segment>
    <Segment StartTime="449.409" EndTime="449.993" Participant="fe016">
      Right.
    </Segment>
    <Segment StartTime="450.260" EndTime="456.125" Participant="fe016">
      So, what happened to - Is it st- possible at all to display the words in their <Emphasis> aligned </Emphasis> locations?
    </Segment>
    <Segment StartTime="455.770" EndTime="456.410" Participant="me011">
      That's what I did.
    </Segment>
    <Segment StartTime="456.125" EndTime="460.595" Participant="fe016">
      O_K. So. Sorry, I missed - O_K. Great! But it - O_K.
    </Segment>
    <Segment StartTime="456.509" EndTime="456.788" Participant="me011">
      Yeah.
    </Segment>
    <Segment StartTime="457.020" EndTime="462.016" Participant="mn017">
      You missed that part. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="458.124" EndTime="460.610" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="458.336" EndTime="460.602" Participant="mn014">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="458.386" EndTime="459.375" Participant="me018">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="458.613" EndTime="461.661" Participant="me013">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="460.595" EndTime="465.190" Participant="fe016">
      I couldn't get the words <Emphasis> and </Emphasis> the waveform at the same time for some reason, and there must be some -
    </Segment>
    <Segment StartTime="465.612" EndTime="468.113" Participant="fe016">
      I'll - I'll work on it with Don and see what I'm doing wrong.
    </Segment>
    <Segment StartTime="468.112" EndTime="470.691" Participant="me011">
      Yeah. <Uncertain> I mean, </Uncertain> just ask - Just come by my office. I can show you as well. <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="470.024" EndTime="472.759" Participant="fe016">
      O_K. Great! Well, thanks a lot. That's really great.
    </Segment>
    <Segment StartTime="472.106" EndTime="472.345" Participant="me011">
      Right. <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="473.042" EndTime="473.420" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="473.420" EndTime="475.233" Participant="me011">
      And for the information retrieval,
    </Segment>
    <Segment StartTime="475.412" EndTime="477.636" Participant="me011">
      uh, Don has been working on that. So.
    </Segment>
    <Segment StartTime="477.736" EndTime="480.610" Participant="me001">
      Yeah. So, it's coming along. Um <Pause/>
    </Segment>
    <Segment StartTime="478.150" EndTime="478.738" Participant="me011">
      It looks like it-
    </Segment>
    <Segment StartTime="481.000" EndTime="486.358" Participant="me001">
      Just hacking <Pause/> Dan's code and m- stepping through it. But, I think <Pause/> it's close.
    </Segment>
    <Segment StartTime="485.200" EndTime="494.242" Participant="fe016">
      <NonVocalSound Description="more minor mike noise"/>
    </Segment>
    <Segment StartTime="486.548" EndTime="486.907" Participant="me011">
      Great. <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="486.701" EndTime="489.512" Participant="me001">
      And we should be there pretty soon, with at least, like -
    </Segment>
    <Segment StartTime="489.174" EndTime="490.047" Participant="me013">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="490.060" EndTime="491.290" Participant="me001">
      at least with,
    </Segment>
    <Segment StartTime="491.470" EndTime="494.447" Participant="me001">
      you know, being able to search over
    </Segment>
    <Segment StartTime="492.737" EndTime="494.062" Participant="me013">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="494.985" EndTime="501.462" Participant="me001">
      c- certain amount of meetings, just, like, really basic stuff. Just asking fo- looking for a word, and looking through a bunch of different meetings.
    </Segment>
    <Segment StartTime="496.219" EndTime="498.384" Participant="mn017">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="500.105" EndTime="506.630" Participant="mn017">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="501.462" EndTime="506.097" Participant="me001">
      And if we have time, I'll also add, you know, like, choosing which speakers you wanna include,
    </Segment>
    <Segment StartTime="506.097" EndTime="507.420" Participant="me001">
      and stuff. But -
    </Segment>
    <Segment StartTime="507.448" EndTime="512.801" Participant="me013">
      <Pronounce Pronunciation="&quot;kay&quot;"> O_K.  </Pronounce> Well, I'm gonna start working on this the week after next, so that's the point when I'll need to look more carefully at what y- what -
    </Segment>
    <Segment StartTime="512.801" EndTime="513.950" Participant="me013">
      what you guys have.
    </Segment>
    <Segment StartTime="513.910" EndTime="516.618" Participant="me001">
      So, is the end of the month still <Pause/> the -
    </Segment>
    <Segment StartTime="516.231" EndTime="517.660" Participant="me013">
      <VocalSound Description="mouth"/> Right. The week after - th- th-
    </Segment>
    <Segment StartTime="517.192" EndTime="517.526" Participant="me001">
      the d- ?
    </Segment>
    <Segment StartTime="517.660" EndTime="521.850" Participant="me013">
      The Monday the week after next is July second, which is the first day I get back. So <Pause/>
    </Segment>
    <Segment StartTime="519.770" EndTime="520.600" Participant="me001">
      O_K.
    </Segment>
    <Segment StartTime="521.771" EndTime="522.206" Participant="me001">
      O_K.
    </Segment>
    <Segment StartTime="523.326" EndTime="524.761" Participant="me011">
      <VocalSound Description="inbreath"/> Yeah. So I think for the L-
    </Segment>
    <Segment StartTime="524.860" EndTime="527.451" Participant="me011">
      stuff Liz was talking about, we have something that'll work <Emphasis> now. </Emphasis>
    </Segment>
    <Segment StartTime="527.590" EndTime="530.479" Participant="me011">
      And Liz can look at it and see if she wants anything else.
    </Segment>
    <Segment StartTime="530.479" EndTime="530.942" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="530.942" EndTime="535.460" Participant="me011">
      Maybe we can work on doing - displaying multiple - or displaying one and playing back the other.
    </Segment>
    <Segment StartTime="535.152" EndTime="540.167" Participant="fe016">
      <VocalSound Description="inbreath"/> So do you think it's <Pause/> reasonable to display more than one <Pause/> before the demo? Cuz - <Comment Description="abandoned"/>
    </Segment>
    <Segment StartTime="539.784" EndTime="540.021" Participant="me011">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="540.280" EndTime="545.840" Participant="me011">
      Um, I think I'd h- I'd have to ask Dave. I did it once before and it was just so slow to scroll,
    </Segment>
    <Segment StartTime="540.814" EndTime="542.306" Participant="me013">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="546.348" EndTime="547.459" Participant="fe016">
      Huh.
    </Segment>
    <Segment StartTime="546.460" EndTime="547.264" Participant="me011">
      that I gave up.
    </Segment>
    <Segment StartTime="547.600" EndTime="548.055" Participant="me011">
      But,
    </Segment>
    <Segment StartTime="548.290" EndTime="553.030" Participant="me011">
      the advantage is that these things are much lower sampling rate. And so then it might be alright.
    </Segment>
    <Segment StartTime="551.012" EndTime="551.870" Participant="fe016">
      Right.
    </Segment>
    <Segment StartTime="552.718" EndTime="554.147" Participant="fe016">
      O_K. Let me know.
    </Segment>
    <Segment StartTime="554.263" EndTime="555.472" Participant="me001">
      Morgan, when's the demo?
    </Segment>
    <Segment StartTime="556.011" EndTime="558.721" Participant="me013">
      <VocalSound Description="mouth"/> Well, uh, I'm giving a talk
    </Segment>
    <Segment StartTime="559.030" EndTime="559.720" Participant="me013">
      on
    </Segment>
    <Segment StartTime="559.720" EndTime="560.327" Participant="me013">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="560.920" EndTime="562.240" Participant="me013">
      July
    </Segment>
    <Segment StartTime="562.590" EndTime="565.257" Participant="me013">
      sixteenth. It's a Mon- Monday in
    </Segment>
    <Segment StartTime="565.585" EndTime="567.580" Participant="mn017">
      <VocalSound Description="mouth"/> <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="566.225" EndTime="567.142" Participant="me013">
      four weeks?
    </Segment>
    <Segment StartTime="567.580" EndTime="568.380" Participant="mn017">
      So -
    </Segment>
    <Segment StartTime="567.894" EndTime="568.589" Participant="me013">
      Three weeks?
    </Segment>
    <Segment StartTime="568.911" EndTime="573.106" Participant="mn017">
      If - if raw speed is the problem <Pause/> This thing is written in Tcl. Right?
    </Segment>
    <Segment StartTime="572.307" EndTime="572.703" Participant="me011">
      Tcl.
    </Segment>
    <Segment StartTime="574.030" EndTime="576.680" Participant="mn017">
      Y- I mean, John Osterhout, uh -
    </Segment>
    <Segment StartTime="576.680" EndTime="583.389" Participant="mn017">
      Y- you know, he started his own company based on Tcl stuff, and maybe they have the native code compiler or something. <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="583.001" EndTime="585.330" Participant="me011">
      I mean, we could check. I <Emphasis> don't </Emphasis> think they do.
    </Segment>
    <Segment StartTime="586.240" EndTime="591.053" Participant="me011">
      Um, there was actually a <Pause/> Java back-end that apparently is actually a little faster.
    </Segment>
    <Segment StartTime="591.172" EndTime="592.379" Participant="me011">
      It generates byte code.
    </Segment>
    <Segment StartTime="591.195" EndTime="591.783" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="591.381" EndTime="591.638" Participant="fe016">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="592.379" EndTime="592.715" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="592.715" EndTime="593.684" Participant="me011">
      But, uh -
    </Segment>
    <Segment StartTime="594.643" EndTime="596.965" Participant="me013">
      It's always exciting to hear that Java's faster than <Emphasis> something. </Emphasis>
    </Segment>
    <Segment StartTime="596.829" EndTime="597.856" Participant="me011">
      Yeah. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="596.965" EndTime="600.367" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="597.173" EndTime="598.476" Participant="me001">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="597.219" EndTime="597.714" Participant="fe008">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="597.690" EndTime="599.806" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="597.856" EndTime="598.528" Participant="me011">
      Well,
    </Segment>
    <Segment StartTime="598.528" EndTime="603.137" Participant="me011">
      e- everything is faster than Tcl-T_K. It's a string substitution language, basically.
    </Segment>
    <Segment StartTime="600.367" EndTime="601.018" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="601.655" EndTime="604.843" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="603.125" EndTime="605.669" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="603.513" EndTime="606.855" Participant="me011">
      I should probably beep that out in <Pronounce> case  </Pronounce> John Osterhout ever listens. But -
    </Segment>
    <Segment StartTime="604.780" EndTime="607.011" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="605.855" EndTime="609.396" Participant="me018">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="605.890" EndTime="607.465" Participant="fe008">
      But Tcl is <Emphasis> wonderful. </Emphasis>
    </Segment>
    <Segment StartTime="607.465" EndTime="611.146" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="607.619" EndTime="611.112" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="607.860" EndTime="608.931" Participant="me011">
      Well, it <Emphasis> is </Emphasis> wonderful.
    </Segment>
    <Segment StartTime="607.911" EndTime="610.460" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="608.017" EndTime="609.380" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="608.670" EndTime="610.271" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="608.706" EndTime="609.660" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="608.781" EndTime="610.564" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="608.931" EndTime="613.124" Participant="me011">
      It is, for prototyping and user interface. It's just really - the language is awful.
    </Segment>
    <Segment StartTime="611.370" EndTime="612.200" Participant="fe008">
      There you go.
    </Segment>
    <Segment StartTime="611.639" EndTime="613.444" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="613.502" EndTime="614.735" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="613.660" EndTime="614.130" Participant="fe008">
      Oh.
    </Segment>
    <Segment StartTime="614.300" EndTime="620.996" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="614.355" EndTime="616.450" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="614.447" EndTime="617.232" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="614.699" EndTime="619.171" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="614.735" EndTime="620.990" Participant="me013">
      Beep. <Comment Description="nonword, imitating censor's tone"/> <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="614.737" EndTime="618.374" Participant="fe008">
      <VocalSound Description="laugh"/> Yeah. <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="614.740" EndTime="619.026" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="615.064" EndTime="616.770" Participant="me011">
      <VocalSound Description="laugh"/> Beep, y- right. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="617.500" EndTime="619.414" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="619.053" EndTime="622.176" Participant="me001">
      <VocalSound Description="breath-laugh"/> <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="619.723" EndTime="622.860" Participant="me011">
      But let me tell you how I <Emphasis> really </Emphasis> feel. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="620.990" EndTime="622.389" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="621.077" EndTime="625.897" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="621.180" EndTime="622.742" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="621.183" EndTime="622.172" Participant="me018">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="621.236" EndTime="622.650" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="621.492" EndTime="623.181" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="622.389" EndTime="624.182" Participant="me013">
      We're all entitled to our opinions here.
    </Segment>
    <Segment StartTime="624.182" EndTime="625.290" Participant="me013">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="624.230" EndTime="624.754" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="624.520" EndTime="625.949" Participant="fe008">
      Yeah. I like it. It's great. <Comment Description="last two words very quietly"/>
    </Segment>
    <Segment StartTime="625.290" EndTime="629.624" Participant="me013">
      Um. Yeah. So it's - Yeah. I think it must be three and a half weeks. Uh, cuz July -
    </Segment>
    <Segment StartTime="630.158" EndTime="632.920" Participant="me013">
      <VocalSound Description="mouth"/> The - the meeting is July sixteenth through eighteenth.
    </Segment>
    <Segment StartTime="633.307" EndTime="635.800" Participant="me013">
      And, uh, my talk's the first day. So.
    </Segment>
    <Segment StartTime="635.132" EndTime="636.010" Participant="me001">
      O_K.
    </Segment>
    <Segment StartTime="637.062" EndTime="639.098" Participant="me013">
      I'm flying out there the Sunday before. So.
    </Segment>
    <Segment StartTime="639.676" EndTime="640.208" Participant="me013">
      Um -
    </Segment>
    <Segment StartTime="640.376" EndTime="642.342" Participant="mn017">
      <VocalSound Description="sniff"/> Hmm.
    </Segment>
    <Segment StartTime="642.342" EndTime="643.908" Participant="mn017">
      <VocalSound Description="three tongue taps"/> <Comment Description="vocal gesture"/>
    </Segment>
    <Segment StartTime="644.495" EndTime="645.600" Participant="me013">
      I guess, you know, it'd be
    </Segment>
    <Segment StartTime="646.040" EndTime="649.891" Participant="me013">
      desirable <Pause/> if, a week ahead of that, we basically had - <Emphasis> thought </Emphasis> we had it,
    </Segment>
    <Segment StartTime="649.891" EndTime="651.216" Participant="me013">
      which would allow a week for
    </Segment>
    <Segment StartTime="651.216" EndTime="653.057" Participant="me013">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="651.813" EndTime="652.928" Participant="me001">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="652.538" EndTime="653.785" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="652.928" EndTime="657.340" Participant="me001">
      For realizing we don't? <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="652.960" EndTime="653.986" Participant="me011">
      Oh, that's right.
    </Segment>
    <Segment StartTime="653.057" EndTime="653.738" Participant="me013">
      re-iterating.
    </Segment>
    <Segment StartTime="653.738" EndTime="656.280" Participant="me013">
      <VocalSound Description="laugh"/> Yeah. Pretty much. Yeah.
    </Segment>
    <Segment StartTime="654.067" EndTime="656.170" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="654.228" EndTime="657.040" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="654.470" EndTime="657.162" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="655.810" EndTime="656.053" Participant="me011">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="656.310" EndTime="657.260" Participant="me011">
      Then the other
    </Segment>
    <Segment StartTime="657.820" EndTime="662.581" Participant="me011">
      issue related to that is data release. If we wanna show this in public, it should be
    </Segment>
    <Segment StartTime="662.759" EndTime="667.245" Participant="me011">
      releas- So, I, uh, haven't gotten <Emphasis> any </Emphasis> other replies from the original email
    </Segment>
    <Segment StartTime="667.590" EndTime="671.052" Participant="me011">
      asking for approval. So I sent out another set this morning.
    </Segment>
    <Segment StartTime="669.554" EndTime="670.284" Participant="fe016">
      <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="669.823" EndTime="670.179" Participant="me013">
      <VocalSound Description="sniff"/>
    </Segment>
    <Segment StartTime="670.459" EndTime="671.598" Participant="me013">
      I th- I saw that.
    </Segment>
    <Segment StartTime="671.573" EndTime="673.595" Participant="me011">
      And, uh, we'll see if we get any responses.
    </Segment>
    <Segment StartTime="672.236" EndTime="676.830" Participant="fe016">
      I just did it. <VocalSound Description="laugh"/> But it <Emphasis> is </Emphasis> - I did wanna say that, um -
    </Segment>
    <Segment StartTime="673.595" EndTime="674.580" Participant="me011">
      Very good. <Comment Description="last two words while laughing"/>
    </Segment>
    <Segment StartTime="674.931" EndTime="675.663" Participant="me018">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="676.898" EndTime="678.730" Participant="me011">
      Did you notice I put in the filter?
    </Segment>
    <Segment StartTime="679.187" EndTime="680.140" Participant="fe016">
      No!
    </Segment>
    <Segment StartTime="680.514" EndTime="683.818" Participant="fe016">
      No! No! <VocalSound Description="laugh"/> I just figured you p-
    </Segment>
    <Segment StartTime="680.609" EndTime="682.951" Participant="me011">
      Go ahead. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="683.479" EndTime="686.408" Participant="me011">
      There's a link there that now says if you want to search by -
    </Segment>
    <Segment StartTime="686.337" EndTime="687.890" Participant="fe016">
      Oh, my gosh.
    </Segment>
    <Segment StartTime="686.685" EndTime="688.890" Participant="me011">
      filter by a regular expression, you can.
    </Segment>
    <Segment StartTime="688.021" EndTime="688.732" Participant="fe016">
      O_K.
    </Segment>
    <Segment StartTime="689.846" EndTime="690.894" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="690.071" EndTime="691.430" Participant="me011">
      I put that in just for you.
    </Segment>
    <Segment StartTime="690.121" EndTime="691.381" Participant="fe016">
      Terrific. <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="691.381" EndTime="696.879" Participant="fe016">
      Well, since you didn't answer the emai- So there was a q- question I had asked Adam whether it's possible to search only for your own
    </Segment>
    <Segment StartTime="691.943" EndTime="692.772" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="697.120" EndTime="701.634" Participant="fe016">
      name - your own utterances, so that you know you don't have to go through the whole meeting, and - <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="701.634" EndTime="707.550" Participant="fe016">
      Um, and I didn't hear back. So I thought "O_K. It's probably too hard. He's overloaded. I won't say anything. I'll just <Emphasis> do </Emphasis> it". <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="706.476" EndTime="709.380" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="707.550" EndTime="707.904" Participant="fe016">
      Great.
    </Segment>
    <Segment StartTime="708.258" EndTime="711.640" Participant="fe016">
      O_K. So, anyway I looked at everybody else's stu- <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="709.380" EndTime="713.773" Participant="me011">
      It's actually an arbitrary - arbitrary regular expression. But if you search your name,
    </Segment>
    <Segment StartTime="712.484" EndTime="713.032" Participant="fe016">
      Good.
    </Segment>
    <Segment StartTime="713.991" EndTime="717.650" Participant="me011">
      you'll get all of the things you said and any time anyone said your name. So.
    </Segment>
    <Segment StartTime="714.040" EndTime="715.920" Participant="fe016">
      So th- <VocalSound Description="click"/>
    </Segment>
    <Segment StartTime="716.979" EndTime="717.944" Participant="fe016">
      That's great.
    </Segment>
    <Segment StartTime="717.309" EndTime="720.060" Participant="me013">
      And it's - it's case insensitive? Yeah.
    </Segment>
    <Segment StartTime="719.173" EndTime="720.190" Participant="me011">
      Correct.
    </Segment>
    <Segment StartTime="719.725" EndTime="720.605" Participant="fe016">
      That's great.
    </Segment>
    <Segment StartTime="722.120" EndTime="725.537" Participant="me018">
      Did you actually look through your <Pause/> transcripts? Or - you just approved <Pronounce Pronunciation="&quot;em&quot;"> them  </Pronounce> all.
    </Segment>
    <Segment StartTime="724.481" EndTime="727.120" Participant="fe016">
      Uh <Pause/> Well <Pause/> <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="726.345" EndTime="728.002" Participant="me018">
      I just approved all mine. I didn't look at them.
    </Segment>
    <Segment StartTime="727.120" EndTime="730.491" Participant="fe016">
      <VocalSound Description="vocal squeak"/> I sort of spot-checked. I was trying to remember -
    </Segment>
    <Segment StartTime="728.002" EndTime="729.657" Participant="me018">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="728.622" EndTime="731.770" Participant="mn017">
      <VocalSound Description="mouth"/> Oh, <Emphasis> darn. </Emphasis> I haven't done that yet. Er - O_K.
    </Segment>
    <Segment StartTime="731.300" EndTime="732.309" Participant="fe016">
      I couldn't find the -
    </Segment>
    <Segment StartTime="732.922" EndTime="735.870" Participant="fe016">
      the keywords for things <VocalSound Description="laugh"/> that I thought I had said wrong. So.
    </Segment>
    <Segment StartTime="735.618" EndTime="739.198" Participant="mn014">
      <Uncertain NumSyllables="1"> @@ </Uncertain> <Comment Description="uncodeable vocal sound"/> <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="736.343" EndTime="737.490" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="736.370" EndTime="737.660" Participant="me011">
      It's hard to find.
    </Segment>
    <Segment StartTime="738.106" EndTime="739.130" Participant="fe016">
      That makes it g-
    </Segment>
    <Segment StartTime="738.430" EndTime="739.543" Participant="me018">
      That's a <Emphasis> compliment </Emphasis> to you.
    </Segment>
    <Segment StartTime="739.815" EndTime="741.810" Participant="me018">
      He said it's hard to find things you say wrong.
    </Segment>
    <Segment StartTime="741.734" EndTime="744.660" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="741.933" EndTime="742.549" Participant="me011">
      Hmm.
    </Segment>
    <Segment StartTime="742.330" EndTime="745.700" Participant="fe016">
      It's hard to find <Emphasis> anything </Emphasis> that you say in these. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="742.542" EndTime="746.799" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="742.670" EndTime="746.320" Participant="me018">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="742.819" EndTime="745.271" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="743.469" EndTime="743.990" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="745.700" EndTime="750.990" Participant="fe016">
      Great. Well, thanks for the filter. Uh, it's really useful, tha- Cuz if you're only at part of a meeting, or something <Pause/>
    </Segment>
    <Segment StartTime="746.032" EndTime="747.595" Participant="me011">
      You really <Emphasis> do </Emphasis> have to sort of r-
    </Segment>
    <Segment StartTime="751.080" EndTime="752.941" Participant="me013">
      So, we have our first information retrieval
    </Segment>
    <Segment StartTime="752.776" EndTime="754.173" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="752.941" EndTime="756.680" Participant="me013">
      example. It's a regular expression <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="753.795" EndTime="754.667" Participant="me011">
      Yeah. That's right.
    </Segment>
    <Segment StartTime="754.667" EndTime="757.535" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="756.442" EndTime="758.870" Participant="fe016">
      Yeah! That's actually - Well, it's <Emphasis> useful. </Emphasis>
    </Segment>
    <Segment StartTime="756.897" EndTime="757.253" Participant="me013">
      searcher.
    </Segment>
    <Segment StartTime="757.253" EndTime="760.595" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="758.390" EndTime="762.124" Participant="me011">
      And it demonstrates why it doesn't work, because you really wanna go acro- more than one meeting.
    </Segment>
    <Segment StartTime="761.600" EndTime="762.034" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="762.542" EndTime="762.991" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="762.840" EndTime="766.199" Participant="me011">
      And you need a better user interface for displaying the results. So.
    </Segment>
    <Segment StartTime="765.550" EndTime="767.030" Participant="fe016">
      But this helps a lot.
    </Segment>
    <Segment StartTime="766.654" EndTime="767.052" Participant="me011">
      Yeah.
    </Segment>
    <Segment StartTime="767.791" EndTime="771.360" Participant="me013">
      You wanna say, "Where are al- where - Find all the contentious things I said."
    </Segment>
    <Segment StartTime="767.995" EndTime="769.199" Participant="fe016">
      Great. Thanks.
    </Segment>
    <Segment StartTime="771.204" EndTime="771.750" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="771.323" EndTime="772.168" Participant="me011">
      Yeah, really. <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="771.360" EndTime="773.615" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="771.690" EndTime="774.347" Participant="me018">
      <VocalSound Description="laugh"/> Find everything that should be bleeped.
    </Segment>
    <Segment StartTime="772.168" EndTime="774.677" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="773.871" EndTime="775.572" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="773.931" EndTime="775.560" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="774.563" EndTime="777.711" Participant="me018">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="774.677" EndTime="777.910" Participant="me011">
      That's right. Th- we <Emphasis> do </Emphasis> have that bi- nice marker - is that, n- n-
    </Segment>
    <Segment StartTime="775.814" EndTime="776.150" Participant="fe008">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="778.030" EndTime="783.613" Participant="me011">
      because we all know we're being recorded, whenever anyone says anything like that, we then have a conversation about bleeping it out. So.
    </Segment>
    <Segment StartTime="779.139" EndTime="788.328" Participant="mn017">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="779.858" EndTime="780.670" Participant="fe016">
      Right.
    </Segment>
    <Segment StartTime="781.164" EndTime="781.924" Participant="fe016">
      We-
    </Segment>
    <Segment StartTime="783.122" EndTime="785.160" Participant="fe016">
      Yeah. You can search for "beep" or "bleep".
    </Segment>
    <Segment StartTime="783.198" EndTime="783.841" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="783.858" EndTime="784.827" Participant="me018">
      <Uncertain NumSyllables="2"> @@ </Uncertain> <Comment Description="whispered"/>
    </Segment>
    <Segment StartTime="784.932" EndTime="785.660" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="785.544" EndTime="786.187" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="786.670" EndTime="788.100" Participant="fe016">
      In somebody <Emphasis> else's </Emphasis> turn.
    </Segment>
    <Segment StartTime="790.713" EndTime="791.069" Participant="me011">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="791.540" EndTime="796.786" Participant="me011">
      Um. <Emphasis> Oh. </Emphasis> And also we actually have a few people who <Pronounce> have  </Pronounce> still not filled out <Emphasis> speaker </Emphasis> forms.
    </Segment>
    <Segment StartTime="797.279" EndTime="827.987" Participant="fe016">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="797.382" EndTime="800.843" Participant="me011">
      Specifically in the N_S_A ones, and I noticed that when I tried to, uh -
    </Segment>
    <Segment StartTime="799.826" EndTime="801.254" Participant="me013">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="801.560" EndTime="802.560" Participant="me011">
      uh, generate
    </Segment>
    <Segment StartTime="803.180" EndTime="806.451" Participant="me011">
      the transcripts for N_S_A. That there are a few with no speaker forms.
    </Segment>
    <Segment StartTime="806.960" EndTime="808.535" Participant="me011">
      And so, uh, I have a -
    </Segment>
    <Segment StartTime="809.510" EndTime="818.030" Participant="me011">
      I sent out yet another this morning, which I think makes six total emails that I've sent to these people, and so I think we need to escalate to some other method of trying to contact them.
    </Segment>
    <Segment StartTime="818.068" EndTime="821.645" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="819.286" EndTime="821.130" Participant="me013">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="821.590" EndTime="822.330" Participant="me013">
      Um <Pause/>
    </Segment>
    <Segment StartTime="823.423" EndTime="823.745" Participant="me013">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="824.263" EndTime="824.873" Participant="me013">
      Right.
    </Segment>
    <Segment StartTime="824.720" EndTime="826.970" Participant="mn017">
      Stalk them <Uncertain NumSyllables="1"> @@ </Uncertain> at their - <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="825.624" EndTime="828.950" Participant="me013">
      <VocalSound Description="inbreath"/> Has - has - has Joachim Sokol replied? Or - ?
    </Segment>
    <Segment StartTime="825.664" EndTime="826.802" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="826.970" EndTime="829.400" Participant="mn017">
      Like in the morning, when I leave for work.
    </Segment>
    <Segment StartTime="828.146" EndTime="828.643" Participant="me011">
      Nope.
    </Segment>
    <Segment StartTime="829.207" EndTime="832.090" Participant="fe008">
      <VocalSound Description="mouth"/> I think, maybe talk to him first in person? That's what I would <Pause/> think.
    </Segment>
    <Segment StartTime="832.703" EndTime="834.910" Participant="me013">
      He's <Pause/> not around, is the only problem.
    </Segment>
    <Segment StartTime="833.140" EndTime="833.881" Participant="fe008">
      <VocalSound Description="clears throat"/> Mm-hmm.
    </Segment>
    <Segment StartTime="834.791" EndTime="837.676" Participant="fe008">
      Oh, is that right? Oh. I saw him - I saw him on Tuesday.
    </Segment>
    <Segment StartTime="835.530" EndTime="837.569" Participant="me013">
      Yeah. Otherwise, it'd be a good idea.
    </Segment>
    <Segment StartTime="836.474" EndTime="837.146" Participant="mn017">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="837.146" EndTime="837.824" Participant="mn017">
      <Uncertain NumSyllables="2"> @@ </Uncertain> <Comment Description="uncodeable vocal sounds"/>
    </Segment>
    <Segment StartTime="837.569" EndTime="837.985" Participant="me013">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="837.824" EndTime="839.140" Participant="mn017">
      <VocalSound Description="inbreath"/> <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="837.985" EndTime="840.556" Participant="me013">
      Yeah. He popped in. But, I mean, he's basically gone.
    </Segment>
    <Segment StartTime="839.090" EndTime="839.960" Participant="fe008">
      Oh, O_K.
    </Segment>
    <Segment StartTime="839.140" EndTime="842.470" Participant="mn017">
      Cold calling at lunch time? Uh, dinner time, I mean. <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="842.470" EndTime="848.633" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="842.635" EndTime="845.660" Participant="me011">
      Well, if I could find phone numbers, that would certainly work. But.
    </Segment>
    <Segment StartTime="842.873" EndTime="847.048" Participant="me018">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="843.542" EndTime="846.324" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="844.660" EndTime="848.020" Participant="fe008">
      Well, did you ask Lila? Cuz I bet <Emphasis> she </Emphasis> has this information.
    </Segment>
    <Segment StartTime="847.061" EndTime="850.250" Participant="me011">
      Yeah, that's a good idea. I'll ask her if she can con- track some of them down.
    </Segment>
    <Segment StartTime="848.820" EndTime="849.280" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="848.960" EndTime="849.315" Participant="fe008">
      Yeah.
    </Segment>
    <Segment StartTime="849.770" EndTime="851.637" Participant="me013">
      Yeah. And - and tell her - You know.
    </Segment>
    <Segment StartTime="851.637" EndTime="854.360" Participant="me013">
      Tell her your specific problem. She'll fix -
    </Segment>
    <Segment StartTime="853.139" EndTime="853.807" Participant="fe008">
      <VocalSound Description="inbreath"/> And it -
    </Segment>
    <Segment StartTime="854.045" EndTime="854.345" Participant="fe008">
      M-
    </Segment>
    <Segment StartTime="854.754" EndTime="856.080" Participant="fe008">
      And, uh, then there's still m- um -
    </Segment>
    <Segment StartTime="856.080" EndTime="860.810" Participant="fe008">
      <Emphasis> Miguel </Emphasis> is still an active member of the group and he's - What I mean is, he's an active member and he's still <Emphasis> here. </Emphasis>
    </Segment>
    <Segment StartTime="859.286" EndTime="860.999" Participant="me001">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="860.142" EndTime="861.540" Participant="me013">
      He's right there. Yeah.
    </Segment>
    <Segment StartTime="861.152" EndTime="861.642" Participant="me011">
      Mm-hmm.
    </Segment>
    <Segment StartTime="861.745" EndTime="862.705" Participant="fe008">
      Very helpful.
    </Segment>
    <Segment StartTime="862.079" EndTime="863.288" Participant="me011">
      Yeah, I didn't actually
    </Segment>
    <Segment StartTime="863.590" EndTime="864.809" Participant="me011">
      see who they all were.
    </Segment>
    <Segment StartTime="865.080" EndTime="872.480" Participant="me011">
      Um, a <Emphasis> couple </Emphasis> of them were, like, people at I_B_M who were here for one of the I_B_M meetings and one - a guy from S_R_I who was at one of the S_R_I meetings.
    </Segment>
    <Segment StartTime="865.380" EndTime="865.643" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="872.480" EndTime="874.650" Participant="me011">
      And so, uh, <Emphasis> those </Emphasis> might be harder to track down.
    </Segment>
    <Segment StartTime="875.095" EndTime="878.655" Participant="fe008">
      <VocalSound Description="mouth"/> <Emphasis> Most </Emphasis> of them, though, really <Emphasis> were </Emphasis> visitors here and Lila should have contacts with them.
    </Segment>
    <Segment StartTime="878.058" EndTime="878.508" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="878.655" EndTime="880.121" Participant="fe008">
      Nice meetings, by the way. Yeah.
    </Segment>
    <Segment StartTime="881.433" EndTime="884.830" Participant="me011">
      I mean, they - they were people who didn't have accounts at ICSI, so they're - they're harder to find.
    </Segment>
    <Segment StartTime="883.800" EndTime="884.347" Participant="fe008">
      <VocalSound Description="mouth"/> <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="884.830" EndTime="886.030" Participant="fe008">
      Well, <Emphasis> not </Emphasis> the ones that -
    </Segment>
    <Segment StartTime="886.030" EndTime="886.835" Participant="fe008">
      Are <Emphasis> you </Emphasis> su- ?
    </Segment>
    <Segment StartTime="887.550" EndTime="888.690" Participant="fe008">
      Are you sure? <Comment Description="rise-fall-rise intonation, surprised"/>
    </Segment>
    <Segment StartTime="888.620" EndTime="889.621" Participant="me011">
      Am I sure about what?
    </Segment>
    <Segment StartTime="889.180" EndTime="891.860" Participant="fe008">
      Yeah. N_S_A-<Emphasis> one </Emphasis> and N_S_A-<Emphasis> three? </Emphasis> We're talking about those?
    </Segment>
    <Segment StartTime="890.709" EndTime="892.243" Participant="me011">
      There were other people also.
    </Segment>
    <Segment StartTime="892.243" EndTime="896.810" Participant="me011">
      There were other c- other people also who didn't ha- fill out the speaker forms, in addition to the N_S_As. Yeah.
    </Segment>
    <Segment StartTime="892.280" EndTime="892.820" Participant="fe008">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="894.720" EndTime="896.660" Participant="me013">
      I- i- in - in other meetings. Yeah.
    </Segment>
    <Segment StartTime="896.660" EndTime="899.112" Participant="fe008">
      Oh, oh. I see. O_K. Fine. <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="897.020" EndTime="900.889" Participant="me013">
      <VocalSound Description="inbreath"/> Well, S_R_I people, easy to f- find. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="899.354" EndTime="901.099" Participant="fe016">
      Yeah. <Comment Description="very quietly"/>
    </Segment>
    <Segment StartTime="899.730" EndTime="900.289" Participant="me011">
      Yeah. <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="900.889" EndTime="904.469" Participant="me013">
      And, uh, I_B_M people, also. Just let us know. I mean -
    </Segment>
    <Segment StartTime="904.712" EndTime="909.532" Participant="fe016">
      <NonVocalSound Description="page being turned"/>
    </Segment>
    <Segment StartTime="904.845" EndTime="906.554" Participant="me013">
      Just - we certainly have their email.
    </Segment>
    <Segment StartTime="906.292" EndTime="908.882" Participant="fe008">
      <VocalSound Description="mouth"/> But I knew everybody in the N_S_A meetings.
    </Segment>
    <Segment StartTime="908.435" EndTime="908.954" Participant="me011">
      Mm-hmm.
    </Segment>
    <Segment StartTime="908.882" EndTime="912.560" Participant="fe008">
      So th- I'm sure that we have, uh, fresh, you know, information on them.
    </Segment>
    <Segment StartTime="911.330" EndTime="911.811" Participant="me011">
      Right.
    </Segment>
    <Segment StartTime="913.454" EndTime="916.737" Participant="me011">
      Yeah. None of the e- emails bounced, so I know they're going somewhere.
    </Segment>
    <Segment StartTime="916.569" EndTime="917.243" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="916.570" EndTime="917.690" Participant="fe008">
      Good. O_K.
    </Segment>
    <Segment StartTime="916.760" EndTime="917.158" Participant="me013">
      Right.
    </Segment>
    <Segment StartTime="917.324" EndTime="918.634" Participant="me013">
      <VocalSound Description="sniff"/>
    </Segment>
    <Segment StartTime="917.870" EndTime="918.610" Participant="me011">
      <VocalSound Description="clears throat"/>
    </Segment>
    <Segment StartTime="921.039" EndTime="922.090" Participant="me011">
      That's all I have.
    </Segment>
    <Segment StartTime="923.130" EndTime="924.770" Participant="me011">
      You wanna talk about recognition?
    </Segment>
    <Segment StartTime="924.347" EndTime="925.830" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="925.830" EndTime="929.140" Participant="mn017">
      Chuck, you wanna talk about recognition? <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="927.146" EndTime="929.614" Participant="me018">
      <VocalSound Description="breath-laugh"/> I haven't done anything. I was -
    </Segment>
    <Segment StartTime="927.552" EndTime="928.369" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="928.369" EndTime="931.906" Participant="me011">
      J- T- Liz, you wanna talk about recognition? Thilo, you wanna talk about recognition? <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="929.254" EndTime="932.070" Participant="me013">
      <VocalSound Description="laugh"/> <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="929.319" EndTime="931.890" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="929.498" EndTime="932.064" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="929.614" EndTime="930.342" Participant="me018">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="929.884" EndTime="931.630" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="930.326" EndTime="930.997" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="930.342" EndTime="932.172" Participant="me018">
      I was away for a couple of days. <Comment Description="while laughing"/> So.
    </Segment>
    <Segment StartTime="930.759" EndTime="932.651" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="932.172" EndTime="933.460" Participant="me018">
      I haven't done a - <Comment Description="abandoned"/>
    </Segment>
    <Segment StartTime="932.651" EndTime="933.874" Participant="fe016">
      We're sort of in a
    </Segment>
    <Segment StartTime="932.806" EndTime="934.336" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="933.187" EndTime="939.154" Participant="me013">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="934.214" EndTime="940.738" Participant="fe016">
      stage where we're - uh, Don's going through getting some of the next meetings that Jane s- has. And, uh,
    </Segment>
    <Segment StartTime="941.534" EndTime="947.780" Participant="fe016">
      you know, creating a second database. So we haven't actually run anything yet. We need to get a critical mass for that.
    </Segment>
    <Segment StartTime="947.353" EndTime="950.599" Participant="mn017">
      However, I just got an email from Thilo saying that
    </Segment>
    <Segment StartTime="950.974" EndTime="952.577" Participant="mn017">
      we are ready to run -
    </Segment>
    <Segment StartTime="952.893" EndTime="955.455" Participant="mn017">
      I mean, we have segmentations for the <Emphasis> old </Emphasis> meetings
    </Segment>
    <Segment StartTime="953.178" EndTime="956.251" Participant="fe016">
      And check the segmentations. Yeah.
    </Segment>
    <Segment StartTime="955.455" EndTime="958.081" Participant="mn017">
      that <Pause/> are from his segmenter, and so -
    </Segment>
    <Segment StartTime="955.612" EndTime="956.512" Participant="me001">
      Mm-hmm.
    </Segment>
    <Segment StartTime="958.081" EndTime="964.470" Participant="mn017">
      <NonVocalSound Description="mike noise"/> You - you had three different versions, with different, like, pause thresholds <Pause/> between the segments?
    </Segment>
    <Segment StartTime="958.590" EndTime="959.295" Participant="fe016">
      Great.
    </Segment>
    <Segment StartTime="961.188" EndTime="962.108" Participant="mn014">
      Yep.
    </Segment>
    <Segment StartTime="962.746" EndTime="963.151" Participant="mn014">
      Yep.
    </Segment>
    <Segment StartTime="964.370" EndTime="968.032" Participant="mn014">
      Yeah. Just - Yeah. Just smooth the - the output of the - of the
    </Segment>
    <Segment StartTime="967.034" EndTime="972.820" Participant="mn017">
      Right. And you recommended using the one with two s- maximum of two seconds? But two s-
    </Segment>
    <Segment StartTime="968.324" EndTime="968.982" Participant="mn014">
      detector.
    </Segment>
    <Segment StartTime="970.461" EndTime="971.105" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="971.690" EndTime="974.440" Participant="fe016">
      A - What do you mean, a different pause threshold? Do you mean a - ?
    </Segment>
    <Segment StartTime="971.691" EndTime="972.172" Participant="mn014">
      Yep.
    </Segment>
    <Segment StartTime="971.884" EndTime="972.620" Participant="me001">
      <VocalSound Description="ingressive whistle"/> <Comment Description="maybe unintentional, otherwise a vocal gesture"/>
    </Segment>
    <Segment StartTime="972.599" EndTime="975.067" Participant="mn014">
      <VocalSound Description="inbreath"/> Or you can - Yeah. You can use
    </Segment>
    <Segment StartTime="975.504" EndTime="981.730" Participant="mn014">
      the one with one second or whatever. I - I - I - There's no - not much difference between the - the one-second and the two-second one.
    </Segment>
    <Segment StartTime="979.484" EndTime="980.109" Participant="mn017">
      Mm-hmm.
    </Segment>
    <Segment StartTime="981.526" EndTime="987.599" Participant="mn017">
      <VocalSound Description="mouth"/> I mean, the only advantage to using the longer threshold would be that you run less risk of missing some -
    </Segment>
    <Segment StartTime="987.688" EndTime="988.382" Participant="me011">
      Backchannel. <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="988.017" EndTime="989.880" Participant="mn017">
      some speech. Right?
    </Segment>
    <Segment StartTime="989.430" EndTime="991.545" Participant="mn014">
      And I think - wouldn't it be better
    </Segment>
    <Segment StartTime="991.545" EndTime="992.560" Participant="mn014">
      to - to have
    </Segment>
    <Segment StartTime="992.560" EndTime="999.509" Participant="mn014">
      a little longer sequences for the recognizer? B- eh- because of the language model? As sometimes it happens that - that it cuts off within a s-
    </Segment>
    <Segment StartTime="995.330" EndTime="996.300" Participant="fe016">
      Yeah.
    </Segment>
    <Segment StartTime="997.002" EndTime="1000.147" Participant="mn017">
      <VocalSound Description="mouth"/> But two seconds is pretty long. So - <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="997.965" EndTime="999.398" Participant="me013">
      <VocalSound Description="clears throat"/>
    </Segment>
    <Segment StartTime="999.398" EndTime="1001.007" Participant="me013">
      <VocalSound Description="sniff"/>
    </Segment>
    <Segment StartTime="1000.155" EndTime="1003.185" Participant="mn014">
      Yeah. But, we can be sure that as - Or we can be - er-
    </Segment>
    <Segment StartTime="1000.360" EndTime="1001.427" Participant="mn017">
      <VocalSound Description="clears throat"/>
    </Segment>
    <Segment StartTime="1000.583" EndTime="1002.850" Participant="fe016">
      No, it's not bad. That's good.
    </Segment>
    <Segment StartTime="1003.493" EndTime="1004.320" Participant="mn014">
      not - not
    </Segment>
    <Segment StartTime="1004.456" EndTime="1005.407" Participant="mn014">
      totally sure, but
    </Segment>
    <Segment StartTime="1005.233" EndTime="1006.057" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="1005.407" EndTime="1010.431" Participant="mn014">
      we can be somehow sure that there is nothing - not - no speech between those. So, it - it -
    </Segment>
    <Segment StartTime="1008.630" EndTime="1010.739" Participant="fe016">
      i- ye- Yeah. th- u-
    </Segment>
    <Segment StartTime="1008.665" EndTime="1009.328" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="1010.059" EndTime="1012.260" Participant="me018">
      What does the two-second threshold mean?
    </Segment>
    <Segment StartTime="1010.431" EndTime="1015.870" Participant="mn014">
      I think i- it doesn't - It's the same as in the - the smoother for the -
    </Segment>
    <Segment StartTime="1011.283" EndTime="1013.009" Participant="fe016">
      I th- I think that's g- that's good.
    </Segment>
    <Segment StartTime="1015.870" EndTime="1016.230" Participant="mn014">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="1016.218" EndTime="1018.389" Participant="fe016">
      <VocalSound Description="inbreath"/> <VocalSound Description="outbreath"/> <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1016.230" EndTime="1017.260" Participant="mn014">
      I_B_M thing.
    </Segment>
    <Segment StartTime="1016.654" EndTime="1019.750" Participant="me011">
      <VocalSound Description="mouth"/> It combines them if it's - if the pause is longer than -
    </Segment>
    <Segment StartTime="1017.923" EndTime="1018.494" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="1018.389" EndTime="1026.080" Participant="fe016">
      It's - it's no more than six words. So - Roughly. On average. That's pretty good, I think. It's better than a -
    </Segment>
    <Segment StartTime="1019.030" EndTime="1019.769" Participant="me018">
      Ah.
    </Segment>
    <Segment StartTime="1019.337" EndTime="1020.264" Participant="mn014">
      So. Yeah.
    </Segment>
    <Segment StartTime="1022.029" EndTime="1023.707" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="1023.132" EndTime="1023.682" Participant="mn014">
      So.
    </Segment>
    <Segment StartTime="1024.560" EndTime="1029.670" Participant="me011">
      Right. So the - the trade-off is you get longer utterances, but you miss fewer utterances.
    </Segment>
    <Segment StartTime="1027.210" EndTime="1027.741" Participant="mn014">
      But -
    </Segment>
    <Segment StartTime="1028.466" EndTime="1031.078" Participant="mn014">
      Yeah. But - but the - the chunks are already sh-
    </Segment>
    <Segment StartTime="1031.078" EndTime="1038.584" Participant="mn014">
      i- in general, are short. So <Pause/> I - I t- I think it would be better to have - to have more of them <Pronounce Pronunciation="&quot;concenated&quot;"> concatenated  </Pronounce> together, in order to have
    </Segment>
    <Segment StartTime="1036.040" EndTime="1036.780" Participant="mn017">
      Mmm.
    </Segment>
    <Segment StartTime="1038.584" EndTime="1040.960" Participant="mn014">
      better language model or language modeling.
    </Segment>
    <Segment StartTime="1041.270" EndTime="1043.160" Participant="mn017">
      I think two seconds - mmm <Pause/>
    </Segment>
    <Segment StartTime="1041.789" EndTime="1042.388" Participant="mn014">
      I don't know.
    </Segment>
    <Segment StartTime="1043.160" EndTime="1046.283" Participant="mn017">
      I would maybe go with <Emphasis> one </Emphasis> second. I don't know, it's a - <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1044.652" EndTime="1045.440" Participant="me011">
      Well, take a look -
    </Segment>
    <Segment StartTime="1044.694" EndTime="1045.073" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="1045.440" EndTime="1046.456" Participant="mn014">
      Do that. Yeah.
    </Segment>
    <Segment StartTime="1046.283" EndTime="1048.883" Participant="mn017">
      See what the length distribution is. Yeah.
    </Segment>
    <Segment StartTime="1047.160" EndTime="1054.062" Participant="mn014">
      Yeah. But - but y- there's - there's really not much difference between the one-second and the two-second. So just <Pause/> take the one - the one-second one.
    </Segment>
    <Segment StartTime="1050.728" EndTime="1051.392" Participant="mn017">
      Really?
    </Segment>
    <Segment StartTime="1051.161" EndTime="1052.332" Participant="me018">
      <VocalSound Description="inbreath"/> I wouldn't think that
    </Segment>
    <Segment StartTime="1052.410" EndTime="1058.190" Participant="mn017">
      <VocalSound Description="inbreath"/> Uh - bu- I'm - I'm - I'm just scared that with two seconds you get - you get, um -
    </Segment>
    <Segment StartTime="1052.706" EndTime="1053.882" Participant="me018">
      the language model would
    </Segment>
    <Segment StartTime="1053.772" EndTime="1054.770" Participant="fe016">
      Yeah.
    </Segment>
    <Segment StartTime="1053.882" EndTime="1055.970" Participant="me018">
      continue across two seconds.
    </Segment>
    <Segment StartTime="1055.953" EndTime="1057.810" Participant="fe016">
      Well, yeah. Y- you <Emphasis> do, </Emphasis> becau-
    </Segment>
    <Segment StartTime="1058.190" EndTime="1060.847" Participant="mn017">
      <VocalSound Description="mouth"/> you - you get false recognitions. You're gonna -
    </Segment>
    <Segment StartTime="1060.847" EndTime="1066.449" Participant="mn017">
      Yeah, you're gonna hurt yourself occasionally by having - missing the language model context. But you might
    </Segment>
    <Segment StartTime="1066.449" EndTime="1068.907" Participant="mn017">
      hurt yourself more by having
    </Segment>
    <Segment StartTime="1068.907" EndTime="1074.810" Participant="mn017">
      misrecognitions due to <Pause/> background speech, or, uh, y- noise, or whatever.
    </Segment>
    <Segment StartTime="1073.942" EndTime="1076.321" Participant="mn014">
      I - I'm not too afraid ab- about that as
    </Segment>
    <Segment StartTime="1076.321" EndTime="1082.400" Participant="mn014">
      w- when there - when there would be something - some background speech or something - there wo- there would be a - a chunk in another
    </Segment>
    <Segment StartTime="1082.063" EndTime="1084.491" Participant="mn017">
      <VocalSound Description="mouth"/> Oh, I see. Then - Oh, I see. O_K.
    </Segment>
    <Segment StartTime="1082.400" EndTime="1082.788" Participant="mn014">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1082.671" EndTime="1084.730" Participant="fe016">
      Yeah. I think it's better.
    </Segment>
    <Segment StartTime="1082.788" EndTime="1089.668" Participant="mn014">
      channel, and when there is s- something in between, I con- I - I do not <Pronounce Pronunciation="&quot;concenate&quot;"> concatenate  </Pronounce> them. It's just when there is - when they are sequentially and -
    </Segment>
    <Segment StartTime="1085.910" EndTime="1087.410" Participant="mn017">
      Oh, right. Oh, that's -
    </Segment>
    <Segment StartTime="1085.940" EndTime="1088.430" Participant="fe016">
      The longer is better. There's - there's - th-
    </Segment>
    <Segment StartTime="1089.150" EndTime="1091.202" Participant="mn017">
      Mm-hmm. O_K. Sure.
    </Segment>
    <Segment StartTime="1090.170" EndTime="1092.104" Participant="mn014">
      So. I wou- I would use th-
    </Segment>
    <Segment StartTime="1090.930" EndTime="1091.536" Participant="fe016">
      Beca-
    </Segment>
    <Segment StartTime="1091.690" EndTime="1094.650" Participant="mn017">
      We can try them all and see which works better. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1092.712" EndTime="1093.128" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="1092.770" EndTime="1095.427" Participant="fe016">
      There's a l- there's a lot of these cases just like now, where I c-
    </Segment>
    <Segment StartTime="1092.880" EndTime="1093.429" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1095.427" EndTime="1101.597" Participant="fe016">
      people say "uh - uh" when they're trying to ta- and there's about a half-second pause to a second in between and then another word,
    </Segment>
    <Segment StartTime="1096.593" EndTime="1097.760" Participant="mn017">
      Mm-hmm.
    </Segment>
    <Segment StartTime="1099.790" EndTime="1100.263" Participant="mn014">
      Yep.
    </Segment>
    <Segment StartTime="1101.145" EndTime="1101.886" Participant="mn017">
      <VocalSound Description="mouth"/> O_K.
    </Segment>
    <Segment StartTime="1101.597" EndTime="1104.030" Participant="fe016">
      and it's much better if we can keep those together, I think.
    </Segment>
    <Segment StartTime="1104.240" EndTime="1108.559" Participant="me011">
      Yeah. It's, uh, funny looking at some of the transcripts. I was filtering by <Pause/> person,
    </Segment>
    <Segment StartTime="1107.785" EndTime="1108.487" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1108.897" EndTime="1109.536" Participant="mn017">
      Mm-hmm.
    </Segment>
    <Segment StartTime="1108.937" EndTime="1112.143" Participant="me011">
      and in one of the - one of the early meetings, one pers- particular person,
    </Segment>
    <Segment StartTime="1112.738" EndTime="1118.020" Participant="me011">
      <Emphasis> almost </Emphasis> the only thing they said the entire meeting was "yeah", "uh-huh". <Comment Description="while laughing"/> It was just a <Emphasis> whole </Emphasis> list of them.
    </Segment>
    <Segment StartTime="1115.830" EndTime="1117.785" Participant="fe016">
      <VocalSound Description="breath-laugh"/> <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="1117.861" EndTime="1119.990" Participant="fe008">
      <VocalSound Description="mouth"/> I bet I know who that was. Yeah.
    </Segment>
    <Segment StartTime="1117.913" EndTime="1124.600" Participant="mn017">
      <VocalSound Description="mouth"/> O_K. So - so we need to split the waveforms, then. Or do you already have them split up? No, you don't. Right?
    </Segment>
    <Segment StartTime="1118.020" EndTime="1119.310" Participant="me011">
      It was very funny.
    </Segment>
    <Segment StartTime="1124.461" EndTime="1125.340" Participant="mn014">
      No.
    </Segment>
    <Segment StartTime="1124.600" EndTime="1128.740" Participant="mn017">
      So - so, I guess Don <Pause/> would need your help to -
    </Segment>
    <Segment StartTime="1128.914" EndTime="1129.903" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1129.020" EndTime="1132.340" Participant="mn017">
      to create a new set of split, uh, meetings.
    </Segment>
    <Segment StartTime="1131.818" EndTime="1132.500" Participant="me001">
      Sure.
    </Segment>
    <Segment StartTime="1132.343" EndTime="1138.316" Participant="fe016">
      You know, you just <Emphasis> fake </Emphasis> the format that you take as input with the synch times to a new set of synch times, and -
    </Segment>
    <Segment StartTime="1132.975" EndTime="1134.160" Participant="mn017">
      If -
    </Segment>
    <Segment StartTime="1133.292" EndTime="1134.042" Participant="mn014">
      We could -
    </Segment>
    <Segment StartTime="1135.366" EndTime="1136.230" Participant="mn017">
      Right.
    </Segment>
    <Segment StartTime="1135.836" EndTime="1136.274" Participant="mn014">
      Mmm.
    </Segment>
    <Segment StartTime="1137.750" EndTime="1139.740" Participant="me011">
      Uh, do we know about disk?
    </Segment>
    <Segment StartTime="1138.143" EndTime="1142.457" Participant="mn017">
      But th- <VocalSound Description="laugh"/> Oh, yeah. There's <Emphasis> that </Emphasis> pressure. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1140.485" EndTime="1141.304" Participant="me001">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1141.260" EndTime="1142.568" Participant="me011">
      Uh, Abbott disk?
    </Segment>
    <Segment StartTime="1143.050" EndTime="1144.100" Participant="me013">
      Uh. I know they're in.
    </Segment>
    <Segment StartTime="1144.203" EndTime="1144.731" Participant="me011">
      O_K.
    </Segment>
    <Segment StartTime="1144.436" EndTime="1149.040" Participant="me013">
      <VocalSound Description="inbreath"/> And, uh - But I don't know. I think he was - Wasn't he asking about - ?
    </Segment>
    <Segment StartTime="1146.229" EndTime="1146.916" Participant="me018">
      Wel-
    </Segment>
    <Segment StartTime="1147.960" EndTime="1149.236" Participant="me018">
      He had a problem. Right?
    </Segment>
    <Segment StartTime="1148.725" EndTime="1149.222" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="1148.877" EndTime="1149.477" Participant="mn014">
      Yep.
    </Segment>
    <Segment StartTime="1149.537" EndTime="1150.119" Participant="fe016">
      Oh.
    </Segment>
    <Segment StartTime="1150.445" EndTime="1153.810" Participant="me013">
      Well, there was an issue. He wanted to take it down. And then he tried -
    </Segment>
    <Segment StartTime="1152.570" EndTime="1155.160" Participant="me018">
      H- he did, and then it didn't work, and <Pause/>
    </Segment>
    <Segment StartTime="1154.539" EndTime="1155.892" Participant="me011">
      Couldn't format <Pronounce Pronunciation="&quot;em&quot;"> them.  </Pronounce> No.
    </Segment>
    <Segment StartTime="1155.684" EndTime="1157.148" Participant="me018">
      I didn't hear anything after that.
    </Segment>
    <Segment StartTime="1155.892" EndTime="1158.378" Participant="me011">
      <VocalSound Description="inbreath"/> The only reason I'm asking is, you're gonna need space to
    </Segment>
    <Segment StartTime="1158.378" EndTime="1161.745" Participant="me011">
      split them up. And so I wanted to make sure we had some available for you.
    </Segment>
    <Segment StartTime="1160.104" EndTime="1164.057" Participant="me001">
      <VocalSound Description="clears throat"/> I still have, like, <VocalSound Description="outbreath"/> probably <Pause/>
    </Segment>
    <Segment StartTime="1162.496" EndTime="1162.956" Participant="mn014">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1164.668" EndTime="1168.290" Participant="me001">
      six, seven, eight gig on my disk.
    </Segment>
    <Segment StartTime="1165.668" EndTime="1166.816" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1166.466" EndTime="1167.930" Participant="me013">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="1166.950" EndTime="1170.330" Participant="me011">
      O_K. So we're O_K - we're <Emphasis> still </Emphasis> O_K for another couple days, then?
    </Segment>
    <Segment StartTime="1167.470" EndTime="1175.140" Participant="fe016">
      And I have - I have like another un-backed - I ha- have another six gig, which Jeremy, if you're not using, can - on X_A.
    </Segment>
    <Segment StartTime="1171.318" EndTime="1172.318" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1171.320" EndTime="1172.130" Participant="me001">
      Yeah.
    </Segment>
    <Segment StartTime="1172.698" EndTime="1173.118" Participant="me013">
      Wel-
    </Segment>
    <Segment StartTime="1173.533" EndTime="1175.471" Participant="me011">
      So we're O_K for - for a couple weeks, then.
    </Segment>
    <Segment StartTime="1173.680" EndTime="1174.805" Participant="me013">
      Markham probably needs u-
    </Segment>
    <Segment StartTime="1175.142" EndTime="1179.940" Participant="me013">
      H- he probably needs us to approve another time to take things down. Right? In order to do that?
    </Segment>
    <Segment StartTime="1176.086" EndTime="1177.362" Participant="fe016">
      <Uncertain> To you. </Uncertain>
    </Segment>
    <Segment StartTime="1178.622" EndTime="1181.698" Participant="me018">
      <VocalSound Description="inbreath"/> Yeah. He - he didn't say - he hasn't said anything to me about it.
    </Segment>
    <Segment StartTime="1181.550" EndTime="1186.269" Participant="me013">
      I thought he - I thought he said in that mail that he would need to take it down another time. So I think he -
    </Segment>
    <Segment StartTime="1185.762" EndTime="1186.534" Participant="me018">
      Yeah. I think -
    </Segment>
    <Segment StartTime="1185.990" EndTime="1187.740" Participant="me011">
      Y- yeah. He just didn't say when. So -
    </Segment>
    <Segment StartTime="1187.370" EndTime="1188.243" Participant="me018">
      Yeah. He -
    </Segment>
    <Segment StartTime="1187.502" EndTime="1189.817" Participant="me013">
      Well, no. I think he wanted u- us to tell him
    </Segment>
    <Segment StartTime="1189.049" EndTime="1190.970" Participant="fe016">
      How about during the picnic?
    </Segment>
    <Segment StartTime="1189.600" EndTime="1190.252" Participant="me011">
      Ah.
    </Segment>
    <Segment StartTime="1190.331" EndTime="1191.011" Participant="me013">
      when.
    </Segment>
    <Segment StartTime="1190.970" EndTime="1193.897" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1191.597" EndTime="1195.493" Participant="me013">
      <VocalSound Description="laugh"/> Yeah. I'm sure he'd love to - <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1191.798" EndTime="1193.628" Participant="me011">
      Yeah. I'm sure he'd appreciate that. <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="1191.829" EndTime="1197.371" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1191.909" EndTime="1194.288" Participant="me018">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1193.897" EndTime="1199.353" Participant="fe016">
      <VocalSound Description="laugh"/> Sorry. <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1194.213" EndTime="1196.445" Participant="me011">
      Well, m- <Emphasis> my </Emphasis> feeling about that
    </Segment>
    <Segment StartTime="1194.480" EndTime="1195.670" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1194.690" EndTime="1196.149" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1195.493" EndTime="1199.640" Participant="me013">
      Fortunately, Markham's not a transcriber. But, um - <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1197.108" EndTime="1197.955" Participant="me011">
      is p- is p-
    </Segment>
    <Segment StartTime="1197.921" EndTime="1198.890" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1199.353" EndTime="1203.033" Participant="fe016">
      Beep. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1199.509" EndTime="1200.778" Participant="me011">
      Well, O_K. That's the point.
    </Segment>
    <Segment StartTime="1199.582" EndTime="1200.820" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1200.418" EndTime="1202.171" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1201.267" EndTime="1203.385" Participant="me011">
      So it's <Emphasis> Jane </Emphasis> <Pause/> that we have to coordinate that through.
    </Segment>
    <Segment StartTime="1203.385" EndTime="1208.789" Participant="me011">
      Uh, what <Emphasis> I </Emphasis> was gonna say is, as soon as possible, and I'm willing to not work for an hour to get it done.
    </Segment>
    <Segment StartTime="1204.415" EndTime="1205.312" Participant="me013">
      <VocalSound Description="sniff"/>
    </Segment>
    <Segment StartTime="1209.530" EndTime="1210.267" Participant="me011">
      Uh, but -
    </Segment>
    <Segment StartTime="1209.824" EndTime="1211.011" Participant="me013">
      "I <Emphasis> might </Emphasis> not work." <Comment Description="whispered"/> <Comment Description="quoting [me011]"/>
    </Segment>
    <Segment StartTime="1209.893" EndTime="1213.545" Participant="mn017">
      <VocalSound Description="laugh"/> <NonVocalSound Description="noise"/> <Comment Description="clinking in background"/> Oops. <Comment Description="interjection"/> I'm sorry.
    </Segment>
    <Segment StartTime="1210.865" EndTime="1213.330" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1210.915" EndTime="1213.310" Participant="fe008">
      <NonVocalSound Description="clinking in background"/>
    </Segment>
    <Segment StartTime="1211.011" EndTime="1212.140" Participant="me013">
      <NonVocalSound Description="sound of clinking in background"/>
    </Segment>
    <Segment StartTime="1212.140" EndTime="1213.543" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1212.294" EndTime="1213.465" Participant="mn014">
      Oops! <Comment Description="interjection"/>
    </Segment>
    <Segment StartTime="1212.425" EndTime="1213.452" Participant="me001">
      Whoa! <Comment Description="interjection"/>
    </Segment>
    <Segment StartTime="1213.543" EndTime="1217.815" Participant="me013">
      "I'm - I'm willing to <Emphasis> not </Emphasis> work for an hour." <Comment Description="quoting [me011]"/> I <Emphasis> know </Emphasis> you're willing to not work for an hour.
    </Segment>
    <Segment StartTime="1213.660" EndTime="1215.170" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1213.819" EndTime="1224.305" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1213.916" EndTime="1214.730" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1215.926" EndTime="1219.991" Participant="me011">
      Because when Abbott's - Yeah, right. Because when Abbott is down you can't work. <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="1216.962" EndTime="1220.733" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1217.241" EndTime="1219.481" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1217.328" EndTime="1220.197" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1217.815" EndTime="1224.320" Participant="me013">
      But, I th- <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1217.957" EndTime="1219.120" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1219.481" EndTime="1223.019" Participant="fe016">
      You're really <Emphasis> dedicated, </Emphasis> if you're - No matter how you parse that one.
    </Segment>
    <Segment StartTime="1220.733" EndTime="1226.576" Participant="me001">
      Yea- <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1220.812" EndTime="1222.570" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1220.857" EndTime="1223.590" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1222.768" EndTime="1225.346" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1222.770" EndTime="1223.820" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1223.019" EndTime="1227.382" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1223.590" EndTime="1224.317" Participant="me011">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1224.305" EndTime="1225.011" Participant="mn017">
      <Uncertain> O_K. </Uncertain>
    </Segment>
    <Segment StartTime="1224.317" EndTime="1226.406" Participant="me011">
      But, uh, I think the per- the people it
    </Segment>
    <Segment StartTime="1225.220" EndTime="1228.811" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1225.346" EndTime="1229.878" Participant="mn014">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="1226.782" EndTime="1229.160" Participant="me011">
      disrupts the most are the transcribers.
    </Segment>
    <Segment StartTime="1228.740" EndTime="1231.504" Participant="fe008">
      Well, you know, uh, I - All I need to do is mail, um -
    </Segment>
    <Segment StartTime="1231.504" EndTime="1235.566" Participant="fe008">
      send them m- a mail like two days in advance so they can schedule their time.
    </Segment>
    <Segment StartTime="1233.196" EndTime="1234.630" Participant="me001">
      <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="1235.566" EndTime="1239.249" Participant="fe008">
      I did that with the last outage. I j- I wrote to <Pronounce> them  </Pronounce> letting them know that this w-
    </Segment>
    <Segment StartTime="1235.670" EndTime="1236.400" Participant="me011">
      O_K.
    </Segment>
    <Segment StartTime="1235.731" EndTime="1237.365" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="1236.371" EndTime="1236.953" Participant="me013">
      I-
    </Segment>
    <Segment StartTime="1239.620" EndTime="1245.140" Participant="me013">
      So, <Pronounce> O_K,  </Pronounce> it sounds like <Emphasis> Markham </Emphasis> should almost decide when he wants to do it, and tell us, as long as -
    </Segment>
    <Segment StartTime="1240.119" EndTime="1241.356" Participant="fe008">
      was not, um <Pause/>
    </Segment>
    <Segment StartTime="1240.209" EndTime="1241.830" Participant="me011">
      So early next week.
    </Segment>
    <Segment StartTime="1243.373" EndTime="1245.410" Participant="me011">
      And just as long as we have a little warning.
    </Segment>
    <Segment StartTime="1245.140" EndTime="1245.870" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="1245.288" EndTime="1246.548" Participant="fe016">
      So that means we can't,
    </Segment>
    <Segment StartTime="1245.510" EndTime="1246.280" Participant="fe008">
      Yeah.
    </Segment>
    <Segment StartTime="1246.920" EndTime="1249.236" Participant="fe016">
      um, save meeting data either. Right?
    </Segment>
    <Segment StartTime="1250.010" EndTime="1253.760" Participant="me011">
      Uh, just not during that time when it's down. But that - it should be down for an hour.
    </Segment>
    <Segment StartTime="1253.210" EndTime="1258.393" Participant="fe016">
      So it just can't - so we can't have two meetings in a row where the first meeting's during that hour. That's all I meant.
    </Segment>
    <Segment StartTime="1257.340" EndTime="1257.838" Participant="me011">
      Right.
    </Segment>
    <Segment StartTime="1258.596" EndTime="1260.317" Participant="me011">
      Well, no, we can store <Pronounce Pronunciation="&quot;em&quot;"> them  </Pronounce> here.
    </Segment>
    <Segment StartTime="1258.946" EndTime="1261.080" Participant="fe016">
      That - that's happening, like, today.
    </Segment>
    <Segment StartTime="1260.640" EndTime="1261.750" Participant="me011">
      We can store <Pronounce Pronunciation="&quot;em&quot;"> them  </Pronounce> here.
    </Segment>
    <Segment StartTime="1262.006" EndTime="1265.687" Participant="me011">
      You - we just run the risk <Pause/> that if you have a crash we lose the data. So.
    </Segment>
    <Segment StartTime="1262.010" EndTime="1263.060" Participant="fe016">
      Temporarily.
    </Segment>
    <Segment StartTime="1265.600" EndTime="1266.698" Participant="me018">
      They're stored on Popcorn.
    </Segment>
    <Segment StartTime="1265.617" EndTime="1268.230" Participant="fe016">
      Well, no. I mea- I mean, the fir- Oh, jus- O_K.
    </Segment>
    <Segment StartTime="1265.710" EndTime="1266.848" Participant="fe008">
      This is on Popcorn? Or - ?
    </Segment>
    <Segment StartTime="1266.241" EndTime="1268.939" Participant="mn014">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="1267.796" EndTime="1268.287" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="1268.441" EndTime="1270.930" Participant="fe008">
      This is on Popcorn or something? Yeah? O_K.
    </Segment>
    <Segment StartTime="1269.653" EndTime="1270.278" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="1271.032" EndTime="1274.474" Participant="me013">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1275.473" EndTime="1277.912" Participant="me013">
      We store - we store our data on Popcorn.
    </Segment>
    <Segment StartTime="1275.710" EndTime="1276.498" Participant="me011">
      Anything else?
    </Segment>
    <Segment StartTime="1277.912" EndTime="1278.901" Participant="me013">
      That's really great.
    </Segment>
    <Segment StartTime="1278.901" EndTime="1286.399" Participant="me013">
      <VocalSound Description="laugh"/> I was just thinking we store our data on Popcorn. How many - how many institutes can you <VocalSound Description="laugh"/> say do that? O_K.
    </Segment>
    <Segment StartTime="1279.058" EndTime="1282.030" Participant="mn017">
      <VocalSound Description="inbreath"/> <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1279.153" EndTime="1286.629" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1279.161" EndTime="1279.933" Participant="me011">
      Excuse me?
    </Segment>
    <Segment StartTime="1279.194" EndTime="1280.930" Participant="mn014">
      <VocalSound Description="laugh-snort"/>
    </Segment>
    <Segment StartTime="1279.369" EndTime="1280.650" Participant="me018">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1279.404" EndTime="1280.294" Participant="me051">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1281.719" EndTime="1284.970" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1281.770" EndTime="1285.470" Participant="me018">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1281.961" EndTime="1284.746" Participant="me011">
      Yep. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1282.030" EndTime="1286.000" Participant="mn017">
      Pop goes the data. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1282.173" EndTime="1283.360" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1283.626" EndTime="1284.860" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1286.399" EndTime="1286.858" Participant="me013">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1286.858" EndTime="1289.703" Participant="me013">
      Uh, g- megabytes and mega- many megabytes, too.
    </Segment>
    <Segment StartTime="1289.703" EndTime="1290.662" Participant="me013">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1289.946" EndTime="1293.146" Participant="me018">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1290.662" EndTime="1293.060" Participant="me013">
      Um, what - what, uh - ?
    </Segment>
    <Segment StartTime="1292.625" EndTime="1294.163" Participant="me011">
      We have a <Emphasis> kernel </Emphasis> on Popcorn, too. <Comment Description="humor, computer science pun"/>
    </Segment>
    <Segment StartTime="1293.799" EndTime="1301.991" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1293.910" EndTime="1299.802" Participant="me018">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1294.046" EndTime="1298.002" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1294.163" EndTime="1297.783" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1294.210" EndTime="1298.433" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1294.460" EndTime="1295.900" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1294.719" EndTime="1296.040" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1294.866" EndTime="1296.027" Participant="mn017">
      Right. <Comment Description="while laughing"/> <Comment Description="with mock groan intonation"/>
    </Segment>
    <Segment StartTime="1296.027" EndTime="1301.095" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1296.858" EndTime="1298.324" Participant="me001">
      <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="1298.002" EndTime="1300.491" Participant="me013">
      <VocalSound Description="outbreath"/> <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1298.210" EndTime="1302.710" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1298.324" EndTime="1302.204" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1299.056" EndTime="1304.405" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1299.580" EndTime="1302.881" Participant="fe008">
      That's very good. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1300.491" EndTime="1306.979" Participant="me013">
      So uh, what - what's on your <Emphasis> queue </Emphasis> for - for recognition experiments? Let's talk about <Emphasis> that </Emphasis> for a second, maybe.
    </Segment>
    <Segment StartTime="1300.500" EndTime="1302.300" Participant="me018">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1301.095" EndTime="1302.026" Participant="mn017">
      u- un-
    </Segment>
    <Segment StartTime="1302.026" EndTime="1305.750" Participant="mn017">
      Can I have butter on my meeting? <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1303.295" EndTime="1304.369" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1304.969" EndTime="1305.364" Participant="me018">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="1306.280" EndTime="1307.520" Participant="fe008">
      What was the question?
    </Segment>
    <Segment StartTime="1306.297" EndTime="1306.828" Participant="me018">
      <VocalSound Description="mouth"/> Uh -
    </Segment>
    <Segment StartTime="1306.979" EndTime="1310.070" Participant="me013">
      What - what was on his queue for recognition experiments?
    </Segment>
    <Segment StartTime="1308.872" EndTime="1309.740" Participant="fe008">
      O_K. <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="1309.561" EndTime="1310.695" Participant="me018">
      <VocalSound Description="mouth"/> Um,
    </Segment>
    <Segment StartTime="1310.695" EndTime="1313.975" Participant="me018">
      I'm rebuilding the net that we're gonna use for the tandem
    </Segment>
    <Segment StartTime="1314.560" EndTime="1316.530" Participant="me018">
      stuff. And so what I'm doing is, <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1316.790" EndTime="1318.429" Participant="me018">
      um, putting in
    </Segment>
    <Segment StartTime="1317.411" EndTime="1317.970" Participant="me011">
      Mmm. <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="1318.765" EndTime="1319.993" Participant="me018">
      the stream
    </Segment>
    <Segment StartTime="1320.389" EndTime="1324.345" Participant="me018">
      reader into the Quicknet libraries for the S_R_I feature files.
    </Segment>
    <Segment StartTime="1322.508" EndTime="1323.888" Participant="me013">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="1323.799" EndTime="1325.762" Participant="fe008">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="1324.345" EndTime="1324.752" Participant="me018">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1324.752" EndTime="1329.360" Participant="me018">
      An- Which is the right way to do it. I mean, when we did our first experiments and I was, <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1329.360" EndTime="1335.452" Participant="me018">
      uh, creating S_R_I feature files from the ICSI front-end, I just had perl scripts, you know, and <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1335.920" EndTime="1340.667" Participant="me018">
      hacked a bunch of stuff together just to get it going. But the r- the <Emphasis> right </Emphasis> way to do it is to
    </Segment>
    <Segment StartTime="1339.630" EndTime="1367.881" Participant="fe016">
      <NonVocalSound Description="mike noise"/> <Comment Description="readjusting mike"/>
    </Segment>
    <Segment StartTime="1340.667" EndTime="1343.376" Participant="me018">
      integrate it in with the ICSI tools. And so, <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1343.376" EndTime="1346.481" Participant="me018">
      that's what I'm doing now. And so once I get that done, then I'll generate
    </Segment>
    <Segment StartTime="1346.481" EndTime="1349.996" Participant="me018">
      the P_files I need. <VocalSound Description="inbreath"/> Cuz we already have the <Emphasis> feature </Emphasis> files
    </Segment>
    <Segment StartTime="1350.378" EndTime="1351.271" Participant="me013">
      Mm-hmm.
    </Segment>
    <Segment StartTime="1350.770" EndTime="1353.912" Participant="me018">
      in the S_R_I format. So what I need to do is, make it so that the -
    </Segment>
    <Segment StartTime="1354.558" EndTime="1357.242" Participant="me018">
      the Quicknet stuff can read those. And, uh -
    </Segment>
    <Segment StartTime="1356.840" EndTime="1365.257" Participant="mn017">
      I- is that <Emphasis> independent </Emphasis> or related to <Pause/> also being able to write out the, uh, <Emphasis> feature </Emphasis> file <Pause/> i- in the S_R_I format f- r-
    </Segment>
    <Segment StartTime="1359.363" EndTime="1362.399" Participant="me018">
      <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="1364.330" EndTime="1366.795" Participant="me018">
      It's both. There's a - there's an input stream and an output stream.
    </Segment>
    <Segment StartTime="1365.164" EndTime="1367.039" Participant="me011">
      Input reader and an output stream. Yeah.
    </Segment>
    <Segment StartTime="1365.257" EndTime="1366.112" Participant="mn017">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1366.112" EndTime="1368.903" Participant="mn017">
      Oh, O_K. So then you could use, um -
    </Segment>
    <Segment StartTime="1367.052" EndTime="1367.479" Participant="me018">
      Yeah.
    </Segment>
    <Segment StartTime="1369.729" EndTime="1371.174" Participant="mn017">
      You could use, um -
    </Segment>
    <Segment StartTime="1371.174" EndTime="1374.621" Participant="mn017">
      <VocalSound Description="mouth smack"/> <Comment Description="vocal gesture of pondering"/> uh, like Feacalc and s- just specify
    </Segment>
    <Segment StartTime="1375.058" EndTime="1376.266" Participant="mn017">
      as an output format
    </Segment>
    <Segment StartTime="1376.379" EndTime="1377.370" Participant="me011">
      Yeah. That's the point.
    </Segment>
    <Segment StartTime="1376.730" EndTime="1377.834" Participant="mn017">
      the - the - Oh, O_K.
    </Segment>
    <Segment StartTime="1377.834" EndTime="1380.192" Participant="mn017">
      Th- t- I'm just ignorant about the <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1378.251" EndTime="1378.725" Participant="me018">
      Yeah.
    </Segment>
    <Segment StartTime="1379.612" EndTime="1380.740" Participant="me018">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1379.740" EndTime="1381.116" Participant="me011">
      Yeah. So, if - if you -
    </Segment>
    <Segment StartTime="1380.192" EndTime="1383.937" Participant="mn017">
      sof- software architecture of this thing. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1381.475" EndTime="1389.054" Participant="me011">
      Right. Quicknet is a very nice stream-based library, so without too much effort, once he has the classes written we can incorporate it into all the standard tools.
    </Segment>
    <Segment StartTime="1383.937" EndTime="1385.010" Participant="mn017">
      Uh-huh.
    </Segment>
    <Segment StartTime="1388.833" EndTime="1389.920" Participant="mn017">
      Oh, cool.
    </Segment>
    <Segment StartTime="1390.185" EndTime="1390.900" Participant="me011">
      So.
    </Segment>
    <Segment StartTime="1391.444" EndTime="1391.981" Participant="mn017">
      Great.
    </Segment>
    <Segment StartTime="1392.748" EndTime="1393.065" Participant="me018">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="1393.570" EndTime="1396.420" Participant="me018">
      So, then it's t- uh, tandem experiments after that?
    </Segment>
    <Segment StartTime="1394.249" EndTime="1398.695" Participant="mn017">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1394.498" EndTime="1395.006" Participant="me013">
      Uh-huh.
    </Segment>
    <Segment StartTime="1395.423" EndTime="1398.568" Participant="me011">
      <VocalSound Description="mouth"/> And at some point, I'd like to get back to, uh,
    </Segment>
    <Segment StartTime="1395.626" EndTime="1395.955" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="1398.805" EndTime="1402.257" Participant="me011">
      porting Quicknet to the multiprocessor Linux box.
    </Segment>
    <Segment StartTime="1403.205" EndTime="1406.214" Participant="me011">
      You know, I - I have forward passes working, but I haven't done training yet.
    </Segment>
    <Segment StartTime="1408.557" EndTime="1411.036" Participant="mn017">
      So, b- speaking of Linux. So - <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1411.036" EndTime="1412.790" Participant="mn017">
      Th- there's some
    </Segment>
    <Segment StartTime="1413.207" EndTime="1415.330" Participant="mn017">
      i- impetus at, um,
    </Segment>
    <Segment StartTime="1415.330" EndTime="1416.392" Participant="mn017">
      <VocalSound Description="mouth"/> S_R_I to actually
    </Segment>
    <Segment StartTime="1416.392" EndTime="1417.360" Participant="mn017">
      <VocalSound Description="breath"/> u- u- p- th- <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="1417.360" EndTime="1421.090" Participant="mn017">
      uh, build - support Linux as a platform. So.
    </Segment>
    <Segment StartTime="1421.090" EndTime="1426.726" Participant="mn017">
      <VocalSound Description="laugh"/> What that means is, <VocalSound Description="laugh"/> once we have, uh, everything running on Linux we can
    </Segment>
    <Segment StartTime="1421.730" EndTime="1422.223" Participant="me011">
      Mm-hmm.
    </Segment>
    <Segment StartTime="1422.759" EndTime="1423.616" Participant="me018">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1426.726" EndTime="1428.780" Participant="mn017">
      also use a Li- <VocalSound Description="inbreath"/> eh-
    </Segment>
    <Segment StartTime="1426.758" EndTime="1428.732" Participant="fe016">
      n- Run all our jobs on <Emphasis> your </Emphasis> machines. <Comment Description="while laughing"/> <Comment Description="completing [mn017]'s utterance"/>
    </Segment>
    <Segment StartTime="1428.732" EndTime="1441.450" Participant="fe016">
      <VocalSound Description="laugh"/> Sorry. <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1428.735" EndTime="1432.872" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1428.780" EndTime="1432.410" Participant="mn017">
      Yeah. Exac- <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1429.372" EndTime="1431.324" Participant="me018">
      <VocalSound Description="mouth"/> We don't have too many. We just have that -
    </Segment>
    <Segment StartTime="1429.640" EndTime="1430.288" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1431.671" EndTime="1433.700" Participant="me018">
      just have a few Linux machin- <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1432.410" EndTime="1440.740" Participant="mn017">
      I mean, if you can't use all the processors on whatever machine, we'll help you with that. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1434.116" EndTime="1439.497" Participant="me018">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1435.015" EndTime="1442.281" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1435.567" EndTime="1436.338" Participant="me011">
      Yeah. That's right. <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="1436.338" EndTime="1438.238" Participant="me011">
      Well, that's the nice thing about it, is that -
    </Segment>
    <Segment StartTime="1436.580" EndTime="1439.333" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1437.655" EndTime="1439.180" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1438.238" EndTime="1441.148" Participant="me011">
      i- since it's coarse parallelism you don't have to do anything special.
    </Segment>
    <Segment StartTime="1438.399" EndTime="1439.289" Participant="me001">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1440.740" EndTime="1442.180" Participant="mn017">
      Right. Exactly.
    </Segment>
    <Segment StartTime="1440.933" EndTime="1442.110" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1441.753" EndTime="1444.539" Participant="me011">
      So. I mean that would be a fine use for b- for that machine.
    </Segment>
    <Segment StartTime="1443.271" EndTime="1443.668" Participant="mn017">
      Yeah.
    </Segment>
    <Segment StartTime="1443.668" EndTime="1445.357" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1445.357" EndTime="1447.140" Participant="mn017">
      So it's just, uh -
    </Segment>
    <Segment StartTime="1446.608" EndTime="1448.348" Participant="me013">
      Oh, I know what it was - <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="1446.760" EndTime="1448.011" Participant="me011">
      Five more processors.
    </Segment>
    <Segment StartTime="1448.267" EndTime="1449.492" Participant="mn017">
      Uh. Yeah.
    </Segment>
    <Segment StartTime="1448.900" EndTime="1449.750" Participant="me013">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1449.750" EndTime="1451.410" Participant="me013">
      Um. Yeah. U- um -
    </Segment>
    <Segment StartTime="1451.695" EndTime="1452.392" Participant="me013">
      Some -
    </Segment>
    <Segment StartTime="1451.830" EndTime="1459.480" Participant="mn017">
      Or if - uh, you know, in the future, if Linux machines become like way cheaper, than, <VocalSound Description="mouth"/> uh, you know, Solaris machines, then <Pause/>
    </Segment>
    <Segment StartTime="1460.180" EndTime="1463.990" Participant="mn017">
      you know, that wouldn't be a reason not to <Pause/> use Linux anymore. So.
    </Segment>
    <Segment StartTime="1463.181" EndTime="1463.751" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="1463.745" EndTime="1464.290" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="1463.990" EndTime="1465.930" Participant="mn017">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1464.785" EndTime="1465.461" Participant="me013">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1465.461" EndTime="1469.674" Participant="me013">
      Um, I think it would be ne- neat at some point <Pause/> in this <Pause/> to do,
    </Segment>
    <Segment StartTime="1470.202" EndTime="1472.030" Participant="me013">
      um, <Emphasis> a </Emphasis> recognition,
    </Segment>
    <Segment StartTime="1472.474" EndTime="1474.180" Participant="me013">
      uh, pass
    </Segment>
    <Segment StartTime="1474.180" EndTime="1476.012" Participant="me013">
      on one of the P_Z_M
    </Segment>
    <Segment StartTime="1476.309" EndTime="1478.600" Participant="me013">
      mikes for these s- same
    </Segment>
    <Segment StartTime="1478.681" EndTime="1480.670" Participant="fe016">
      <VocalSound Description="inbreath"/> Yeah.
    </Segment>
    <Segment StartTime="1478.730" EndTime="1479.510" Participant="mn017">
      For the meeting?
    </Segment>
    <Segment StartTime="1479.061" EndTime="1481.195" Participant="me013">
      meetings that you've been gi- n- gi- bi- I mean,
    </Segment>
    <Segment StartTime="1481.195" EndTime="1485.234" Participant="me013">
      it's gonna be terrible, but, you know, we - we just don't know <Emphasis> how </Emphasis> terrible.
    </Segment>
    <Segment StartTime="1481.811" EndTime="1482.523" Participant="mn017">
      Mm-hmm.
    </Segment>
    <Segment StartTime="1485.230" EndTime="1489.310" Participant="fe016">
      Ye- It's also an interesting problem to come up with the reference.
    </Segment>
    <Segment StartTime="1485.234" EndTime="1486.045" Participant="me013">
      And -
    </Segment>
    <Segment StartTime="1485.324" EndTime="1485.943" Participant="mn017">
      Mm-hmm.
    </Segment>
    <Segment StartTime="1490.373" EndTime="1491.680" Participant="fe016">
      <VocalSound Description="inbreath"/> So,
    </Segment>
    <Segment StartTime="1491.848" EndTime="1493.025" Participant="fe016">
      the reference file
    </Segment>
    <Segment StartTime="1493.239" EndTime="1496.931" Participant="fe016">
      for the relative <Emphasis> times </Emphasis> at which - So -
    </Segment>
    <Segment StartTime="1495.110" EndTime="1497.000" Participant="me011">
      Oh, yeah. That's really <Emphasis> hard. </Emphasis>
    </Segment>
    <Segment StartTime="1496.562" EndTime="1497.486" Participant="mn017">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1497.248" EndTime="1501.730" Participant="fe016">
      Well, i- it's an interesting question, because I was thinking "well you can force align
    </Segment>
    <Segment StartTime="1498.840" EndTime="1500.560" Participant="me011">
      It's not determined.
    </Segment>
    <Segment StartTime="1501.730" EndTime="1506.871" Participant="fe016">
      the transcriber transcripts, and then of course, you try to merge them in time.
    </Segment>
    <Segment StartTime="1504.080" EndTime="1505.647" Participant="me013">
      Mm-hmm. <Comment Description="very quietly"/>
    </Segment>
    <Segment StartTime="1506.871" EndTime="1508.820" Participant="fe016">
      But how do you <Emphasis> score? </Emphasis>"
    </Segment>
    <Segment StartTime="1508.820" EndTime="1512.748" Participant="me011">
      <VocalSound Description="inbreath"/> I think the first pass is <Pause/> throw out words which are overlapped.
    </Segment>
    <Segment StartTime="1509.236" EndTime="1510.170" Participant="mn017">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1510.010" EndTime="1512.230" Participant="fe016">
      I mean it's just an interesting problem.
    </Segment>
    <Segment StartTime="1513.320" EndTime="1514.899" Participant="me011">
      That would be a good first pass.
    </Segment>
    <Segment StartTime="1515.438" EndTime="1518.000" Participant="me011">
      Just ignore everything that has any overlap.
    </Segment>
    <Segment StartTime="1515.770" EndTime="1516.500" Participant="mn017">
      Mmm.
    </Segment>
    <Segment StartTime="1515.995" EndTime="1519.989" Participant="fe016">
      Right. I mean, there's a whole sort of dis- Right. We - Right.
    </Segment>
    <Segment StartTime="1516.107" EndTime="1521.590" Participant="me013">
      <VocalSound Description="inbreath"/> Yeah. Cuz you have a set of scores about that, so maybe then, that wouldn't be so bad.
    </Segment>
    <Segment StartTime="1519.989" EndTime="1527.351" Participant="fe016">
      But there's a whole interesting discussion, cuz of course the <Emphasis> alignments </Emphasis> are not perfect <Emphasis> either, </Emphasis> and so - um. In fact, we actually
    </Segment>
    <Segment StartTime="1523.385" EndTime="1524.158" Participant="me013">
      Oh. <Comment Description="whispered"/>
    </Segment>
    <Segment StartTime="1524.368" EndTime="1525.045" Participant="me011">
      Right.
    </Segment>
    <Segment StartTime="1526.043" EndTime="1526.651" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="1527.351" EndTime="1529.170" Participant="fe016">
      don't <Emphasis> have </Emphasis> a - a p-
    </Segment>
    <Segment StartTime="1529.228" EndTime="1532.367" Participant="me013">
      But - that'd be a hell of a lot better than what we do with just these. And -
    </Segment>
    <Segment StartTime="1529.270" EndTime="1531.190" Participant="fe016">
      I mean, it- it's worth trying. I -
    </Segment>
    <Segment StartTime="1532.010" EndTime="1533.213" Participant="fe016">
      Right. Right.
    </Segment>
    <Segment StartTime="1532.367" EndTime="1536.479" Participant="me013">
      and - and, again, if you rule out the overlap, you have some numbers for that, cuz that's yet another -
    </Segment>
    <Segment StartTime="1536.479" EndTime="1538.475" Participant="me013">
      But, I - I'm just concerned, of course, about that -
    </Segment>
    <Segment StartTime="1538.150" EndTime="1542.840" Participant="fe016">
      Oh, I see. You mean when only one person is <Pause/> talki- Yes. We could - we could try tha- Right.
    </Segment>
    <Segment StartTime="1539.904" EndTime="1540.322" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="1541.070" EndTime="1544.155" Participant="me013">
      Cuz you have scores for that for the other case. And -
    </Segment>
    <Segment StartTime="1542.887" EndTime="1543.797" Participant="mn017">
      Mm-hmm.
    </Segment>
    <Segment StartTime="1543.060" EndTime="1544.330" Participant="fe016">
      Right. Exactly.
    </Segment>
    <Segment StartTime="1544.531" EndTime="1546.510" Participant="fe016">
      Yeah. We - we should try -
    </Segment>
    <Segment StartTime="1544.993" EndTime="1546.064" Participant="me013">
      We just don't know how
    </Segment>
    <Segment StartTime="1546.650" EndTime="1550.444" Participant="me013">
      bad it will be. I mean, one of the things that Dave was noticing - we were talking this morning -
    </Segment>
    <Segment StartTime="1549.826" EndTime="1550.400" Participant="me001">
      Hmm.
    </Segment>
    <Segment StartTime="1551.072" EndTime="1554.238" Participant="me013">
      is that i- it <Emphasis> seems </Emphasis> like - and we do- don't know
    </Segment>
    <Segment StartTime="1554.238" EndTime="1560.256" Participant="me013">
      this in detail, but it <Emphasis> seems </Emphasis> like you're getting a <Emphasis> lot </Emphasis> from the channel adaptation, the speaker adaptation, and so forth.
    </Segment>
    <Segment StartTime="1554.986" EndTime="1555.558" Participant="mn017">
      Mmm.
    </Segment>
    <Segment StartTime="1558.782" EndTime="1559.402" Participant="mn017">
      Mm-hmm.
    </Segment>
    <Segment StartTime="1560.256" EndTime="1561.058" Participant="me013">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1561.058" EndTime="1565.384" Participant="me013">
      Um. So you are, already, in that recognizer, doing <Emphasis> something </Emphasis>
    </Segment>
    <Segment StartTime="1565.323" EndTime="1565.953" Participant="mn017">
      Mmm.
    </Segment>
    <Segment StartTime="1566.150" EndTime="1573.105" Participant="me013">
      that is likely to affect, uh, the - the far-field microphone, uh, <Uncertain> formant </Uncertain>. So, it may not -
    </Segment>
    <Segment StartTime="1569.966" EndTime="1571.060" Participant="mn017">
      Right.
    </Segment>
    <Segment StartTime="1573.105" EndTime="1574.419" Participant="me013">
      I mean, it's gonna be bad,
    </Segment>
    <Segment StartTime="1574.244" EndTime="1574.772" Participant="mn017">
      Right.
    </Segment>
    <Segment StartTime="1574.954" EndTime="1580.623" Participant="me013">
      but, uh, it may not be, like, "won't decode" kind of bad. It might - might <Emphasis> only </Emphasis> be that it
    </Segment>
    <Segment StartTime="1578.270" EndTime="1578.960" Participant="mn017">
      Right.
    </Segment>
    <Segment StartTime="1578.426" EndTime="1578.901" Participant="me001">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1580.545" EndTime="1581.558" Participant="me011">
      Yeah, eighty percent.
    </Segment>
    <Segment StartTime="1580.623" EndTime="1583.544" Participant="me013">
      goes from forty percent to eighty, or something like that. <Comment Description="last three words while laughing"/>
    </Segment>
    <Segment StartTime="1581.558" EndTime="1583.030" Participant="me011">
      <Pronounce Pronunciation="&quot;mm-yeah&quot;, due to laughing"> Yeah.  </Pronounce> <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="1583.010" EndTime="1583.436" Participant="mn017">
      Right.
    </Segment>
    <Segment StartTime="1583.470" EndTime="1585.580" Participant="fe016">
      Do you assume you know the speaker
    </Segment>
    <Segment StartTime="1585.957" EndTime="1587.091" Participant="fe016">
      when you do this?
    </Segment>
    <Segment StartTime="1586.844" EndTime="1589.722" Participant="me013">
      I want us to assume the exac- whatever it was you assumed
    </Segment>
    <Segment StartTime="1590.162" EndTime="1592.890" Participant="me013">
      when you did the other - the - the close mike.
    </Segment>
    <Segment StartTime="1590.240" EndTime="1591.235" Participant="mn017">
      You just di- you just -
    </Segment>
    <Segment StartTime="1591.539" EndTime="1599.540" Participant="fe016">
      Well, th- r- Well, <Emphasis> there, </Emphasis> there's only one person who it <Emphasis> can </Emphasis> be, because they own that microphone. I'm just wondering - There's -
    </Segment>
    <Segment StartTime="1592.790" EndTime="1593.720" Participant="mn017">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="1596.160" EndTime="1596.657" Participant="mn017">
      Right.
    </Segment>
    <Segment StartTime="1598.058" EndTime="1600.920" Participant="mn017">
      Yeah. That - that becomes another problem, actually.
    </Segment>
    <Segment StartTime="1599.516" EndTime="1601.889" Participant="me001">
      Well, and then you have the gender detection <Pause/> <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1599.950" EndTime="1601.193" Participant="fe016">
      No- It's not a pro-
    </Segment>
    <Segment StartTime="1600.012" EndTime="1600.764" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="1600.093" EndTime="1600.879" Participant="me011">
      Well, but for a - <Comment Description="loud, clipped"/>
    </Segment>
    <Segment StartTime="1601.193" EndTime="1602.820" Participant="fe016">
      Right. And -
    </Segment>
    <Segment StartTime="1601.251" EndTime="1605.046" Participant="me011">
      But, for s- for <Emphasis> scoring, </Emphasis> you can do it or not do it as you choose. So.
    </Segment>
    <Segment StartTime="1602.163" EndTime="1603.613" Participant="me013">
      Oh, I'm -
    </Segment>
    <Segment StartTime="1603.478" EndTime="1607.860" Participant="fe016">
      R- right. But in terms of - for norm- for adaptation. Right.
    </Segment>
    <Segment StartTime="1603.584" EndTime="1604.939" Participant="mn017">
      <VocalSound Description="inbreath"/> n- s-
    </Segment>
    <Segment StartTime="1604.599" EndTime="1607.180" Participant="me013">
      But you're saying for this - For the <Emphasis> adaptation, </Emphasis> you mean.
    </Segment>
    <Segment StartTime="1605.168" EndTime="1605.587" Participant="mn017">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1605.587" EndTime="1613.611" Participant="mn017">
      So, d- s- so - fo- not - well, for everything. For - s- for - f- even <Emphasis> feature </Emphasis> normalization, for, uh, vocal tract length estimation, <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1608.030" EndTime="1611.170" Participant="fe016">
      You know, do you do a supervised adaptation - ?
    </Segment>
    <Segment StartTime="1611.575" EndTime="1612.490" Participant="me013">
      Right.
    </Segment>
    <Segment StartTime="1612.229" EndTime="1615.030" Participant="fe016">
      Ye- <Emphasis> All </Emphasis> of these adaptations ha-
    </Segment>
    <Segment StartTime="1613.611" EndTime="1617.000" Participant="mn017">
      all of - all of these assume you know who's speaking. So,
    </Segment>
    <Segment StartTime="1615.500" EndTime="1617.243" Participant="fe016">
      assume that the same person -
    </Segment>
    <Segment StartTime="1616.197" EndTime="1616.703" Participant="mn014">
      Mmm.
    </Segment>
    <Segment StartTime="1617.000" EndTime="1621.618" Participant="mn017">
      you would have to do a speaker segmentation first on the far-field micropho- signal.
    </Segment>
    <Segment StartTime="1617.066" EndTime="1621.997" Participant="me018">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="1617.243" EndTime="1618.420" Participant="fe016">
      Yeah.
    </Segment>
    <Segment StartTime="1619.873" EndTime="1620.456" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1620.268" EndTime="1622.115" Participant="me011">
      <VocalSound Description="mouth"/> N- well, but you can use the -
    </Segment>
    <Segment StartTime="1622.115" EndTime="1626.567" Participant="me011">
      when you're doing the <Emphasis> scoring </Emphasis> - Since you're - you're gonna be scoring against <Emphasis> transcript, </Emphasis> you can use -
    </Segment>
    <Segment StartTime="1624.928" EndTime="1626.568" Participant="me013">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1626.518" EndTime="1627.661" Participant="mn017">
      You mean you wanna cheat.
    </Segment>
    <Segment StartTime="1628.190" EndTime="1629.303" Participant="me011">
      Well, you're doing that anyway.
    </Segment>
    <Segment StartTime="1629.255" EndTime="1632.585" Participant="fe008">
      Ooo. <Comment Description="rhymes with &quot;cool&quot;"/> I don't like that term. I <Emphasis> don't </Emphasis> like that term.
    </Segment>
    <Segment StartTime="1629.278" EndTime="1632.600" Participant="fe016">
      N- Well. I was just aski-
    </Segment>
    <Segment StartTime="1629.431" EndTime="1631.203" Participant="mn017">
      No. If - i- i-
    </Segment>
    <Segment StartTime="1630.010" EndTime="1633.378" Participant="me011">
      So ch- chea- try to cheat in the same way that you're doing with the close-talking.
    </Segment>
    <Segment StartTime="1633.150" EndTime="1635.032" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1633.917" EndTime="1635.507" Participant="fe008">
      I <Emphasis> don't </Emphasis> like that term.
    </Segment>
    <Segment StartTime="1634.070" EndTime="1636.160" Participant="fe016">
      Actually, the s- Those are two -
    </Segment>
    <Segment StartTime="1634.070" EndTime="1634.440" Participant="me013">
      I -
    </Segment>
    <Segment StartTime="1634.970" EndTime="1635.701" Participant="me013">
      I have -
    </Segment>
    <Segment StartTime="1635.032" EndTime="1637.480" Participant="mn017">
      O_K. We g- we're gonna bleep that out. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1635.801" EndTime="1636.181" Participant="me013">
      I -
    </Segment>
    <Segment StartTime="1636.640" EndTime="1639.233" Participant="me013">
      I have a suggestion. Do the simplest thing first.
    </Segment>
    <Segment StartTime="1638.862" EndTime="1639.803" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1639.020" EndTime="1641.058" Participant="me011">
      Yeah, right. <Comment Description="while laughing"/> <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1639.233" EndTime="1642.180" Participant="me013">
      <VocalSound Description="inbreath"/> Because we're gonna want to know that anyway.
    </Segment>
    <Segment StartTime="1642.070" EndTime="1645.020" Participant="fe016">
      So, <Uncertain> wait </Uncertain>, the simplest thing is you cheat, saying -
    </Segment>
    <Segment StartTime="1642.180" EndTime="1643.352" Participant="me013">
      In other words, if you d-
    </Segment>
    <Segment StartTime="1644.563" EndTime="1647.950" Participant="me013">
      No. It's - No. It's even simpler thing than that, is just that you <Emphasis> don't </Emphasis> know.
    </Segment>
    <Segment StartTime="1648.427" EndTime="1651.640" Participant="mn017">
      You mean you do- you don't do all those normalizations.
    </Segment>
    <Segment StartTime="1650.810" EndTime="1651.200" Participant="me011">
      Yeah.
    </Segment>
    <Segment StartTime="1651.200" EndTime="1651.734" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="1651.960" EndTime="1655.470" Participant="fe016">
      Oh, you m- y- totally unadapted. Uh. Right.
    </Segment>
    <Segment StartTime="1652.500" EndTime="1654.240" Participant="me011">
      Just do - or a free ri- Yeah.
    </Segment>
    <Segment StartTime="1653.930" EndTime="1658.438" Participant="me013">
      Yeah. Because you can get a number - uh, for that with the other as well. Right? You can turn those things off. Right?
    </Segment>
    <Segment StartTime="1658.479" EndTime="1664.292" Participant="mn017">
      <VocalSound Description="mouth"/> Um, actually - <VocalSound Description="laugh"/> We don't <Emphasis> have </Emphasis> any models. Um, you can -
    </Segment>
    <Segment StartTime="1658.641" EndTime="1660.180" Participant="fe016">
      <VocalSound Description="vocal squeak"/> Yeah.
    </Segment>
    <Segment StartTime="1659.687" EndTime="1663.070" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1660.180" EndTime="1661.277" Participant="fe016">
      i- i-
    </Segment>
    <Segment StartTime="1662.561" EndTime="1664.170" Participant="fe016">
      No. You can -
    </Segment>
    <Segment StartTime="1664.292" EndTime="1666.171" Participant="mn017">
      Um - <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="1665.240" EndTime="1668.060" Participant="fe016">
      You can use a speaker - eh - What about <Emphasis> gender </Emphasis> detection?
    </Segment>
    <Segment StartTime="1666.171" EndTime="1673.760" Participant="mn017">
      A- actually, it's that - it's - it's - We would have to retrain models that are not - that have none of that stuff,
    </Segment>
    <Segment StartTime="1674.037" EndTime="1674.753" Participant="mn017">
      uh,
    </Segment>
    <Segment StartTime="1674.753" EndTime="1677.220" Participant="mn017">
      in it. But actually we could -
    </Segment>
    <Segment StartTime="1678.678" EndTime="1684.260" Participant="mn017">
      We can just run it, assuming that it's all one speaker, basically. And see what happens.
    </Segment>
    <Segment StartTime="1682.220" EndTime="1682.774" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="1683.532" EndTime="1684.018" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="1684.018" EndTime="1689.858" Participant="me013">
      <VocalSound Description="inbreath"/> Yeah. And then put it in correctly and see how much that helps. I mean, I <Pronounce Pronunciation="reduced"> was  </Pronounce> just thinking, do the one that's easiest first,
    </Segment>
    <Segment StartTime="1684.260" EndTime="1686.062" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1686.406" EndTime="1687.450" Participant="fe016">
      Yeah.
    </Segment>
    <Segment StartTime="1686.709" EndTime="1687.738" Participant="mn017">
      <Uncertain> Yeah. </Uncertain>
    </Segment>
    <Segment StartTime="1689.858" EndTime="1693.240" Participant="me013">
      because you wanna know <Pause/> how much that's helping you in these cases anyhow.
    </Segment>
    <Segment StartTime="1691.260" EndTime="1693.040" Participant="mn017">
      <VocalSound Description="inbreath"/> Actually -
    </Segment>
    <Segment StartTime="1692.133" EndTime="1692.631" Participant="mn014">
      But d-
    </Segment>
    <Segment StartTime="1692.900" EndTime="1694.801" Participant="mn014">
      do you have gender-dependent models?
    </Segment>
    <Segment StartTime="1693.240" EndTime="1694.712" Participant="mn017">
      Actually, no. Th- th-
    </Segment>
    <Segment StartTime="1694.712" EndTime="1695.550" Participant="mn017">
      Sorry.
    </Segment>
    <Segment StartTime="1694.801" EndTime="1697.008" Participant="mn014">
      A- Are the models gender-dependent?
    </Segment>
    <Segment StartTime="1696.014" EndTime="1696.797" Participant="me001">
      Mmm.
    </Segment>
    <Segment StartTime="1697.157" EndTime="1698.020" Participant="mn017">
      Yeah.
    </Segment>
    <Segment StartTime="1697.420" EndTime="1700.233" Participant="me011">
      Yeah. They're all gender-dependent. So we would have to at least do that.
    </Segment>
    <Segment StartTime="1697.580" EndTime="1702.337" Participant="fe016">
      Yeah. <Pause/> Yeah. So. You can r- No. You can run <Emphasis> both. </Emphasis> And you can s-
    </Segment>
    <Segment StartTime="1697.628" EndTime="1698.832" Participant="mn014">
      So <Pause/>
    </Segment>
    <Segment StartTime="1698.020" EndTime="1699.910" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1699.910" EndTime="1701.300" Participant="mn017">
      No, actually -
    </Segment>
    <Segment StartTime="1700.090" EndTime="1700.811" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="1701.436" EndTime="1701.995" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="1701.560" EndTime="1705.708" Participant="mn017">
      No, actually, what - Here's - here's what we would usually do u- under these circumstances.
    </Segment>
    <Segment StartTime="1701.790" EndTime="1702.937" Participant="me011">
      And pick whichever's better.
    </Segment>
    <Segment StartTime="1703.601" EndTime="1704.510" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1705.708" EndTime="1709.180" Participant="mn017">
      We would actually - we would run some sort of segmentation.
    </Segment>
    <Segment StartTime="1706.301" EndTime="1706.914" Participant="mn014">
      <VocalSound Description="clears throat"/>
    </Segment>
    <Segment StartTime="1709.530" EndTime="1712.486" Participant="mn017">
      Thilo's is as <Pause/> good as any, probably.
    </Segment>
    <Segment StartTime="1712.042" EndTime="1712.900" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1712.486" EndTime="1719.256" Participant="mn017">
      Um, and then we would do an unsupervised clustering of - of the segments, to - and - and put
    </Segment>
    <Segment StartTime="1719.256" EndTime="1726.884" Participant="mn017">
      the similar ones into bins that would be sort of pseudo-speakers. And then we would do our standard processing on these pseudo-speakers.
    </Segment>
    <Segment StartTime="1722.200" EndTime="1723.162" Participant="me013">
      Mm-hmm.
    </Segment>
    <Segment StartTime="1726.884" EndTime="1733.612" Participant="mn017">
      <VocalSound Description="inbreath"/> And that turns out to work very well on Broadcast News, SPINE - those types of tasks, where you don't have
    </Segment>
    <Segment StartTime="1733.612" EndTime="1735.580" Participant="mn017">
      the speaker segmentation given to you.
    </Segment>
    <Segment StartTime="1735.696" EndTime="1739.690" Participant="me011">
      S- does the clustering - ? Do you give it sort of a target number of clusters? Or is it - ? <Comment Description="leading question"/>
    </Segment>
    <Segment StartTime="1740.245" EndTime="1746.806" Participant="mn017">
      <VocalSound Description="mouth"/> Um, you can either do it by target number or by some measure of dissimilarity that you use as a threshold.
    </Segment>
    <Segment StartTime="1740.904" EndTime="1741.975" Participant="me011">
      adapted in some way?
    </Segment>
    <Segment StartTime="1745.077" EndTime="1745.616" Participant="me011">
      Mm-hmm.
    </Segment>
    <Segment StartTime="1745.895" EndTime="1751.376" Participant="me011">
      That's what I'm just thinking one of the big differences with Broadcast News and these meetings is we have m- many fewer participants.
    </Segment>
    <Segment StartTime="1752.068" EndTime="1754.600" Participant="fe016">
      The other thing is that you actually have
    </Segment>
    <Segment StartTime="1753.570" EndTime="1754.280" Participant="mn017">
      Right.
    </Segment>
    <Segment StartTime="1755.136" EndTime="1756.634" Participant="fe016">
      direction here. So,
    </Segment>
    <Segment StartTime="1757.290" EndTime="1762.777" Participant="fe016">
      unlike these corpora that are recorded with other microphones, like - The <Emphasis> right </Emphasis> way to do this I guess, w-
    </Segment>
    <Segment StartTime="1762.777" EndTime="1770.760" Participant="fe016">
      you know, in the <Emphasis> future </Emphasis> would be, well, in general, Thilo's sitting there, and this P_Z_M is gonna - You know, he's roughly <Pause/> in that location.
    </Segment>
    <Segment StartTime="1764.520" EndTime="1765.792" Participant="me011">
      Speaker I_D. Yeah.
    </Segment>
    <Segment StartTime="1767.910" EndTime="1773.353" Participant="me011">
      <VocalSound Description="inbreath"/> Well, there're different ways of thinking about it. I mean that - that would be true if w- you had a meeting situation with multiple mikes.
    </Segment>
    <Segment StartTime="1773.353" EndTime="1775.894" Participant="me011">
      But if you only had your P_D_A sitting in front of you -
    </Segment>
    <Segment StartTime="1776.620" EndTime="1782.700" Participant="fe016">
      Well, any case where the people are not all sitting at the same place and they're not moving around <Emphasis> too </Emphasis> much.
    </Segment>
    <Segment StartTime="1782.140" EndTime="1783.660" Participant="me011">
      <Emphasis> And </Emphasis> you have more than one mike.
    </Segment>
    <Segment StartTime="1784.002" EndTime="1788.050" Participant="me013">
      Yeah. If you don't have one - more than one mike, you don't have a very good handle on location.
    </Segment>
    <Segment StartTime="1788.326" EndTime="1791.290" Participant="fe016">
      W- well, you have distance, and you have -
    </Segment>
    <Segment StartTime="1790.987" EndTime="1792.490" Participant="me013">
      That's - it's not enough.
    </Segment>
    <Segment StartTime="1792.088" EndTime="1792.957" Participant="fe016">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1793.900" EndTime="1794.712" Participant="fe016">
      I mean, that,
    </Segment>
    <Segment StartTime="1794.970" EndTime="1802.110" Participant="fe016">
      um, Jane's - i- n- The pickup of Adam on this mike is gonna be different than me, in terms of energy and so forth over the whole meeting.
    </Segment>
    <Segment StartTime="1800.881" EndTime="1804.970" Participant="me011">
      <VocalSound Description="mouth"/> Oh, so just from clustering. You might be able to cluster it better because of that.
    </Segment>
    <Segment StartTime="1803.710" EndTime="1810.340" Participant="fe016">
      You might get some clustering from the speaker and some of it from the characteristics of the distance, and - <Comment Description="abandoned"/>
    </Segment>
    <Segment StartTime="1804.771" EndTime="1805.650" Participant="me013">
      Mmm. <Comment Description="doubtfully"/>
    </Segment>
    <Segment StartTime="1807.543" EndTime="1808.963" Participant="me011">
      From mike. Yeah.
    </Segment>
    <Segment StartTime="1808.738" EndTime="1809.302" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="1809.302" EndTime="1809.730" Participant="me013">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1809.730" EndTime="1811.215" Participant="me013">
      But, say, if - if you had a -
    </Segment>
    <Segment StartTime="1811.347" EndTime="1812.673" Participant="me011">
      And transfer functions.
    </Segment>
    <Segment StartTime="1811.940" EndTime="1815.098" Participant="me013">
      a cardioid mike or something sitting someplace, then - <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1812.151" EndTime="1813.920" Participant="fe016">
      Right. Exactly.
    </Segment>
    <Segment StartTime="1815.098" EndTime="1819.973" Participant="me013">
      sitting <Emphasis> there, </Emphasis> then it would - Its - its response to <Emphasis> him </Emphasis> would be about the same as the response to him, and so on.
    </Segment>
    <Segment StartTime="1816.146" EndTime="1817.700" Participant="mn017">
      Well, you can do -
    </Segment>
    <Segment StartTime="1817.637" EndTime="1819.054" Participant="me011">
      Well, I think there're lots of -
    </Segment>
    <Segment StartTime="1818.433" EndTime="1823.642" Participant="mn017">
      <VocalSound Description="mouth"/> You can - you can - you can do certain normalizations like, you know, gain control, <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1818.810" EndTime="1821.620" Participant="fe016">
      They're <Emphasis> both </Emphasis> picked up in the clustering, I guess.
    </Segment>
    <Segment StartTime="1819.226" EndTime="1820.580" Participant="me011">
      lots of ways of doing it.
    </Segment>
    <Segment StartTime="1821.871" EndTime="1822.504" Participant="me013">
      Mm-hmm. <Comment Description="very quietly"/>
    </Segment>
    <Segment StartTime="1823.642" EndTime="1828.000" Participant="mn017">
      uh, before you do the clustering to rule out those - those types of things.
    </Segment>
    <Segment StartTime="1826.880" EndTime="1830.000" Participant="fe016">
      Or to just do the clustering, knowing that you're capturing
    </Segment>
    <Segment StartTime="1830.268" EndTime="1834.350" Participant="fe016">
      <Emphasis> both. </Emphasis> It's just that the kind of clustering we've done before hasn't had that,
    </Segment>
    <Segment StartTime="1831.075" EndTime="1831.845" Participant="me013">
      I see.
    </Segment>
    <Segment StartTime="1834.920" EndTime="1836.585" Participant="fe016">
      uh, distance factor, or
    </Segment>
    <Segment StartTime="1836.585" EndTime="1839.890" Participant="fe016">
      location factor in it in the same way. And so we're not really
    </Segment>
    <Segment StartTime="1836.960" EndTime="1837.437" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="1839.890" EndTime="1846.437" Participant="fe016">
      modeling it directly. If somebody <Emphasis> does, </Emphasis> we sh- maybe we <Emphasis> add </Emphasis> that because I think it would be a pretty big difference. When you <Emphasis> listen, </Emphasis>
    </Segment>
    <Segment StartTime="1841.194" EndTime="1842.415" Participant="mn017">
      Mmm. <Comment Description="very quietly"/>
    </Segment>
    <Segment StartTime="1842.008" EndTime="1843.523" Participant="me011">
      Yeah. That's an <Emphasis> interesting </Emphasis> -
    </Segment>
    <Segment StartTime="1843.881" EndTime="1845.630" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1846.437" EndTime="1850.560" Participant="fe016">
      you can sort of tell where people <Emphasis> are, </Emphasis> not which s- you know, <Emphasis> side, </Emphasis> but - <Comment Description="abandoned"/>
    </Segment>
    <Segment StartTime="1846.538" EndTime="1847.921" Participant="me011">
      It's a big difference. Yeah.
    </Segment>
    <Segment StartTime="1848.103" EndTime="1849.990" Participant="mn017">
      <Pronounce Pronunciation="&quot;kay&quot;"> O_K.  </Pronounce> That would be fun -
    </Segment>
    <Segment StartTime="1850.246" EndTime="1851.840" Participant="mn017">
      fun to try.
    </Segment>
    <Segment StartTime="1851.160" EndTime="1852.775" Participant="me011">
      Well, humans are really good at that -
    </Segment>
    <Segment StartTime="1852.775" EndTime="1856.516" Participant="me011">
      transfer function through the head, and things like that. So you know, even if you only have one ear,
    </Segment>
    <Segment StartTime="1854.120" EndTime="1856.740" Participant="fe016">
      Right. Even with one microphone.
    </Segment>
    <Segment StartTime="1855.081" EndTime="1855.820" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="1857.104" EndTime="1859.638" Participant="me011">
      you can still get - get good transfers. So <Pause/>
    </Segment>
    <Segment StartTime="1857.139" EndTime="1857.618" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="1857.880" EndTime="1865.722" Participant="fe016">
      Right. So our - our clustering is not gonna be intelligent that way. It's just gonna pick up whatever energy difference, or whatever, is - <Comment Description="abandoned"/>
    </Segment>
    <Segment StartTime="1864.645" EndTime="1865.236" Participant="me011">
      Yeah.
    </Segment>
    <Segment StartTime="1864.930" EndTime="1866.055" Participant="me013">
      <VocalSound Description="inbreath"/> But -
    </Segment>
    <Segment StartTime="1866.055" EndTime="1871.470" Participant="me013">
      Anyway, I - I'd - it'd be neat to have that, because <Uncertain> you know </Uncertain> we've been at this for a little while and we don't have - have -
    </Segment>
    <Segment StartTime="1870.957" EndTime="1871.533" Participant="mn017">
      Mm-hmm.
    </Segment>
    <Segment StartTime="1871.470" EndTime="1875.465" Participant="me013">
      any results yet with - with conversational speech at a distance.
    </Segment>
    <Segment StartTime="1872.996" EndTime="1873.750" Participant="mn017">
      O_K.
    </Segment>
    <Segment StartTime="1875.465" EndTime="1876.428" Participant="me013">
      So, um -
    </Segment>
    <Segment StartTime="1876.428" EndTime="1877.000" Participant="me013">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1876.596" EndTime="1877.120" Participant="me011">
      Yeah. <Comment Description="very quietly"/>
    </Segment>
    <Segment StartTime="1876.650" EndTime="1877.000" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="1877.000" EndTime="1878.655" Participant="me013">
      We should at least get <Pause/> a first one.
    </Segment>
    <Segment StartTime="1878.280" EndTime="1879.190" Participant="me011">
      Something. Yeah.
    </Segment>
    <Segment StartTime="1878.676" EndTime="1879.951" Participant="fe016">
      O_K. <Uncertain> Hmm. </Uncertain>
    </Segment>
    <Segment StartTime="1879.014" EndTime="1879.910" Participant="fe008">
      Mm-hmm. <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="1879.140" EndTime="1883.091" Participant="me013">
      Um, and the other thing - This would kind of be a Hail Mary, but - but, uh - <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1882.797" EndTime="1885.108" Participant="me011">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1883.576" EndTime="1890.238" Participant="me013">
      Uh, Dave does have this stuff that is helping on digits, and, you know, and so with - then it'd be, you know, just throw that in and see.
    </Segment>
    <Segment StartTime="1886.930" EndTime="1888.250" Participant="mn017">
      Oh, yeah. Then we should -
    </Segment>
    <Segment StartTime="1886.985" EndTime="1889.119" Participant="me011">
      Yeah. So it'd be <Emphasis> cool </Emphasis> to see if it helped.
    </Segment>
    <Segment StartTime="1889.470" EndTime="1893.984" Participant="mn017">
      Well, first you have to filter the whole training set and retrain. <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1889.687" EndTime="1890.640" Participant="fe016">
      Yeah.
    </Segment>
    <Segment StartTime="1889.886" EndTime="1890.692" Participant="me001">
      Mm-hmm.
    </Segment>
    <Segment StartTime="1892.660" EndTime="1893.159" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="1893.771" EndTime="1894.373" Participant="fe016">
      <Uncertain> Mm-hmm. </Uncertain>
    </Segment>
    <Segment StartTime="1893.900" EndTime="1895.469" Participant="me011">
      That would be - ft- quick,
    </Segment>
    <Segment StartTime="1895.178" EndTime="1896.670" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1895.352" EndTime="1896.711" Participant="me013">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1895.469" EndTime="1896.360" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1896.360" EndTime="1900.438" Participant="me011">
      since I think he did it in Matlab. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1897.970" EndTime="1899.403" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1898.517" EndTime="1900.160" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1898.655" EndTime="1899.530" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1898.790" EndTime="1899.800" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1899.403" EndTime="1903.199" Participant="me013">
      Uh, well, he can do it in something else. But, I mean - <VocalSound Description="laugh"/> you know, it's -
    </Segment>
    <Segment StartTime="1899.452" EndTime="1900.340" Participant="fe008">
      Hmm. Interesting. <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="1901.380" EndTime="1902.011" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="1902.130" EndTime="1903.662" Participant="mn017">
      <VocalSound Description="laugh"/> Right. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1903.600" EndTime="1905.775" Participant="me011">
      Can't you export C_ from Matlab,
    </Segment>
    <Segment StartTime="1903.960" EndTime="1904.468" Participant="me013">
      Um -
    </Segment>
    <Segment StartTime="1905.470" EndTime="1909.723" Participant="me013">
      Actually, we're experimenting with phase stuff now, and - and, uh, thi- this, uh -
    </Segment>
    <Segment StartTime="1905.775" EndTime="1907.300" Participant="me011">
      or is that Mathematica?
    </Segment>
    <Segment StartTime="1910.070" EndTime="1913.661" Participant="me013">
      <Emphasis> first </Emphasis> result he got, uh, was really great. It actually, uh - uh -
    </Segment>
    <Segment StartTime="1913.661" EndTime="1917.435" Participant="me013">
      <VocalSound Description="breath"/> didn't exactly eliminate the reverberation, but it <Emphasis> completely </Emphasis> got rid of the speech.
    </Segment>
    <Segment StartTime="1917.597" EndTime="1921.263" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1917.604" EndTime="1920.350" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1917.754" EndTime="1920.870" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1917.856" EndTime="1927.053" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1918.050" EndTime="1919.320" Participant="mn017">
      <VocalSound Description="inbreath"/> Oh.
    </Segment>
    <Segment StartTime="1918.154" EndTime="1919.530" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1919.951" EndTime="1923.227" Participant="mn017">
      That would solve all of our problems. Wouldn't it? <VocalSound Description="laugh"/> <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="1920.172" EndTime="1921.069" Participant="fe008">
      Wow. <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="1920.870" EndTime="1923.240" Participant="me013">
      Yeah. Well, I was thinking that. So, it's - <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1921.085" EndTime="1922.800" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1921.263" EndTime="1923.556" Participant="me011">
      Well, so just take the inverse and you're fine. <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="1923.227" EndTime="1932.048" Participant="mn017">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="1923.240" EndTime="1923.987" Participant="me013">
      That's -
    </Segment>
    <Segment StartTime="1923.449" EndTime="1925.763" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1923.987" EndTime="1926.162" Participant="me013">
      You think I didn't s- <VocalSound Description="laugh"/> tell him that?
    </Segment>
    <Segment StartTime="1924.664" EndTime="1926.814" Participant="me051">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1924.845" EndTime="1927.577" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1925.401" EndTime="1926.796" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1926.162" EndTime="1931.556" Participant="me013">
      No. I got pretty excited, because it <Emphasis> completely </Emphasis> got rid of the speech. So, I was thinking "well, so,
    </Segment>
    <Segment StartTime="1927.546" EndTime="1928.540" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1929.560" EndTime="1931.820" Participant="me011">
      So that's a speech detector. That's great.
    </Segment>
    <Segment StartTime="1931.556" EndTime="1934.937" Participant="me013">
      you know, it could be useful for lots of things". So, we have to - <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1931.660" EndTime="1932.648" Participant="fe008">
      That's interesting.
    </Segment>
    <Segment StartTime="1933.335" EndTime="1934.604" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="1933.380" EndTime="1935.002" Participant="fe008">
      Did it get rid of other stuff, too, though?
    </Segment>
    <Segment StartTime="1934.937" EndTime="1935.295" Participant="me013">
      What?
    </Segment>
    <Segment StartTime="1935.002" EndTime="1940.772" Participant="fe008">
      Did it get rid of other stuff besides the speech? O_K. <Comment Description="while laughing"/> Wow. Interesting.
    </Segment>
    <Segment StartTime="1935.090" EndTime="1936.570" Participant="me011">
      Yeah. Just subtract that -
    </Segment>
    <Segment StartTime="1936.490" EndTime="1938.830" Participant="me013">
      Well, we have to sort of check that out. That's -
    </Segment>
    <Segment StartTime="1937.300" EndTime="1940.040" Participant="me011">
      subtract that from the original signal and you're set.
    </Segment>
    <Segment StartTime="1937.841" EndTime="1939.262" Participant="fe016">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1938.129" EndTime="1939.160" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1939.967" EndTime="1940.577" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="1940.135" EndTime="1941.461" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1940.407" EndTime="1941.590" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1940.866" EndTime="1941.462" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="1941.457" EndTime="1943.057" Participant="me011">
      <VocalSound Description="breath-laugh"/> <Comment Description="very quietly"/>
    </Segment>
    <Segment StartTime="1942.346" EndTime="1944.101" Participant="fe008">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1943.040" EndTime="1945.960" Participant="mn017">
      Right. Then you can do like an - you can estimate the -
    </Segment>
    <Segment StartTime="1946.457" EndTime="1949.030" Participant="me011">
      Noise estimate. Signal to noise. That's great.
    </Segment>
    <Segment StartTime="1946.622" EndTime="1949.100" Participant="mn017">
      the noise estimates. Right? Yeah.
    </Segment>
    <Segment StartTime="1949.125" EndTime="1949.728" Participant="fe016">
      Mm-hmm.
    </Segment>
    <Segment StartTime="1950.048" EndTime="1954.326" Participant="me013">
      Reminds me of when th- when Herve and I were first playing with c- uh, context-dependent things for nets, and -
    </Segment>
    <Segment StartTime="1954.326" EndTime="1957.349" Participant="me013">
      and at one point we took out the speech input, so we <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="1957.349" EndTime="1959.140" Participant="me013">
      only had priors, and our performance went up.
    </Segment>
    <Segment StartTime="1959.140" EndTime="1960.805" Participant="me011">
      <VocalSound Description="laugh"/> No. <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="1959.470" EndTime="1965.469" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1959.711" EndTime="1960.150" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1959.964" EndTime="1960.261" Participant="fe008">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="1960.150" EndTime="1964.417" Participant="mn017">
      Mmm <Pause/> <VocalSound Description="laugh"/> Yeah. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1960.587" EndTime="1962.664" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1960.805" EndTime="1962.954" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1960.941" EndTime="1961.928" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1961.440" EndTime="1962.450" Participant="fe008">
      Wow.
    </Segment>
    <Segment StartTime="1962.954" EndTime="1968.847" Participant="me011">
      I guess that's why Herve always talks about using the priors as one of the mixtures in - in his all-ways combos.
    </Segment>
    <Segment StartTime="1966.337" EndTime="1966.994" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="1966.690" EndTime="1967.417" Participant="me013">
      S- Yeah. <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="1968.570" EndTime="1971.279" Participant="me013">
      Well, of course it was a bug. But, I mean, it's - it's a - but it was pretty
    </Segment>
    <Segment StartTime="1969.953" EndTime="1971.220" Participant="me011">
      But still - <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1971.279" EndTime="1974.275" Participant="me013">
      f- f- <VocalSound Description="laugh"/> It was pretty funny anyway. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1971.660" EndTime="1972.731" Participant="fe008">
      Wow. Interesting.
    </Segment>
    <Segment StartTime="1974.371" EndTime="1974.792" Participant="mn017">
      Yeah.
    </Segment>
    <Segment StartTime="1974.929" EndTime="1975.719" Participant="fe008">
      Hmm. <Comment Description="very quietly"/>
    </Segment>
    <Segment StartTime="1975.389" EndTime="1977.136" Participant="me013">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1975.670" EndTime="1979.979" Participant="me011">
      So if you run - er, your recognizer with all probabilities equal, what do you get out?
    </Segment>
    <Segment StartTime="1980.182" EndTime="1981.725" Participant="mn017">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="1980.800" EndTime="1981.816" Participant="me011">
      Probably garbage.
    </Segment>
    <Segment StartTime="1982.446" EndTime="1984.500" Participant="me018">
      Whatever the language model says. <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1982.481" EndTime="1983.431" Participant="me011">
      I bet the pruning -
    </Segment>
    <Segment StartTime="1983.355" EndTime="1983.983" Participant="mn014">
      Yep.
    </Segment>
    <Segment StartTime="1983.894" EndTime="1985.690" Participant="me011">
      The pruning probably prunes everything out.
    </Segment>
    <Segment StartTime="1985.017" EndTime="1989.460" Participant="mn017">
      You get out Switchboard. That's just the lang- the language model. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1986.490" EndTime="1989.270" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1986.536" EndTime="1988.992" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1986.671" EndTime="1989.573" Participant="me018">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="1986.674" EndTime="1989.170" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1986.770" EndTime="1989.774" Participant="me011">
      Yeah. That's right. <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="1988.664" EndTime="1990.345" Participant="fe008">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="1989.710" EndTime="1990.890" Participant="me013">
      So <Emphasis> that's </Emphasis> how it was generated.
    </Segment>
    <Segment StartTime="1989.958" EndTime="1990.724" Participant="mn017">
      Ha- <Comment Description="very quietly"/>
    </Segment>
    <Segment StartTime="1990.724" EndTime="1993.148" Participant="mn017">
      Yeah. So we have this new speaker adaptation.
    </Segment>
    <Segment StartTime="1993.480" EndTime="1994.190" Participant="mn017">
      Um <Pause/>
    </Segment>
    <Segment StartTime="1994.576" EndTime="1997.250" Participant="mn017">
      A - b- Oh, it's a s- sort of feature normalization t-
    </Segment>
    <Segment StartTime="1997.250" EndTime="1999.810" Participant="mn017">
      uh, like f- speaker adaptation, which,
    </Segment>
    <Segment StartTime="1999.970" EndTime="2006.520" Participant="mn017">
      uh, which I wr- which I <Pause/> wrote about in the last status report, which seems to be helping about a percent and a half on
    </Segment>
    <Segment StartTime="2006.520" EndTime="2007.316" Participant="mn017">
      Hub-five.
    </Segment>
    <Segment StartTime="2008.070" EndTime="2013.210" Participant="mn017">
      So, um. We haven't tried that yet on the meetings, uh, but hopefully it'll help there, too.
    </Segment>
    <Segment StartTime="2012.870" EndTime="2018.157" Participant="fe008">
      <VocalSound Description="inbreath"/> I wanna ask, um - So, you know that the data - I'v- I have upgraded it considerably. So I've probably made -
    </Segment>
    <Segment StartTime="2018.157" EndTime="2019.874" Participant="fe008">
      I probably corrected something like,
    </Segment>
    <Segment StartTime="2019.129" EndTime="2021.971" Participant="me001">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="2020.277" EndTime="2024.921" Participant="fe008">
      well, s- It's really a substantial amount of things that I've caught, changed - um, added to it.
    </Segment>
    <Segment StartTime="2024.921" EndTime="2030.562" Participant="fe008">
      Including a lot of, uh, backchannels. So, when you're d- running things, if you run it on the old -
    </Segment>
    <Segment StartTime="2028.248" EndTime="2028.683" Participant="me011">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="2030.562" EndTime="2034.916" Participant="fe008">
      So. If you - if you - run it on the <Emphasis> new </Emphasis> version, then the numbers <Emphasis> will </Emphasis> be -
    </Segment>
    <Segment StartTime="2033.009" EndTime="2033.406" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="2034.916" EndTime="2042.247" Participant="fe008">
      um, and you compare it to the - to runs on the <Emphasis> old </Emphasis> version, then you're gonna end up with more of an improvement than would actually be the case.
    </Segment>
    <Segment StartTime="2035.461" EndTime="2036.054" Participant="me011">
      Different.
    </Segment>
    <Segment StartTime="2037.976" EndTime="2038.561" Participant="mn017">
      Mm-hmm.
    </Segment>
    <Segment StartTime="2040.830" EndTime="2047.910" Participant="mn017">
      Well, we - we have a frozen - we do all our experiments with a frozen version of the transcripts as of, I don't know <Pause/>
    </Segment>
    <Segment StartTime="2048.000" EndTime="2049.086" Participant="fe008">
      As of February?
    </Segment>
    <Segment StartTime="2048.316" EndTime="2052.570" Participant="mn017">
      A- a- as of - no, a little - I don't know. When - when did we
    </Segment>
    <Segment StartTime="2051.795" EndTime="2053.679" Participant="me001">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="2052.522" EndTime="2053.291" Participant="fe008">
      So a l-
    </Segment>
    <Segment StartTime="2052.580" EndTime="2054.360" Participant="mn017">
      grab the transcripts?
    </Segment>
    <Segment StartTime="2053.654" EndTime="2054.965" Participant="fe008">
      Like the H_L_T paper?
    </Segment>
    <Segment StartTime="2053.679" EndTime="2057.560" Participant="me001">
      The b- for the f- ? It was m- m- March, probably.
    </Segment>
    <Segment StartTime="2054.638" EndTime="2056.163" Participant="fe016">
      F- Sorry.
    </Segment>
    <Segment StartTime="2056.120" EndTime="2057.186" Participant="mn017">
      The - <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2056.163" EndTime="2057.560" Participant="fe016">
      For these meetings?
    </Segment>
    <Segment StartTime="2057.420" EndTime="2063.570" Participant="mn017">
      We're talking about which - which version we're using for evaluating the recognition. Which version of the transcripts.
    </Segment>
    <Segment StartTime="2062.526" EndTime="2065.470" Participant="fe016">
      Right. They're somewhere in between
    </Segment>
    <Segment StartTime="2065.470" EndTime="2069.100" Participant="fe016">
      January and late March, or something like that?
    </Segment>
    <Segment StartTime="2068.222" EndTime="2069.610" Participant="me001">
      Yeah. Something around there.
    </Segment>
    <Segment StartTime="2068.270" EndTime="2071.907" Participant="fe008">
      O_K. Well, so long as you have the same baseline, then <Pause/> you'll be able to tell.
    </Segment>
    <Segment StartTime="2069.372" EndTime="2071.223" Participant="me011">
      But they are channelized ones, though?
    </Segment>
    <Segment StartTime="2070.422" EndTime="2073.010" Participant="mn017">
      No, no. Yeah. O- obviously. Yeah.
    </Segment>
    <Segment StartTime="2071.311" EndTime="2072.286" Participant="fe016">
      N-
    </Segment>
    <Segment StartTime="2072.286" EndTime="2072.862" Participant="fe016">
      N-
    </Segment>
    <Segment StartTime="2072.862" EndTime="2076.620" Participant="fe016">
      They're <Emphasis> meetings </Emphasis> that are <Emphasis> now </Emphasis> channelized, but they were not from -
    </Segment>
    <Segment StartTime="2073.010" EndTime="2074.435" Participant="mn017">
      The - uh, the other thing is -
    </Segment>
    <Segment StartTime="2073.010" EndTime="2079.356" Participant="fe008">
      So long as i- so long as it's the same baseline you'll be able to tell. But - but I'm just gonn- I'm just saying that if you were to compare that with running that on the <Emphasis> new </Emphasis> data,
    </Segment>
    <Segment StartTime="2074.435" EndTime="2076.720" Participant="mn017">
      Th- the - the -
    </Segment>
    <Segment StartTime="2079.220" EndTime="2080.373" Participant="mn017">
      Hmm. And -
    </Segment>
    <Segment StartTime="2079.220" EndTime="2084.332" Participant="me001">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="2079.356" EndTime="2082.430" Participant="fe008">
      that it would be an - a more optimistic <Pronounce Pronunciation="&quot;outscome&quot;"> outcome.  </Pronounce>
    </Segment>
    <Segment StartTime="2080.600" EndTime="2088.850" Participant="mn017">
      Well, the - and the other thing is, it takes only a - a minute to rescore all the old outputs with - If you had new transcripts, then
    </Segment>
    <Segment StartTime="2088.850" EndTime="2091.610" Participant="mn017">
      we j- we just re- rescore the old -
    </Segment>
    <Segment StartTime="2090.090" EndTime="2091.580" Participant="me011">
      Right. Cuz you haven't done any training.
    </Segment>
    <Segment StartTime="2091.720" EndTime="2092.480" Participant="mn017">
      Sorry?
    </Segment>
    <Segment StartTime="2091.976" EndTime="2093.880" Participant="me011">
      Right. Cuz we're not doing it for training.
    </Segment>
    <Segment StartTime="2091.996" EndTime="2094.848" Participant="me013">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="2092.339" EndTime="2097.160" Participant="fe016">
      Yeah. We haven't a- modified the recognizer at all. So, actually, um,
    </Segment>
    <Segment StartTime="2094.860" EndTime="2096.969" Participant="me011">
      Right. So it's really - it would be really easy to re-do it.
    </Segment>
    <Segment StartTime="2095.001" EndTime="2097.510" Participant="mn017">
      W- we just - We save -
    </Segment>
    <Segment StartTime="2097.160" EndTime="2103.803" Participant="fe016">
      at some point we <Emphasis> should </Emphasis> update and rescore everything with, you know, the corrected transcripts,
    </Segment>
    <Segment StartTime="2101.855" EndTime="2105.377" Participant="me011">
      Yeah. It'd be interesting just to see i- how much it changes. I bet it wouldn't change a lot.
    </Segment>
    <Segment StartTime="2101.910" EndTime="2103.740" Participant="mn017">
      <VocalSound Description="inbreath"/> Well, the - the -
    </Segment>
    <Segment StartTime="2103.803" EndTime="2105.220" Participant="fe016">
      just to h-
    </Segment>
    <Segment StartTime="2104.399" EndTime="2106.110" Participant="mn017">
      Right. <Comment Description="lengthened"/>
    </Segment>
    <Segment StartTime="2105.165" EndTime="2107.914" Participant="me001">
      <VocalSound Description="mouth"/> What are the nature of most of the changes?
    </Segment>
    <Segment StartTime="2105.377" EndTime="2105.843" Participant="me011">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2105.742" EndTime="2106.297" Participant="fe008">
      Yeah.
    </Segment>
    <Segment StartTime="2105.843" EndTime="2108.440" Participant="me011">
      We can take - we can have a pool. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2107.495" EndTime="2110.021" Participant="fe016">
      Sometimes the changes are, um,
    </Segment>
    <Segment StartTime="2110.558" EndTime="2116.296" Participant="fe016">
      cases where the recognizer would get it wrong anyway, cuz it was some word that we didn't have in the vocabulary, or -
    </Segment>
    <Segment StartTime="2113.233" EndTime="2114.910" Participant="me011">
      That's just what I was thinking.
    </Segment>
    <Segment StartTime="2113.312" EndTime="2114.696" Participant="mn017">
      <VocalSound Description="loud inbreath"/>
    </Segment>
    <Segment StartTime="2115.130" EndTime="2117.330" Participant="mn017">
      So, I - I - The thing is, when -
    </Segment>
    <Segment StartTime="2116.014" EndTime="2116.960" Participant="me013">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="2116.593" EndTime="2120.170" Participant="fe016">
      But it <Emphasis> does </Emphasis> help to get the backchannels back in, and things like that.
    </Segment>
    <Segment StartTime="2117.900" EndTime="2118.610" Participant="mn017">
      Right.
    </Segment>
    <Segment StartTime="2119.147" EndTime="2119.443" Participant="mn017">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="2119.810" EndTime="2124.430" Participant="mn017">
      So, whenever the - Right now, the s- the scoring is based on segments.
    </Segment>
    <Segment StartTime="2124.430" EndTime="2125.040" Participant="mn017">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="2125.040" EndTime="2128.053" Participant="mn017">
      Um, which is not great because, for instance -
    </Segment>
    <Segment StartTime="2128.053" EndTime="2133.304" Participant="mn017">
      So - so, the - the <NonVocalSound Description="mike noise"/> other way to do the scoring is using a - a NIST format called
    </Segment>
    <Segment StartTime="2133.304" EndTime="2137.404" Participant="mn017">
      S_T_M. S- uh, Segment Time Marked, where -
    </Segment>
    <Segment StartTime="2135.276" EndTime="2136.850" Participant="me011">
      Yep. I - we know about it.
    </Segment>
    <Segment StartTime="2137.404" EndTime="2146.744" Participant="mn017">
      <VocalSound Description="mouth"/> So, um, I have to convert the, uh, transcripts into this format, and then the scoring program actually looks at the times. And,
    </Segment>
    <Segment StartTime="2138.295" EndTime="2139.831" Participant="me001">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="2145.360" EndTime="2158.108" Participant="me001">
      <NonVocalSound Description="mike noise"/> <Comment Description="readjusting mike"/>
    </Segment>
    <Segment StartTime="2146.744" EndTime="2153.460" Participant="mn017">
      uh, you know, it - You can have a different segmentation in your recognizer output than in your references, and it will still do the right thing.
    </Segment>
    <Segment StartTime="2153.460" EndTime="2157.400" Participant="mn017">
      So, that's what we need to basically, uh, to - Hmm?
    </Segment>
    <Segment StartTime="2155.470" EndTime="2156.117" Participant="me011">
      Tran-
    </Segment>
    <Segment StartTime="2156.650" EndTime="2159.120" Participant="me011">
      Transcriber will export S_T_M.
    </Segment>
    <Segment StartTime="2157.283" EndTime="2158.806" Participant="me013">
      <NonVocalSound Description="mike noise"/> <NonVocalSound Description="thump in background"/>
    </Segment>
    <Segment StartTime="2159.207" EndTime="2159.721" Participant="mn017">
      Oh.
    </Segment>
    <Segment StartTime="2159.409" EndTime="2160.518" Participant="me011">
      In case you care.
    </Segment>
    <Segment StartTime="2159.990" EndTime="2163.561" Participant="mn017">
      Well. But then there's other changes. So. I mean, there's other -
    </Segment>
    <Segment StartTime="2162.011" EndTime="2162.475" Participant="me011">
      O_K.
    </Segment>
    <Segment StartTime="2163.561" EndTime="2165.640" Participant="mn017">
      We - we strip away a lot of the
    </Segment>
    <Segment StartTime="2163.645" EndTime="2165.926" Participant="me013">
      <VocalSound Description="inbreath"/> <VocalSound Description="clears throat"/>
    </Segment>
    <Segment StartTime="2166.030" EndTime="2166.710" Participant="mn017">
      mark-up,
    </Segment>
    <Segment StartTime="2166.710" EndTime="2169.430" Participant="mn017">
      uh, in the transcripts, which,
    </Segment>
    <Segment StartTime="2169.430" EndTime="2173.110" Participant="mn017">
      you know, isn't relevant to the scoring of the speech recognition output. So <Pause/>
    </Segment>
    <Segment StartTime="2170.720" EndTime="2171.460" Participant="me011">
      Right.
    </Segment>
    <Segment StartTime="2172.618" EndTime="2173.270" Participant="fe016">
      Would that - ?
    </Segment>
    <Segment StartTime="2173.270" EndTime="2178.488" Participant="fe016">
      Doesn't that only change the scoring if a word has moved into a different segment? I mean, I don't
    </Segment>
    <Segment StartTime="2179.053" EndTime="2182.780" Participant="fe016">
      think that's - har- I hardly ever see that. I think most of them are pretty good.
    </Segment>
    <Segment StartTime="2183.700" EndTime="2188.339" Participant="mn017">
      <VocalSound Description="inbreath"/> No. I mean if - suppo- I - I assume you also changed some boundaries. Right?
    </Segment>
    <Segment StartTime="2187.370" EndTime="2188.077" Participant="fe008">
      Mmm. I did.
    </Segment>
    <Segment StartTime="2188.339" EndTime="2194.827" Participant="mn017">
      So, if we want to use new transcripts with a different segmentation, then we can't use them in the current way we do scoring.
    </Segment>
    <Segment StartTime="2195.600" EndTime="2198.450" Participant="mn017">
      We have to - m- m- switch to this o-
    </Segment>
    <Segment StartTime="2197.580" EndTime="2200.040" Participant="fe016">
      Oh, you mean you would need to re-run the recognition.
    </Segment>
    <Segment StartTime="2199.808" EndTime="2200.997" Participant="mn017">
      No. We have to r-
    </Segment>
    <Segment StartTime="2201.244" EndTime="2202.793" Participant="mn017">
      There's a difference -
    </Segment>
    <Segment StartTime="2201.358" EndTime="2205.780" Participant="fe016">
      I mean, <Emphasis> if </Emphasis> you re-ran the recognition, then you just run it - Oh, I see. If you wanna use the old -
    </Segment>
    <Segment StartTime="2203.365" EndTime="2205.007" Participant="mn017">
      Right. Right.
    </Segment>
    <Segment StartTime="2206.140" EndTime="2209.990" Participant="fe016">
      You can actually <Emphasis> never, </Emphasis> though, really infer what you would get with a different -
    </Segment>
    <Segment StartTime="2209.793" EndTime="2210.361" Participant="me013">
      <VocalSound Description="clears throat"/>
    </Segment>
    <Segment StartTime="2209.807" EndTime="2210.727" Participant="mn017">
      That's true.
    </Segment>
    <Segment StartTime="2210.300" EndTime="2213.800" Participant="fe016">
      You know? It's - it's probably more fair to re-run the whole th- re-run that.
    </Segment>
    <Segment StartTime="2214.200" EndTime="2215.530" Participant="fe016">
      Uh, in other words, it's not
    </Segment>
    <Segment StartTime="2215.990" EndTime="2218.600" Participant="fe016">
      really a scoring script problem. It's -
    </Segment>
    <Segment StartTime="2218.665" EndTime="2221.181" Participant="mn017">
      No. But if you w- just want to see what -
    </Segment>
    <Segment StartTime="2221.181" EndTime="2227.041" Participant="mn017">
      Like, suppose you fixed some typ- No, Jane fixed some typos and you wanna see what effect does that have on the
    </Segment>
    <Segment StartTime="2227.350" EndTime="2228.329" Participant="mn017">
      word error, then
    </Segment>
    <Segment StartTime="2228.654" EndTime="2229.500" Participant="me011">
      Right. But - but -
    </Segment>
    <Segment StartTime="2228.789" EndTime="2229.910" Participant="mn017">
      we can f- we c-
    </Segment>
    <Segment StartTime="2229.032" EndTime="2230.950" Participant="fe016">
      Y- Yeah. Yeah.
    </Segment>
    <Segment StartTime="2229.650" EndTime="2231.593" Participant="me011">
      If the segments change, that won't work.
    </Segment>
    <Segment StartTime="2231.830" EndTime="2237.270" Participant="fe016">
      You - It'll sort of work, but it's not <Emphasis> exactly </Emphasis> what you would maybe get from recognition. You never know. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2235.111" EndTime="2238.094" Participant="me011">
      Well, I mean, what happens if you break one segment into two?
    </Segment>
    <Segment StartTime="2238.429" EndTime="2239.200" Participant="fe016">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2238.684" EndTime="2242.030" Participant="me011">
      Suddenly the- don- they don't match at all and you can't line <Pronounce Pronunciation="&quot;em&quot;"> them  </Pronounce> up anymore.
    </Segment>
    <Segment StartTime="2241.749" EndTime="2246.599" Participant="mn017">
      No, but that's what the pr- Well, that's what I'm saying. You <Emphasis> can </Emphasis> line them up based on the times.
    </Segment>
    <Segment StartTime="2242.078" EndTime="2242.632" Participant="fe008">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="2242.105" EndTime="2243.131" Participant="fe016">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2243.131" EndTime="2245.080" Participant="fe016">
      Well, you c- You - you s-
    </Segment>
    <Segment StartTime="2245.080" EndTime="2246.208" Participant="fe016">
      You can -
    </Segment>
    <Segment StartTime="2246.208" EndTime="2247.190" Participant="fe016">
      Yeah.
    </Segment>
    <Segment StartTime="2246.599" EndTime="2250.374" Participant="mn017">
      So the scoring program with - if you give it an S_T_M reference file,
    </Segment>
    <Segment StartTime="2250.834" EndTime="2256.083" Participant="mn017">
      it will actually compare the words based on their time marks. So therefore, you <Emphasis> can, </Emphasis>
    </Segment>
    <Segment StartTime="2253.662" EndTime="2254.370" Participant="me011">
      Right.
    </Segment>
    <Segment StartTime="2254.458" EndTime="2254.831" Participant="fe016">
      Yeah.
    </Segment>
    <Segment StartTime="2256.588" EndTime="2257.155" Participant="mn017">
      um -
    </Segment>
    <Segment StartTime="2257.024" EndTime="2259.758" Participant="me011">
      Does S_T_M do it per word, or por- per utterance?
    </Segment>
    <Segment StartTime="2259.440" EndTime="2266.727" Participant="mn017">
      It's per <Emphasis> utterance, </Emphasis> but it - it allows - As long as you hypothesize the word in the right segment in the reference, it gives you
    </Segment>
    <Segment StartTime="2260.525" EndTime="2261.307" Participant="me011">
      Th- what I thought.
    </Segment>
    <Segment StartTime="2265.473" EndTime="2266.178" Participant="me011">
      Alright.
    </Segment>
    <Segment StartTime="2267.010" EndTime="2268.990" Participant="mn017">
      credit for that. So it does a - <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2268.963" EndTime="2269.490" Participant="fe016">
      Mm-hmm.
    </Segment>
    <Segment StartTime="2268.990" EndTime="2277.436" Participant="mn017">
      it does a <Emphasis> word </Emphasis> alignment, like you have to do for scoring, but it constrains the words to lie within the t- the <Emphasis> time </Emphasis> bins of the reference. <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2271.431" EndTime="2273.120" Participant="me011">
      And then it does also some -
    </Segment>
    <Segment StartTime="2275.088" EndTime="2276.474" Participant="me011">
      Within the segment. Yep.
    </Segment>
    <Segment StartTime="2276.773" EndTime="2277.414" Participant="me011">
      I see.
    </Segment>
    <Segment StartTime="2277.985" EndTime="2280.750" Participant="mn017">
      And - fo- for you to get a credit for it. So.
    </Segment>
    <Segment StartTime="2281.370" EndTime="2282.131" Participant="fe008">
      That sounds great.
    </Segment>
    <Segment StartTime="2282.175" EndTime="2287.600" Participant="mn017">
      So - so, it's - it should be just a straightforward re-formatting issue of the references. So.
    </Segment>
    <Segment StartTime="2288.430" EndTime="2290.944" Participant="fe008">
      Yeah. I mean, t- y- I was thinking, th- the other day, that this
    </Segment>
    <Segment StartTime="2290.944" EndTime="2291.470" Participant="fe008">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2291.470" EndTime="2296.554" Participant="fe008">
      is not just conversational speech. The fact that is has so much technological jargon in it. I- it makes it -
    </Segment>
    <Segment StartTime="2296.554" EndTime="2297.079" Participant="fe008">
      u-
    </Segment>
    <Segment StartTime="2297.079" EndTime="2299.390" Participant="fe008">
      considerably harder to, um -
    </Segment>
    <Segment StartTime="2299.783" EndTime="2300.660" Participant="fe008">
      me- e- e- e-
    </Segment>
    <Segment StartTime="2300.660" EndTime="2304.876" Participant="fe008">
      to transcribe and to - to double check and <Emphasis> all </Emphasis> those things. So, <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2304.876" EndTime="2309.213" Participant="fe008">
      I think you're gonna find a substantial gain in terms of the word ac- accuracy.
    </Segment>
    <Segment StartTime="2309.213" EndTime="2312.525" Participant="fe008">
      So long as those words are in your <Emphasis> vocabulary. </Emphasis> <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2311.260" EndTime="2315.590" Participant="fe016">
      Well, it definitely helps with, um, forced alignment, <Emphasis> too, </Emphasis> because,
    </Segment>
    <Segment StartTime="2312.174" EndTime="2314.000" Participant="me011">
      One percent. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2312.974" EndTime="2314.290" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2314.046" EndTime="2314.758" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="2315.929" EndTime="2323.930" Participant="fe016">
      you know, whe- when we know the - the - the true words and we're adding them to the vocabulary and training a language model and so forth for future meetings, especially the -
    </Segment>
    <Segment StartTime="2324.215" EndTime="2327.338" Participant="fe016">
      like, the front-end meetings or meetings with a lot of <Pause/> jargon in them, that
    </Segment>
    <Segment StartTime="2327.823" EndTime="2330.240" Participant="fe016">
      is not represented in Switchboard or
    </Segment>
    <Segment StartTime="2330.240" EndTime="2331.082" Participant="fe016">
      CallHome, then it -
    </Segment>
    <Segment StartTime="2331.468" EndTime="2335.788" Participant="fe008">
      Well, <Emphasis> all </Emphasis> of them are. All of our - all of our meetings have a lot of jargon in them. But we know what those <Emphasis> words </Emphasis> are.
    </Segment>
    <Segment StartTime="2331.526" EndTime="2333.340" Participant="fe016">
      then it really helps. So <Pause/>
    </Segment>
    <Segment StartTime="2334.283" EndTime="2335.260" Participant="fe016">
      Yeah.
    </Segment>
    <Segment StartTime="2335.788" EndTime="2336.638" Participant="fe008">
      I mean, I don't know
    </Segment>
    <Segment StartTime="2337.123" EndTime="2339.311" Participant="fe008">
      how you get them <Emphasis> into </Emphasis> the vocabulary, but it would seem <Uncertain NumSyllables="1"> @@ </Uncertain> -
    </Segment>
    <Segment StartTime="2339.311" EndTime="2345.520" Participant="fe008">
      Now, I have to - I have to <VocalSound Description="laugh"/> - I really need to, uh, raise a question about the term "cheating". <VocalSound Description="laugh"/> O_K?
    </Segment>
    <Segment StartTime="2345.520" EndTime="2348.489" Participant="fe008">
      And the reason is, um, if I understand that,
    </Segment>
    <Segment StartTime="2349.040" EndTime="2355.281" Participant="fe008">
      "cheating" is a term which is used to apply for basically what all of linguistics - corpus linguistics does, and what -
    </Segment>
    <Segment StartTime="2355.281" EndTime="2361.463" Participant="fe008">
      what my transcribers are doing and what <Emphasis> I </Emphasis> do, <VocalSound Description="laugh"/> which is a methodology whereby you actually physically mark
    </Segment>
    <Segment StartTime="2361.436" EndTime="2362.650" Participant="me011">
      <VocalSound Description="mouth"/> All - al-
    </Segment>
    <Segment StartTime="2361.463" EndTime="2366.120" Participant="fe008">
      things <Emphasis> in </Emphasis> the data. Like the transcription, like the words that are actually there.
    </Segment>
    <Segment StartTime="2364.610" EndTime="2368.161" Participant="me011">
      All we mean by that is that we're giving the recognizer more information
    </Segment>
    <Segment StartTime="2368.640" EndTime="2369.521" Participant="me011">
      than it would
    </Segment>
    <Segment StartTime="2369.521" EndTime="2371.405" Participant="me011">
      have if you were running it raw,
    </Segment>
    <Segment StartTime="2371.502" EndTime="2373.140" Participant="fe008">
      <VocalSound Description="inbreath"/> <Pronounce Pronunciation="&quot;yeah-mmm&quot;"> Yeah.  </Pronounce>
    </Segment>
    <Segment StartTime="2371.643" EndTime="2374.773" Participant="me011">
      over a meeting that no person has ever listened to or transcribed.
    </Segment>
    <Segment StartTime="2374.350" EndTime="2379.642" Participant="fe008">
      O_K. I mean, that - that part is O_K. But I - but I <Emphasis> do </Emphasis> wonder sometimes if it might be possible to use
    </Segment>
    <Segment StartTime="2379.642" EndTime="2384.965" Participant="fe008">
      <VocalSound Description="inbreath"/> a term that's a little bit less evaluative, like, um, "hand-marked". <Comment Description="while smiling"/> <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2380.317" EndTime="2382.231" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2381.480" EndTime="2386.500" Participant="me013">
      <VocalSound Description="inbreath"/> Uh, but cheating is - is - p- Sure. But chea- but cheating is - is - is -
    </Segment>
    <Segment StartTime="2383.922" EndTime="2384.567" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2387.307" EndTime="2390.641" Participant="me013">
      pretty commonly used to mean this. It really is - <VocalSound Description="clears throat"/>
    </Segment>
    <Segment StartTime="2387.362" EndTime="2389.218" Participant="mn017">
      It's a - it's a technical term.
    </Segment>
    <Segment StartTime="2388.067" EndTime="2391.700" Participant="fe016">
      It - I guess it - It's been around for a long time, and it -
    </Segment>
    <Segment StartTime="2388.917" EndTime="2391.054" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2388.960" EndTime="2390.150" Participant="me011">
      It is. Yep.
    </Segment>
    <Segment StartTime="2389.215" EndTime="2392.141" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2391.270" EndTime="2391.673" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="2391.700" EndTime="2393.804" Participant="fe016">
      it's sort of, um - Because it's so
    </Segment>
    <Segment StartTime="2393.804" EndTime="2400.980" Participant="fe016">
      strong a word, people don't take it that seriously, in - It's not negatively viewed. It just really means -
    </Segment>
    <Segment StartTime="2394.437" EndTime="2396.395" Participant="me001">
      <VocalSound Description="loud outbreath"/>
    </Segment>
    <Segment StartTime="2395.033" EndTime="2402.159" Participant="mn017">
      <NonVocalSound Description="mike noise"/> <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2396.282" EndTime="2399.138" Participant="me013">
      <VocalSound Description="inbreath"/> Yeah. I- i- i- i- <Pause/>
    </Segment>
    <Segment StartTime="2399.138" EndTime="2400.300" Participant="me013">
      I- it's - i-
    </Segment>
    <Segment StartTime="2400.300" EndTime="2403.700" Participant="me013">
      It's even more than that. I think it - it <Emphasis> really </Emphasis> gives a very strong
    </Segment>
    <Segment StartTime="2404.027" EndTime="2406.002" Participant="me013">
      perspective, that <Emphasis> you </Emphasis> <Emphasis> know </Emphasis>
    </Segment>
    <Segment StartTime="2406.440" EndTime="2410.330" Participant="me013">
      that what you were doing is not an un-biased experiment.
    </Segment>
    <Segment StartTime="2410.890" EndTime="2416.899" Participant="me013">
      If you <Emphasis> don't </Emphasis> say that, then people think "oh, they did that and they threw that, but that doesn't represent what would happen in the real world".
    </Segment>
    <Segment StartTime="2416.899" EndTime="2417.345" Participant="me013">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2417.345" EndTime="2421.844" Participant="me013">
      But if you say "we did a n- cheating experiment", which is really the standard way you'd say it,
    </Segment>
    <Segment StartTime="2421.844" EndTime="2427.720" Participant="me013">
      it says you <Emphasis> deliberately </Emphasis> put in a piece of me- information that you would <Emphasis> not </Emphasis> have in the real world so that you can learn something.
    </Segment>
    <Segment StartTime="2428.200" EndTime="2432.202" Participant="me013">
      Just as part - as your process. So it's - I d- I don't know, I - <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2430.850" EndTime="2433.473" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="2431.840" EndTime="2438.425" Participant="fe008">
      O_K. I guess what I'm - I'm thinking, is just in - if - uh, when these are presented in an inte- interdisciplinary context,
    </Segment>
    <Segment StartTime="2432.202" EndTime="2433.420" Participant="me013">
      i- <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="2433.473" EndTime="2436.893" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2438.425" EndTime="2444.051" Participant="fe008">
      it might be nice to add that explanation? Cuz otherwise it sounds like a pejorative statement on an alternative methodology?
    </Segment>
    <Segment StartTime="2440.563" EndTime="2442.776" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="2444.341" EndTime="2445.435" Participant="fe008">
      Cuz it sounds like -
    </Segment>
    <Segment StartTime="2445.730" EndTime="2447.579" Participant="fe008">
      it sort of devalues the -
    </Segment>
    <Segment StartTime="2447.579" EndTime="2448.585" Participant="fe008">
      the, um,
    </Segment>
    <Segment StartTime="2448.585" EndTime="2451.786" Participant="fe008">
      all - other approach, which is to - to put those distinctions <Emphasis> in. </Emphasis>
    </Segment>
    <Segment StartTime="2448.830" EndTime="2449.740" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="2450.699" EndTime="2454.030" Participant="mn017">
      <VocalSound Description="breath-laugh"/> <Comment Description="very quietly"/>
    </Segment>
    <Segment StartTime="2451.470" EndTime="2454.524" Participant="me013">
      Maybe. I - I've heard this at I_C_S_L_P for years and years and years, though.
    </Segment>
    <Segment StartTime="2454.524" EndTime="2456.998" Participant="me013">
      I mean, it's - it's pretty - pretty common. People say this.
    </Segment>
    <Segment StartTime="2454.760" EndTime="2463.497" Participant="mn017">
      If - i- Well, you i- i- in a - To a broader audience you could call it a diagnostic experiment, rather than a cheating experiment. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2454.993" EndTime="2456.587" Participant="fe016">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2460.311" EndTime="2461.434" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2460.668" EndTime="2461.314" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2460.692" EndTime="2469.910" Participant="fe016">
      Actually, Jane is right. Like, in conversation analysis, I've never heard people <Emphasis> use </Emphasis> this, because they're not using an automatic system. So, it really -
    </Segment>
    <Segment StartTime="2468.380" EndTime="2469.680" Participant="mn017">
      But they don't run experiments.
    </Segment>
    <Segment StartTime="2468.481" EndTime="2469.056" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="2469.910" EndTime="2480.987" Participant="fe016">
      Right. I'm j- Well, you can do experiments, but the - it's cheating relative to what we call a system where we can completely control this black box. And it's not a very
    </Segment>
    <Segment StartTime="2469.940" EndTime="2470.970" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2470.098" EndTime="2480.034" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2470.110" EndTime="2474.108" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2470.151" EndTime="2470.388" Participant="fe008">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="2470.154" EndTime="2473.591" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2470.640" EndTime="2471.540" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2470.922" EndTime="2472.148" Participant="fe008">
      They can do experiments.
    </Segment>
    <Segment StartTime="2472.699" EndTime="2473.669" Participant="me051">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="2473.825" EndTime="2474.518" Participant="me013">
      <VocalSound Description="cough"/>
    </Segment>
    <Segment StartTime="2480.987" EndTime="2483.720" Participant="fe016">
      <Emphasis> smart </Emphasis> system. It - it only knows what we <Emphasis> give </Emphasis> it.
    </Segment>
    <Segment StartTime="2483.720" EndTime="2488.657" Participant="fe016">
      And if w- If it knows <Emphasis> more </Emphasis> than we would really give it when it runs, we call it cheating. But it's -
    </Segment>
    <Segment StartTime="2488.657" EndTime="2491.610" Participant="fe016">
      Yeah, it's f- only used in a community that
    </Segment>
    <Segment StartTime="2492.310" EndTime="2501.165" Participant="fe016">
      does some type of computational modeling, I think. It's really not used in any kind of community doing experiments on human perception, or -
    </Segment>
    <Segment StartTime="2494.370" EndTime="2496.622" Participant="me013">
      Yeah. In pattern recognition, or in -
    </Segment>
    <Segment StartTime="2496.998" EndTime="2497.475" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="2501.042" EndTime="2501.564" Participant="me013">
      Right.
    </Segment>
    <Segment StartTime="2501.165" EndTime="2503.610" Participant="fe016">
      You know. So <Pause/> she certainly can
    </Segment>
    <Segment StartTime="2502.160" EndTime="2508.195" Participant="me013">
      Yeah. It's machine - machine experiments. Um, <VocalSound Description="clears throat"/> I mean, w- when - when the, uh, the neural net
    </Segment>
    <Segment StartTime="2504.212" EndTime="2505.160" Participant="me001">
      <VocalSound Description="clears throat"/>
    </Segment>
    <Segment StartTime="2504.890" EndTime="2505.970" Participant="fe016">
      define it.
    </Segment>
    <Segment StartTime="2508.195" EndTime="2511.192" Participant="me013">
      w- uh - wave hit in the mid-eighties,
    </Segment>
    <Segment StartTime="2511.580" EndTime="2516.764" Participant="me013">
      and by the late eighties there were - uh, we were reviewing <Emphasis> thousands </Emphasis> of papers that were
    </Segment>
    <Segment StartTime="2516.430" EndTime="2517.310" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2516.764" EndTime="2520.305" Participant="me013">
      <VocalSound Description="laugh"/> coming out in neural nets. It was really hot. Everybody thought it would do everything.
    </Segment>
    <Segment StartTime="2520.305" EndTime="2520.950" Participant="me013">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2520.950" EndTime="2526.753" Participant="me013">
      Um, and a really common error that people were making was, they were just reporting
    </Segment>
    <Segment StartTime="2526.753" EndTime="2531.172" Participant="me013">
      their, uh - uh, classification results on the data that they were training on.
    </Segment>
    <Segment StartTime="2531.460" EndTime="2533.580" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2531.709" EndTime="2533.800" Participant="me013">
      And so I think it was - it was -
    </Segment>
    <Segment StartTime="2533.120" EndTime="2534.020" Participant="mn017">
      <VocalSound Description="laugh-breath"/>
    </Segment>
    <Segment StartTime="2533.800" EndTime="2540.561" Participant="me013">
      it was very important for people then who were doing something <Emphasis> diagnostic </Emphasis> to say "hey, I <Emphasis> know </Emphasis> I'm doing something that isn't kosher".
    </Segment>
    <Segment StartTime="2541.142" EndTime="2541.724" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="2541.330" EndTime="2546.063" Participant="me013">
      And to make it <Emphasis> really </Emphasis> clear that they knew it. So that was, I think, why - why it became a popular term.
    </Segment>
    <Segment StartTime="2545.460" EndTime="2551.409" Participant="fe008">
      <VocalSound Description="inbreath"/> Yeah, it just seems like, you know, i- if it were <Emphasis> "bootstrapping </Emphasis>", or if it were - I mean, there are other ways to maybe get the point across.
    </Segment>
    <Segment StartTime="2550.700" EndTime="2553.430" Participant="me013">
      But it <Emphasis> isn't </Emphasis> bootstrapping. Uh, it's - it's <Emphasis> really </Emphasis> <Emphasis> cheating. </Emphasis>
    </Segment>
    <Segment StartTime="2551.137" EndTime="2560.710" Participant="mn017">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="2552.889" EndTime="2556.759" Participant="me011">
      It's us- it's using information you wouldn't normally have. So.
    </Segment>
    <Segment StartTime="2554.130" EndTime="2555.885" Participant="me013">
      It's just ng- Yeah. Yeah. <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2554.975" EndTime="2557.390" Participant="fe008">
      <VocalSound Description="inbreath"/> Hu- Well. O_K.
    </Segment>
    <Segment StartTime="2555.885" EndTime="2560.631" Participant="me013">
      But it's cheating in a way that's pril- It's - it's announcing to everybody "hey, I'm cheating, by doing this."
    </Segment>
    <Segment StartTime="2560.710" EndTime="2562.730" Participant="mn017">
      <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="2561.100" EndTime="2564.453" Participant="me013">
      It's saying - so it's all above the table. I'm actually using this other thing.
    </Segment>
    <Segment StartTime="2564.630" EndTime="2565.199" Participant="fe008">
      O_K.
    </Segment>
    <Segment StartTime="2564.735" EndTime="2566.316" Participant="me013">
      That's - that's - that's the reason. <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2565.490" EndTime="2566.358" Participant="fe008">
      O_K. Alright.
    </Segment>
    <Segment StartTime="2566.316" EndTime="2569.783" Participant="me013">
      Bootstrapping would imply it was actually <Emphasis> legitimate </Emphasis> in some kind of way.
    </Segment>
    <Segment StartTime="2568.880" EndTime="2569.391" Participant="me011">
      Hmm.
    </Segment>
    <Segment StartTime="2569.640" EndTime="2570.730" Participant="fe008">
      <VocalSound Description="mouth"/> Well, I - I think -
    </Segment>
    <Segment StartTime="2570.730" EndTime="2571.996" Participant="me013">
      But - <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="2571.290" EndTime="2574.678" Participant="me011">
      We're not de-legitimizing the <Emphasis> data. </Emphasis> We're de-legitimizing the <Emphasis> experiment. </Emphasis>
    </Segment>
    <Segment StartTime="2571.300" EndTime="2572.167" Participant="fe008">
      <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="2575.213" EndTime="2575.866" Participant="fe008">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2575.432" EndTime="2580.492" Participant="me011">
      We're not saying that the <Emphasis> data </Emphasis> is cheating <Emphasis> data. </Emphasis> We're saying th- we are <Emphasis> cheating </Emphasis> by <Emphasis> using </Emphasis> this data.
    </Segment>
    <Segment StartTime="2578.389" EndTime="2578.885" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="2580.669" EndTime="2581.325" Participant="fe008">
      O_K.
    </Segment>
    <Segment StartTime="2581.190" EndTime="2581.604" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="2581.247" EndTime="2583.582" Participant="me011">
      Because normally you wouldn't have that data available.
    </Segment>
    <Segment StartTime="2581.660" EndTime="2582.096" Participant="fe008">
      O_K.
    </Segment>
    <Segment StartTime="2583.790" EndTime="2584.302" Participant="fe008">
      O_K.
    </Segment>
    <Segment StartTime="2583.920" EndTime="2584.441" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="2585.350" EndTime="2589.157" Participant="fe008">
      <VocalSound Description="inbreath"/> Does seem to me that it - that it carries over some baggage with it. That, uh,
    </Segment>
    <Segment StartTime="2585.391" EndTime="2586.067" Participant="me013">
      Anyway.
    </Segment>
    <Segment StartTime="2589.157" EndTime="2589.658" Participant="fe008">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2589.658" EndTime="2592.154" Participant="fe008">
      i- that - I can understand it in context as you described it.
    </Segment>
    <Segment StartTime="2592.154" EndTime="2594.796" Participant="fe008">
      <VocalSound Description="inbreath"/> But, it - it seems to me that -
    </Segment>
    <Segment StartTime="2594.796" EndTime="2598.925" Participant="fe008">
      u- um, it does import some negative evaluation, that,
    </Segment>
    <Segment StartTime="2599.264" EndTime="2603.063" Participant="fe008">
      um, would maybe <Emphasis> not </Emphasis> be good if - It has shock value, and
    </Segment>
    <Segment StartTime="2600.452" EndTime="2601.844" Participant="me013">
      It has some shock value.
    </Segment>
    <Segment StartTime="2602.919" EndTime="2604.380" Participant="me013">
      Which is part of why it's used.
    </Segment>
    <Segment StartTime="2603.063" EndTime="2608.930" Participant="fe008">
      to the wr- to the <Emphasis> wrong </Emphasis> audience, I think that that might be t- <VocalSound Description="laugh"/> a negative shock. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2604.637" EndTime="2609.466" Participant="fe016">
      <VocalSound Description="mouth"/> Yeah. I thi- I think to the wrong audience, I agree that, um, p-
    </Segment>
    <Segment StartTime="2606.387" EndTime="2606.982" Participant="me013">
      Uh-huh.
    </Segment>
    <Segment StartTime="2607.461" EndTime="2609.810" Participant="mn017">
      <VocalSound Description="mouth"/> It's like a disclaimer.
    </Segment>
    <Segment StartTime="2609.910" EndTime="2614.790" Participant="fe016">
      to the wrong audie- we should just explain what it - what it means here, and that it's a common term, and
    </Segment>
    <Segment StartTime="2612.800" EndTime="2617.027" Participant="fe008">
      Yeah. We started with hand-marked data, or with, uh, t- you know, hand-transcribed data, or - <Comment Description="trailing off"/>
    </Segment>
    <Segment StartTime="2615.570" EndTime="2619.517" Participant="mn017">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="2615.720" EndTime="2616.550" Participant="fe016">
      we're -
    </Segment>
    <Segment StartTime="2617.083" EndTime="2617.631" Participant="me011">
      Shoot. <Comment Description="interjection"/> <Comment Description="after potential damage to the mike"/>
    </Segment>
    <Segment StartTime="2617.898" EndTime="2618.833" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2618.040" EndTime="2619.747" Participant="fe008">
      Yeah. Did you break the <Emphasis> microphone? </Emphasis>
    </Segment>
    <Segment StartTime="2618.596" EndTime="2623.964" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2618.605" EndTime="2619.063" Participant="fe016">
      Yeah.
    </Segment>
    <Segment StartTime="2619.908" EndTime="2622.020" Participant="me011">
      Just the clip. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2620.720" EndTime="2624.717" Participant="fe008">
      Nnn. O_K. <VocalSound Description="laugh"/> O_K. Well, I feel better now. Thank you very much.
    </Segment>
    <Segment StartTime="2621.150" EndTime="2622.500" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2621.882" EndTime="2622.650" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2622.047" EndTime="2625.200" Participant="mn017">
      Now we have to l- delete that expletive. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2624.117" EndTime="2631.022" Participant="fe016">
      <Uncertain> Hmm. </Uncertain> The first time I heard that, I - I thought, you know, the same thing and I guess you ju- after a while it becomes
    </Segment>
    <Segment StartTime="2624.581" EndTime="2629.271" Participant="me001">
      <VocalSound Description="laugh"/> Thank you, Andreas. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2624.717" EndTime="2627.640" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2625.200" EndTime="2630.226" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2625.941" EndTime="2629.599" Participant="me018">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2626.318" EndTime="2630.189" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2630.811" EndTime="2632.986" Participant="me011">
      It's p- it - it really is part of the jargon.
    </Segment>
    <Segment StartTime="2631.022" EndTime="2634.527" Participant="fe016">
      almost a - It's a bu- bit of a -
    </Segment>
    <Segment StartTime="2633.586" EndTime="2635.054" Participant="me013">
      <VocalSound Description="breath"/> But it's also is -
    </Segment>
    <Segment StartTime="2634.527" EndTime="2643.430" Participant="fe016">
      a <Emphasis> humbling </Emphasis> thing when somebody says that. If they get good results but we were cheating on this feature because we took it for granted even though we <Pause/> can't really assume that,
    </Segment>
    <Segment StartTime="2635.047" EndTime="2640.246" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2636.620" EndTime="2637.000" Participant="me013">
      Right.
    </Segment>
    <Segment StartTime="2643.430" EndTime="2645.540" Participant="fe016">
      then it's actually the opposite. It's -
    </Segment>
    <Segment StartTime="2644.910" EndTime="2648.770" Participant="fe008">
      <VocalSound Description="inbreath"/> The <Emphasis> trouble </Emphasis> is, that - you know, I understand it in that context, but it - but it
    </Segment>
    <Segment StartTime="2648.770" EndTime="2650.729" Participant="fe008">
      is almost resentful. It's almost like,
    </Segment>
    <Segment StartTime="2650.729" EndTime="2653.821" Participant="fe008">
      you know, resentful of the data, resentful of the, um,
    </Segment>
    <Segment StartTime="2650.784" EndTime="2651.959" Participant="mn017">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="2654.361" EndTime="2659.524" Participant="fe008">
      the <Emphasis> hard </Emphasis> work that's going into preparing the data is - <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2654.699" EndTime="2655.392" Participant="me011">
      <VocalSound Description="mouth"/> Well -
    </Segment>
    <Segment StartTime="2656.506" EndTime="2657.212" Participant="me011">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="2657.380" EndTime="2661.128" Participant="me013">
      No, no. It has - I don't think you're saying the <Emphasis> data </Emphasis> is cheating. I think you're saying
    </Segment>
    <Segment StartTime="2657.790" EndTime="2659.370" Participant="fe016">
      Well, it's - <Uncertain> it's </Uncertain> -
    </Segment>
    <Segment StartTime="2660.190" EndTime="2661.440" Participant="me011">
      That's what I was saying.
    </Segment>
    <Segment StartTime="2660.428" EndTime="2661.968" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2660.779" EndTime="2661.363" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="2661.128" EndTime="2665.818" Participant="me013">
      "in my experiment, I cheated in this way". I mean, another is what Liz was talking about, how
    </Segment>
    <Segment StartTime="2664.373" EndTime="2665.424" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2666.380" EndTime="2669.618" Participant="me013">
      in the <Emphasis> Switchboard </Emphasis> tests - in all the Switchboard tests we've been g- doing,
    </Segment>
    <Segment StartTime="2669.618" EndTime="2674.301" Participant="me013">
      we've been making the same er- uh, same - uh, uh, standard -
    </Segment>
    <Segment StartTime="2674.646" EndTime="2677.834" Participant="me013">
      using the same standard way of - of getting the data to test on.
    </Segment>
    <Segment StartTime="2676.654" EndTime="2677.036" Participant="me011">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="2678.300" EndTime="2683.507" Participant="me013">
      Which means that we weren't actually running it on - on, uh, on data that had - had no speech.
    </Segment>
    <Segment StartTime="2683.940" EndTime="2685.608" Participant="me013">
      And, in a sense, that was cheating.
    </Segment>
    <Segment StartTime="2685.379" EndTime="2686.863" Participant="mn014">
      <NonVocalSound Description="mike noise"/> <VocalSound Description="sniff"/>
    </Segment>
    <Segment StartTime="2685.610" EndTime="2686.080" Participant="fe008">
      O_K.
    </Segment>
    <Segment StartTime="2686.210" EndTime="2690.349" Participant="me013">
      So i- so it - it's - it's a good wake-up call if people "well,
    </Segment>
    <Segment StartTime="2690.349" EndTime="2694.167" Participant="me013">
      we have this performance but you have to keep in mind, we're doing - we're - we're not
    </Segment>
    <Segment StartTime="2695.000" EndTime="2698.143" Participant="me013">
      doing the whole real task in this way and this way and this way. <Emphasis> But, </Emphasis>
    </Segment>
    <Segment StartTime="2698.466" EndTime="2703.812" Participant="me013">
      I'll convince you that - that it's still important for you to listen to what I have to say next because of this and this."
    </Segment>
    <Segment StartTime="2703.230" EndTime="2704.268" Participant="fe008">
      O_K. O_K.
    </Segment>
    <Segment StartTime="2703.617" EndTime="2705.016" Participant="fe016">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="2703.812" EndTime="2707.630" Participant="me013">
      And so it's just a way of putting it all out on the table, uh, as opposed to -
    </Segment>
    <Segment StartTime="2706.230" EndTime="2712.581" Participant="me011">
      Uh, and it's used for a lot of different types of data. So, segment- whether you have segmentation or not, is it male or female or not -
    </Segment>
    <Segment StartTime="2712.581" EndTime="2712.842" Participant="me011">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2712.752" EndTime="2713.189" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="2712.842" EndTime="2716.458" Participant="me011">
      Um, do you know the signal-to-noise? Like, that's <Emphasis> another </Emphasis> one I see all the time, where
    </Segment>
    <Segment StartTime="2713.774" EndTime="2714.359" Participant="fe008">
      Mmm.
    </Segment>
    <Segment StartTime="2715.580" EndTime="2716.621" Participant="me013">
      Yeah. Sure.
    </Segment>
    <Segment StartTime="2715.779" EndTime="2716.906" Participant="mn017">
      <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="2716.458" EndTime="2716.651" Participant="me011">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2716.651" EndTime="2717.744" Participant="me011">
      you assume it's known.
    </Segment>
    <Segment StartTime="2717.800" EndTime="2718.439" Participant="fe008">
      Oh, interesting.
    </Segment>
    <Segment StartTime="2718.010" EndTime="2718.424" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="2718.136" EndTime="2721.150" Participant="me011">
      And you say it's <Emphasis> cheating </Emphasis> because <Pause/> you don't actually compute it.
    </Segment>
    <Segment StartTime="2718.316" EndTime="2718.930" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="2718.636" EndTime="2719.790" Participant="fe008">
      I didn't know that. Huh!
    </Segment>
    <Segment StartTime="2719.015" EndTime="2719.391" Participant="me013">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="2719.738" EndTime="2720.222" Participant="me013">
      In -
    </Segment>
    <Segment StartTime="2721.052" EndTime="2727.215" Participant="me013">
      In the multi-band experiments, in the first one, we really wanted to find out, "what if you <Emphasis> knew </Emphasis> which band was really noisy?"
    </Segment>
    <Segment StartTime="2721.425" EndTime="2722.300" Participant="fe008">
      <VocalSound Description="mouth"/> Uh-huh!
    </Segment>
    <Segment StartTime="2726.415" EndTime="2727.553" Participant="me011">
      N- Right.
    </Segment>
    <Segment StartTime="2727.732" EndTime="2729.564" Participant="me013">
      I mean, suppose you just <Emphasis> know </Emphasis> that.
    </Segment>
    <Segment StartTime="2728.720" EndTime="2729.500" Participant="me011">
      Same cheating.
    </Segment>
    <Segment StartTime="2730.130" EndTime="2735.909" Participant="me013">
      And then - then, even if you know that, can that help you? I mean, what can - what strategy can you do, to do well
    </Segment>
    <Segment StartTime="2733.257" EndTime="2734.160" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="2735.909" EndTime="2738.332" Participant="me013">
      without that particular band in the spectrum? <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2738.332" EndTime="2740.380" Participant="me013">
      And so that was important to <Emphasis> know </Emphasis> as a baseline.
    </Segment>
    <Segment StartTime="2740.380" EndTime="2744.734" Participant="me013">
      And - and then, once you knew that, then s- you go "well, now how do I know that that's noisy?"
    </Segment>
    <Segment StartTime="2744.477" EndTime="2745.590" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="2744.920" EndTime="2745.673" Participant="fe008">
      O_K.
    </Segment>
    <Segment StartTime="2745.530" EndTime="2753.130" Participant="me013">
      And - and - and, uh - But, y- you know, I - I should also say that there's a <Emphasis> lot </Emphasis> of - It's not just the word "cheating". But there's <Emphasis> lots </Emphasis> of other things that we talk about, <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2753.130" EndTime="2756.209" Participant="me013">
      which, as soon as you go outside of your narrow little group,
    </Segment>
    <Segment StartTime="2756.209" EndTime="2758.149" Participant="me013">
      it gets very, very confusing to people.
    </Segment>
    <Segment StartTime="2757.796" EndTime="2761.065" Participant="fe008">
      <VocalSound Description="inbreath"/> Oh, sure. Terminology is always <Pronounce Pronunciation="&quot;consext&quot;"> context  </Pronounce> dependent. No question about it.
    </Segment>
    <Segment StartTime="2757.890" EndTime="2758.437" Participant="me011">
      Mm-hmm.
    </Segment>
    <Segment StartTime="2758.149" EndTime="2761.933" Participant="me013">
      And - I- i- i- i- <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="2761.065" EndTime="2765.690" Participant="fe008">
      It's just <Pause/> that it seems like <Pause/> the - the point without the negative evaluation would be -
    </Segment>
    <Segment StartTime="2765.690" EndTime="2771.978" Participant="fe008">
      However, you know, if- i- I understand your point. That it - th- it has a long tradition in this field - I didn't realize. And is used - This is interesting what -
    </Segment>
    <Segment StartTime="2768.084" EndTime="2768.500" Participant="me013">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2768.386" EndTime="2772.133" Participant="mn017">
      <VocalSound Description="inbreath"/> Uh <Pause/> <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="2768.685" EndTime="2770.470" Participant="me011">
      It's not even really negative.
    </Segment>
    <Segment StartTime="2771.978" EndTime="2773.859" Participant="fe008">
      what Adam said about it being used also for
    </Segment>
    <Segment StartTime="2773.859" EndTime="2775.550" Participant="fe008">
      a bunch of other, uh, dimensions.
    </Segment>
    <Segment StartTime="2774.770" EndTime="2781.406" Participant="me013">
      Well, the example I was thinking of, also, was this w- thing that - that Herve and - and Hynek and I made about, uh, increasing the error rate. <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2781.406" EndTime="2788.056" Participant="me013">
      And, so we - we did a number of papers, and talks, and so forth about, uh - w- uh, i- the - the virtues of increasing the error rate.
    </Segment>
    <Segment StartTime="2788.042" EndTime="2790.325" Participant="fe008">
      Oh. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2788.056" EndTime="2788.702" Participant="me013">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2788.362" EndTime="2789.570" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2788.702" EndTime="2794.229" Participant="me013">
      O_K. And - I mean, and the whole <Emphasis> point </Emphasis> of it was, not that it was <Emphasis> good </Emphasis> to increase your error rate, but it was good to
    </Segment>
    <Segment StartTime="2790.950" EndTime="2793.320" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2794.633" EndTime="2795.899" Participant="me013">
      be <Emphasis> willing </Emphasis>
    </Segment>
    <Segment StartTime="2795.899" EndTime="2803.274" Participant="me013">
      to <Emphasis> risk </Emphasis> increasing your error rate by trying risky things, and trying - Because there's this notion of a local minimum, that if you just have
    </Segment>
    <Segment StartTime="2803.274" EndTime="2803.701" Participant="me013">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2803.701" EndTime="2806.415" Participant="me013">
      some system that's very complex, and you t- you -
    </Segment>
    <Segment StartTime="2804.550" EndTime="2808.110" Participant="mn017">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2806.415" EndTime="2811.995" Participant="me013">
      <VocalSound Description="breath"/> you turn some knobs to try to make it better and better, you'll never get out of this local minimum. You have to be willing
    </Segment>
    <Segment StartTime="2806.508" EndTime="2807.834" Participant="me011">
      Fine-tune little bits.
    </Segment>
    <Segment StartTime="2808.110" EndTime="2808.980" Participant="mn017">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2810.822" EndTime="2811.876" Participant="fe008">
      Mm-hmm. <Comment Description="very quietly"/>
    </Segment>
    <Segment StartTime="2810.943" EndTime="2812.001" Participant="mn017">
      Mmm.
    </Segment>
    <Segment StartTime="2811.995" EndTime="2812.403" Participant="me013">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2812.095" EndTime="2812.731" Participant="fe008">
      Interesting.
    </Segment>
    <Segment StartTime="2812.403" EndTime="2818.100" Participant="me013">
      to <Emphasis> jump </Emphasis> to something that's quite different. And the first time you jump to something quite different, or maybe the first ten times or a hundred times you do,
    </Segment>
    <Segment StartTime="2818.100" EndTime="2818.476" Participant="me013">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2818.476" EndTime="2821.484" Participant="me013">
      it's gonna be much much worse, because you've optimized the other system.
    </Segment>
    <Segment StartTime="2821.484" EndTime="2821.731" Participant="me013">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2821.731" EndTime="2825.267" Participant="me013">
      So the effect - the <Emphasis> immediate </Emphasis> effect is gonna be to increase your error rate. <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="2824.750" EndTime="2825.385" Participant="fe008">
      Interesting.
    </Segment>
    <Segment StartTime="2825.267" EndTime="2829.883" Participant="me013">
      And so it was - we had a couple papers like "Towards increasing the error rate in speech" and - and so on.
    </Segment>
    <Segment StartTime="2827.942" EndTime="2828.468" Participant="me011">
      Hmm.
    </Segment>
    <Segment StartTime="2828.396" EndTime="2829.243" Participant="fe008">
      <VocalSound Description="mouth"/> Huh!
    </Segment>
    <Segment StartTime="2828.643" EndTime="2829.652" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2829.539" EndTime="2830.550" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2829.883" EndTime="2836.254" Participant="me013">
      And we really did get feedback from a few people, some of whom were, you know, fairly senior, that - that, um,
    </Segment>
    <Segment StartTime="2836.623" EndTime="2842.066" Participant="me013">
      "Well, you know, you sh- you - you c- We're c- really concerned about you <Emphasis> misleading </Emphasis> people into thinking they should be increasing the error rate."
    </Segment>
    <Segment StartTime="2840.967" EndTime="2842.233" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="2841.482" EndTime="2844.690" Participant="fe008">
      Yeah. <Comment Description="while laughing"/> <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2842.066" EndTime="2846.296" Participant="me013">
      <VocalSound Description="breath"/> And, <VocalSound Description="laugh"/> uh, and we go - we thought "Oh, come <Emphasis> on. </Emphasis>" But. <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2842.310" EndTime="2843.664" Participant="mn014">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="2843.664" EndTime="2849.580" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2845.620" EndTime="2846.780" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2845.732" EndTime="2847.830" Participant="me011">
      Did you read the paper? <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2845.832" EndTime="2846.720" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2846.613" EndTime="2851.114" Participant="mn017">
      But weren't you cheating in those experiments? <VocalSound Description="laugh"/> <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="2846.793" EndTime="2848.100" Participant="fe008">
      Yeah. <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="2846.806" EndTime="2847.574" Participant="me013">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="2847.153" EndTime="2848.200" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2847.431" EndTime="2850.440" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2848.400" EndTime="2850.504" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2848.410" EndTime="2850.300" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2848.515" EndTime="2849.620" Participant="me018">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2848.640" EndTime="2849.810" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2848.770" EndTime="2850.450" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2850.500" EndTime="2851.620" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2850.500" EndTime="2853.800" Participant="me013">
      That's - we <Emphasis> weren't. </Emphasis> That's why we increased the error rate. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2852.207" EndTime="2854.935" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2852.300" EndTime="2855.500" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2852.320" EndTime="2853.520" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2852.480" EndTime="2853.540" Participant="fe008">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2853.909" EndTime="2854.526" Participant="fe008">
      Interesting.
    </Segment>
    <Segment StartTime="2853.960" EndTime="2864.390" Participant="fe016">
      Actually, I think of cheating as <Pause/> a way to do some work, where you can't address all of the computational tasks. Like, if we wanna study,
    </Segment>
    <Segment StartTime="2864.390" EndTime="2871.891" Participant="fe016">
      um, speaker habits, but we can't do speaker detection. But we wanna assume - Let's say, we <Emphasis> know </Emphasis> this is Jane and we <Emphasis> know </Emphasis> this is Chuck, even though
    </Segment>
    <Segment StartTime="2871.383" EndTime="2871.916" Participant="me013">
      Mm-hmm.
    </Segment>
    <Segment StartTime="2871.891" EndTime="2876.540" Participant="fe016">
      <Emphasis> automatically </Emphasis> to look at the habits, we would need to also first figure that out. But we can sort of
    </Segment>
    <Segment StartTime="2877.050" EndTime="2881.373" Participant="fe016">
      cheat on that factor because it's somebody else's research. And then we just -
    </Segment>
    <Segment StartTime="2881.373" EndTime="2884.430" Participant="fe016">
      So it's - Then we wanna make a proof of concept for -
    </Segment>
    <Segment StartTime="2882.211" EndTime="2883.643" Participant="me011">
      <VocalSound Description="mouth"/> What - what this - w-
    </Segment>
    <Segment StartTime="2883.643" EndTime="2887.970" Participant="me011">
      Right. What would this be like if - w- it were perfect? If this component were perfect?
    </Segment>
    <Segment StartTime="2885.600" EndTime="2887.100" Participant="fe016">
      If someone else -
    </Segment>
    <Segment StartTime="2887.100" EndTime="2892.880" Participant="fe016">
      Right. Eh - or - w- But we really can't work on that problem. We don't have time, we're not interested, or w- whatever -
    </Segment>
    <Segment StartTime="2887.121" EndTime="2887.608" Participant="mn014">
      Yep.
    </Segment>
    <Segment StartTime="2892.194" EndTime="2892.595" Participant="me011">
      Hmm.
    </Segment>
    <Segment StartTime="2892.880" EndTime="2897.540" Participant="fe016">
      And so you - you assume it's given because it's, um -
    </Segment>
    <Segment StartTime="2893.070" EndTime="2895.760" Participant="me011">
      It's too hard, usually. <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="2897.921" EndTime="2902.190" Participant="fe016">
      you wanna go forward with your research, and assume that you have that information. So,
    </Segment>
    <Segment StartTime="2901.154" EndTime="2901.884" Participant="fe008">
      Mm-hmm. <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="2902.050" EndTime="2902.455" Participant="me013">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2902.190" EndTime="2907.090" Participant="fe016">
      I guess if - I don't ever think of it as negative. More like, it's something <Emphasis> we're </Emphasis> not building.
    </Segment>
    <Segment StartTime="2904.208" EndTime="2905.850" Participant="me011">
      It's not <Pronounce Pronunciation="&quot;perjortative&quot;"> pejorative  </Pronounce> at all.
    </Segment>
    <Segment StartTime="2906.221" EndTime="2908.852" Participant="fe008">
      Well. <VocalSound Description="laugh"/> The term itself <Emphasis> is. </Emphasis> You know?
    </Segment>
    <Segment StartTime="2908.364" EndTime="2908.542" Participant="me011">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="2908.799" EndTime="2911.708" Participant="me011">
      But, it's not pejorative towards the <Emphasis> data. </Emphasis> It's pejorative towards -
    </Segment>
    <Segment StartTime="2908.852" EndTime="2914.432" Participant="fe008">
      But - but it - And it's pejorative with respect to a certain purpose. I understand. I - I wanted to raise the issue.
    </Segment>
    <Segment StartTime="2911.707" EndTime="2912.320" Participant="mn017">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="2912.637" EndTime="2914.696" Participant="me011">
      It's pejorative to o- ourselves. Right?
    </Segment>
    <Segment StartTime="2913.522" EndTime="2916.410" Participant="mn017">
      <VocalSound Description="breath"/> S- self - <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="2914.432" EndTime="2915.698" Participant="fe008">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2914.696" EndTime="2921.290" Participant="me011">
      To say "I am cheating in this experiment" is not saying that the <Emphasis> data </Emphasis> is bad. It's saying that my <Emphasis> experiment </Emphasis> is bad. <Comment Description="last four words while laughing"/>
    </Segment>
    <Segment StartTime="2915.340" EndTime="2916.329" Participant="me013">
      Yeah. <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2916.558" EndTime="2917.801" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2916.586" EndTime="2917.267" Participant="me013">
      But -
    </Segment>
    <Segment StartTime="2916.640" EndTime="2917.228" Participant="fe008">
      Yeah.
    </Segment>
    <Segment StartTime="2918.690" EndTime="2919.121" Participant="me013">
      Yeah.
    </Segment>
    <Segment StartTime="2919.608" EndTime="2920.861" Participant="fe008">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2919.840" EndTime="2923.938" Participant="me013">
      <VocalSound Description="inbreath"/> Anyway, it's colloquial, an- and - and, uh, it's -
    </Segment>
    <Segment StartTime="2920.861" EndTime="2921.488" Participant="fe008">
      Anyway.
    </Segment>
    <Segment StartTime="2923.938" EndTime="2928.349" Participant="me013">
      it's interesting to hear that - that someone coming from a different direction - it s- it sounds the way it sounds to you.
    </Segment>
    <Segment StartTime="2928.349" EndTime="2930.389" Participant="me013">
      But - but, uh, I - I'd never heard that before.
    </Segment>
    <Segment StartTime="2930.350" EndTime="2931.678" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="2930.960" EndTime="2934.309" Participant="fe008">
      O_K. That's interesting. O_K. Well, I wanted to raise the issue, and um,
    </Segment>
    <Segment StartTime="2931.258" EndTime="2932.202" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2931.430" EndTime="2933.760" Participant="me013">
      So, so i- so, that's interesting.
    </Segment>
    <Segment StartTime="2933.747" EndTime="2937.133" Participant="mn017">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2933.984" EndTime="2934.702" Participant="me013">
      Mm-hmm. <Comment Description="quietly"/>
    </Segment>
    <Segment StartTime="2934.309" EndTime="2937.050" Participant="fe008">
      I - I, uh <Comment Description="while laughing"/> - I appreciate the discussion.
    </Segment>
    <Segment StartTime="2937.050" EndTime="2941.918" Participant="fe008">
      <VocalSound Description="inbreath"/> I wanted to ask one other question which is a different <Pause/> matter, which is with respect to, um, this
    </Segment>
    <Segment StartTime="2941.918" EndTime="2945.594" Participant="fe008">
      thing that you've been r- working on for the recording monitoring script?
    </Segment>
    <Segment StartTime="2943.508" EndTime="2944.626" Participant="me011">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="2946.228" EndTime="2952.573" Participant="fe008">
      So. The idea - You had this script that you're working on, to be sure that the microphone values are in - are kept in - ?
    </Segment>
    <Segment StartTime="2950.603" EndTime="2951.295" Participant="me011">
      Oh, right.
    </Segment>
    <Segment StartTime="2951.680" EndTime="2954.000" Participant="me011">
      I haven't gotten back to that recently. But <Pause/>
    </Segment>
    <Segment StartTime="2953.160" EndTime="2954.471" Participant="fe008">
      O_K. <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="2954.704" EndTime="2955.176" Participant="fe008">
      <Pronounce Pronunciation="&quot;kay&quot;"> O_K.  </Pronounce>
    </Segment>
    <Segment StartTime="2955.153" EndTime="2956.891" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="2955.449" EndTime="2957.819" Participant="me011">
      I assume you're saying you want me to <Pause/> get back to it.
    </Segment>
    <Segment StartTime="2955.540" EndTime="2956.506" Participant="fe008">
      Cuz, y- you know -
    </Segment>
    <Segment StartTime="2957.671" EndTime="2959.584" Participant="fe008">
      <VocalSound Description="mouth"/> Well, I was just wondering if - i- i-
    </Segment>
    <Segment StartTime="2958.650" EndTime="2960.360" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2959.584" EndTime="2961.808" Participant="fe008">
      Because I - I <Emphasis> am </Emphasis> finding that,
    </Segment>
    <Segment StartTime="2961.808" EndTime="2964.653" Participant="fe008">
      um, in double checking, I've run across, u- um,
    </Segment>
    <Segment StartTime="2964.928" EndTime="2968.787" Participant="fe008">
      one data set where the microphone was off early on,
    </Segment>
    <Segment StartTime="2968.100" EndTime="2972.274" Participant="me001">
      <NonVocalSound Description="mike noises"/>
    </Segment>
    <Segment StartTime="2969.105" EndTime="2973.442" Participant="fe008">
      and then two other <Pause/> speakers - their microphones went off later. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2971.398" EndTime="2972.650" Participant="me013">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="2973.232" EndTime="2973.772" Participant="me011">
      Hmm.
    </Segment>
    <Segment StartTime="2973.442" EndTime="2977.439" Participant="fe008">
      And this is out of, like, seven speakers. So out of seven speakers, four microphones -
    </Segment>
    <Segment StartTime="2976.269" EndTime="2980.245" Participant="me001">
      <NonVocalSound Description="mike noises"/>
    </Segment>
    <Segment StartTime="2977.439" EndTime="2984.960" Participant="fe008">
      Y- you know, basically, three - three microphones bak- basically flaked, which makes it <Emphasis> hard </Emphasis> for the data to be <Pause/> <Emphasis> used </Emphasis> for all possible purposes.
    </Segment>
    <Segment StartTime="2982.813" EndTime="2983.294" Participant="me018">
      D-
    </Segment>
    <Segment StartTime="2982.870" EndTime="2984.170" Participant="me011">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2983.404" EndTime="2983.970" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="2983.970" EndTime="2985.599" Participant="me018">
      Do you think the battery ran out? Or - ?
    </Segment>
    <Segment StartTime="2984.960" EndTime="2988.067" Participant="fe008">
      That's what <Emphasis> I </Emphasis> think. I think the two that flaked <Emphasis> late, </Emphasis> I think it was the battery.
    </Segment>
    <Segment StartTime="2988.067" EndTime="2992.755" Participant="fe008">
      But, you know, if the script could, um, you know, alert the <Emphasis> recording </Emphasis> person to that -
    </Segment>
    <Segment StartTime="2992.755" EndTime="2997.326" Participant="fe008">
      I mean, I don't know if there's a way to <Emphasis> replace </Emphasis> a battery, if it happens in the middle of a meeting. Maybe that's hopeless anyway. But -
    </Segment>
    <Segment StartTime="2992.843" EndTime="2994.186" Participant="me011">
      Well - <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="2996.849" EndTime="2999.144" Participant="me011">
      Well, you <Emphasis> can, </Emphasis> but you'll lose a lot of data. But, I mean, that
    </Segment>
    <Segment StartTime="2997.617" EndTime="2998.448" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="2999.144" EndTime="3002.777" Participant="me011">
      doesn't <Emphasis> really </Emphasis> help because often the recording person isn't in the room.
    </Segment>
    <Segment StartTime="3002.985" EndTime="3003.893" Participant="fe008">
      <VocalSound Description="mouth"/> Oh, I see.
    </Segment>
    <Segment StartTime="3004.070" EndTime="3008.127" Participant="me011">
      So, what are you gonna do? I mean it will - w- the person - if you're looking up at the board and I
    </Segment>
    <Segment StartTime="3008.127" EndTime="3012.300" Participant="me011">
      disable the screensaver, you will <Emphasis> see </Emphasis> that the mike is off, but that doesn't necessarily help.
    </Segment>
    <Segment StartTime="3012.064" EndTime="3014.657" Participant="fe008">
      <VocalSound Description="mouth"/> O_K. Then I guess that raises the question of, <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="3012.300" EndTime="3012.895" Participant="me011">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="3012.895" EndTime="3013.730" Participant="me011">
      So <Pause/>
    </Segment>
    <Segment StartTime="3014.657" EndTime="3017.119" Participant="fe008">
      uh, whether we should <Emphasis> screen </Emphasis> the data before they get
    </Segment>
    <Segment StartTime="3017.321" EndTime="3023.814" Participant="fe008">
      <Emphasis> transcribed. </Emphasis> Because, although I think that the data are <Emphasis> still </Emphasis> useful in terms of providing <Emphasis> content, </Emphasis> <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="3023.814" EndTime="3031.558" Participant="fe008">
      and - and <Emphasis> that, </Emphasis> I know that having three out of seven microphones out of commission during some part of the meeting restricts the,
    </Segment>
    <Segment StartTime="3031.258" EndTime="3032.467" Participant="me013">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="3031.558" EndTime="3034.526" Participant="fe008">
      uh, usefulness of the data <Pause/> for other purposes. <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="3032.220" EndTime="3032.927" Participant="fe016">
      <Uncertain> Yeah. </Uncertain>
    </Segment>
    <Segment StartTime="3032.709" EndTime="3033.211" Participant="me011">
      Mm-hmm.
    </Segment>
    <Segment StartTime="3034.236" EndTime="3038.438" Participant="fe016">
      I - I think that's a <Emphasis> really </Emphasis> good idea. Because for all kinds of studies, we
    </Segment>
    <Segment StartTime="3034.526" EndTime="3035.789" Participant="fe008">
      And maybe you don't wanna have it -
    </Segment>
    <Segment StartTime="3038.438" EndTime="3045.643" Participant="fe016">
      don't really enjoy meetings where some of the - where the signal's going off at times. It just makes it hard to study <Emphasis> any </Emphasis> kind of parameters. So,
    </Segment>
    <Segment StartTime="3046.010" EndTime="3051.587" Participant="fe016">
      if there's a way to check the signal quality before transcribing it and you find any problem at <Emphasis> all, </Emphasis>
    </Segment>
    <Segment StartTime="3050.533" EndTime="3055.510" Participant="me018">
      <VocalSound Description="yawn"/>
    </Segment>
    <Segment StartTime="3051.587" EndTime="3054.320" Participant="fe016">
      it'd be much better to go to another <Emphasis> meeting, </Emphasis> <Emphasis> I </Emphasis> think.
    </Segment>
    <Segment StartTime="3053.600" EndTime="3060.936" Participant="fe008">
      <VocalSound Description="inbreath"/> I actually think that Thilo's - n- in a - in a - k- a - You know, when you do the pre-segmenter and you - and you run across troubles - He - he runs across s- some of these that are -
    </Segment>
    <Segment StartTime="3057.832" EndTime="3058.364" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="3058.820" EndTime="3065.275" Participant="mn014">
      Yeah. Some of these <Pause/> things are captured by - Yeah. Um, by looking at - at the minimum and maximum and whatever you find in the channels, and -
    </Segment>
    <Segment StartTime="3061.613" EndTime="3063.085" Participant="fe008">
      So, we have <Emphasis> some </Emphasis> part of that.
    </Segment>
    <Segment StartTime="3064.863" EndTime="3066.050" Participant="fe008">
      <VocalSound Description="inbreath"/> I c-
    </Segment>
    <Segment StartTime="3065.107" EndTime="3068.544" Participant="me011">
      <VocalSound Description="mouth"/> It's just hard to tell between that and just someone not talking.
    </Segment>
    <Segment StartTime="3065.275" EndTime="3065.948" Participant="mn014">
      But -
    </Segment>
    <Segment StartTime="3066.226" EndTime="3067.778" Participant="fe008">
      <VocalSound Description="clears throat"/>
    </Segment>
    <Segment StartTime="3066.304" EndTime="3067.444" Participant="mn014">
      Yep. Yep.
    </Segment>
    <Segment StartTime="3068.471" EndTime="3069.536" Participant="mn014">
      Yeah. Yeah.
    </Segment>
    <Segment StartTime="3069.980" EndTime="3070.425" Participant="me011">
      So.
    </Segment>
    <Segment StartTime="3070.330" EndTime="3072.561" Participant="fe008">
      Well, there- the - And the other aspect of it is,
    </Segment>
    <Segment StartTime="3072.561" EndTime="3077.110" Participant="fe008">
      um, that, when the microphone is not well-adjusted, then y- i- i-
    </Segment>
    <Segment StartTime="3076.636" EndTime="3077.922" Participant="mn017">
      <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="3077.110" EndTime="3082.568" Participant="fe008">
      even if it's not a <Emphasis> lapel </Emphasis> mike, you can get lapel mike type behavior. Cuz I'm - I'm expecting, for example, with this, that you're gonna end up,
    </Segment>
    <Segment StartTime="3079.564" EndTime="3080.215" Participant="me011">
      Right.
    </Segment>
    <Segment StartTime="3082.187" EndTime="3083.600" Participant="fe016">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="3082.568" EndTime="3083.080" Participant="fe008">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="3083.080" EndTime="3084.814" Participant="fe008">
      you know, picking up other peoples' signals.
    </Segment>
    <Segment StartTime="3083.600" EndTime="3086.222" Participant="fe016">
      I've already written some notes here, so I - <Comment Description="last two words while laughing"/>
    </Segment>
    <Segment StartTime="3085.782" EndTime="3086.325" Participant="me011">
      Mm-hmm.
    </Segment>
    <Segment StartTime="3086.222" EndTime="3088.312" Participant="fe016">
      Um, right. <Uncertain> Uh - </Uncertain>
    </Segment>
    <Segment StartTime="3086.300" EndTime="3088.514" Participant="fe008">
      Yeah. I mean, it's just - <Uncertain> you know </Uncertain>, the microphone is
    </Segment>
    <Segment StartTime="3088.564" EndTime="3093.410" Participant="me011">
      <VocalSound Description="mouth"/> Well, we should be getting new equipment in, so we don't have to use the earplug any more. <Comment Description="last four words while laughing"/>
    </Segment>
    <Segment StartTime="3088.690" EndTime="3090.697" Participant="fe008">
      intended for certain adjustments.
    </Segment>
    <Segment StartTime="3090.866" EndTime="3091.428" Participant="fe016">
      O_K.
    </Segment>
    <Segment StartTime="3092.157" EndTime="3095.335" Participant="fe016">
      But it's my ear. I'm sure it's something wrong with my head, but -
    </Segment>
    <Segment StartTime="3094.068" EndTime="3098.661" Participant="me011">
      It's your <Uncertain> hair </Uncertain>. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="3094.459" EndTime="3095.367" Participant="mn014">
      Hmm.
    </Segment>
    <Segment StartTime="3094.540" EndTime="3099.965" Participant="me013">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="3095.335" EndTime="3104.930" Participant="fe016">
      Um, actually, <Emphasis> is </Emphasis> there a way, though, to use whatever you're using for background noise to check post hoc that a microphone was constantly on? I mean -
    </Segment>
    <Segment StartTime="3095.360" EndTime="3096.750" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="3095.426" EndTime="3098.020" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="3102.136" EndTime="3102.709" Participant="me011">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="3104.280" EndTime="3106.893" Participant="me011">
      I mean, y- d- you - you can do <Emphasis> sort </Emphasis> of a check,
    </Segment>
    <Segment StartTime="3104.930" EndTime="3106.720" Participant="fe016">
      Uh, like this o-
    </Segment>
    <Segment StartTime="3106.893" EndTime="3111.087" Participant="me011">
      but it will be very hard to tell the difference between that and, um, someone not talking.
    </Segment>
    <Segment StartTime="3111.130" EndTime="3114.774" Participant="me013">
      Just so it doesn't - when the microphone's dead, it doesn't put out zeros? Or - ? Mm-hmm.
    </Segment>
    <Segment StartTime="3113.595" EndTime="3114.155" Participant="me011">
      No.
    </Segment>
    <Segment StartTime="3114.155" EndTime="3115.509" Participant="me011">
      <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="3114.957" EndTime="3116.500" Participant="mn017">
      What <Emphasis> does </Emphasis> it put out?
    </Segment>
    <Segment StartTime="3115.395" EndTime="3115.920" Participant="me001">
      Really?
    </Segment>
    <Segment StartTime="3115.920" EndTime="3117.560" Participant="me013">
      Johnson noise? Yeah.
    </Segment>
    <Segment StartTime="3116.200" EndTime="3119.670" Participant="fe016">
      But then how are y- how are you detecting during a meeting?
    </Segment>
    <Segment StartTime="3116.251" EndTime="3117.194" Participant="me011">
      Little bit of noise.
    </Segment>
    <Segment StartTime="3118.870" EndTime="3119.918" Participant="me011">
      I use a threshold.
    </Segment>
    <Segment StartTime="3120.260" EndTime="3124.618" Participant="me011">
      If it's below a <Pronounce Pronunciation="&quot;partircular&quot;"> particular  </Pronounce> value, it - it flashes yellow. So as I - eh - but it's not <Emphasis> perfect. </Emphasis>
    </Segment>
    <Segment StartTime="3124.880" EndTime="3126.320" Participant="fe016">
      But is it better than nothing?
    </Segment>
    <Segment StartTime="3126.113" EndTime="3127.109" Participant="me011">
      Yeah, probably.
    </Segment>
    <Segment StartTime="3127.140" EndTime="3128.710" Participant="fe016">
      Cuz it would really be -
    </Segment>
    <Segment StartTime="3127.438" EndTime="3130.993" Participant="me011">
      I mean, this is the reason why I haven't gotten back to it, is cuz my first pass at it
    </Segment>
    <Segment StartTime="3130.993" EndTime="3135.967" Participant="me011">
      didn't really work because all the mikes have different noise levels. And so I have to do something a little more clever.
    </Segment>
    <Segment StartTime="3133.600" EndTime="3134.370" Participant="fe016">
      Right.
    </Segment>
    <Segment StartTime="3133.892" EndTime="3134.764" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="3135.620" EndTime="3138.790" Participant="me001">
      It's just the noise from the connections and the - everything.
    </Segment>
    <Segment StartTime="3136.100" EndTime="3137.020" Participant="fe016">
      But you could -
    </Segment>
    <Segment StartTime="3137.496" EndTime="3138.015" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="3138.410" EndTime="3141.836" Participant="fe016">
      So, you <Emphasis> could </Emphasis> run that post hoc on a already recorded meeting,
    </Segment>
    <Segment StartTime="3138.787" EndTime="3139.179" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="3142.320" EndTime="3149.418" Participant="fe016">
      in the sense that, you know, not everyone - as you just said, you won't <Emphasis> be </Emphasis> there. And then if we find any problems, have the transcribers listen and - I <Emphasis> really </Emphasis> think it's
    </Segment>
    <Segment StartTime="3146.863" EndTime="3147.382" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="3149.418" EndTime="3154.080" Participant="fe016">
      better not to transcribe a meeting that's gonna have problems once you've spent all this effort.
    </Segment>
    <Segment StartTime="3152.355" EndTime="3152.874" Participant="me011">
      Yep.
    </Segment>
    <Segment StartTime="3153.850" EndTime="3158.461" Participant="fe008">
      <VocalSound Description="inbreath"/> The only argument for doing so would be with reference to the content. Content maps.
    </Segment>
    <Segment StartTime="3157.479" EndTime="3159.480" Participant="fe016">
      Yeah. Yeah.
    </Segment>
    <Segment StartTime="3158.461" EndTime="3164.140" Participant="fe008">
      <VocalSound Description="inbreath"/> Uh - but mayb- maybe <Emphasis> then </Emphasis> it should be done in the old <Emphasis> original </Emphasis> way, w- instead of channelizing and having the, uh,
    </Segment>
    <Segment StartTime="3162.226" EndTime="3162.760" Participant="mn017">
      Mmm.
    </Segment>
    <Segment StartTime="3164.139" EndTime="3167.046" Participant="mn017">
      <VocalSound Description="inbreath"/> Yeah. You could still - eh <Pause/> <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="3164.480" EndTime="3166.236" Participant="fe008">
      uh, you know, elaborate coding of the
    </Segment>
    <Segment StartTime="3164.533" EndTime="3167.541" Participant="me011">
      <VocalSound Description="inbreath"/> Yeah. There's - eh - the s- the standard deviation of the signal
    </Segment>
    <Segment StartTime="3166.803" EndTime="3167.768" Participant="fe008">
      backchannels.
    </Segment>
    <Segment StartTime="3167.541" EndTime="3170.713" Participant="me011">
      gives you a good clue. I mean, if that is too low,
    </Segment>
    <Segment StartTime="3168.867" EndTime="3169.723" Participant="fe016">
      Mm-hmm.
    </Segment>
    <Segment StartTime="3170.705" EndTime="3171.313" Participant="me001">
      Hmm.
    </Segment>
    <Segment StartTime="3170.720" EndTime="3171.213" Participant="mn014">
      Yeah.
    </Segment>
    <Segment StartTime="3171.070" EndTime="3173.280" Participant="me011">
      then you can be pretty sure that it's, uh, empty.
    </Segment>
    <Segment StartTime="3171.367" EndTime="3172.473" Participant="mn017">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="3173.342" EndTime="3177.656" Participant="mn014">
      Yeah. Sometimes you - you capture that when you mix it together, and - Yeah. Yeah.
    </Segment>
    <Segment StartTime="3173.470" EndTime="3174.452" Participant="fe008">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="3176.590" EndTime="3177.570" Participant="me011">
      Right. Exactly.
    </Segment>
    <Segment StartTime="3177.089" EndTime="3184.655" Participant="fe008">
      And actually, an - an alternative to even doing <Emphasis> that </Emphasis> level of transcription would be to have a transcriber listen and - and m- you know, maybe just -
    </Segment>
    <Segment StartTime="3185.201" EndTime="3186.473" Participant="fe008">
      Oh, I don't know if a transcr- uh -
    </Segment>
    <Segment StartTime="3186.473" EndTime="3187.060" Participant="fe008">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="3187.060" EndTime="3192.572" Participant="fe008">
      Well, <Emphasis> someone </Emphasis> who's associated with the meeting could have, like, a <Emphasis> summary </Emphasis> of points handled in the meeting. You know, maybe if we could
    </Segment>
    <Segment StartTime="3193.310" EndTime="3197.630" Participant="fe008">
      pay one person who knows that subject matter to do an outline of the meeting's content.
    </Segment>
    <Segment StartTime="3196.194" EndTime="3196.636" Participant="me011">
      <Uncertain> Uh-huh. </Uncertain>
    </Segment>
    <Segment StartTime="3198.009" EndTime="3201.728" Participant="mn017">
      <VocalSound Description="mouth"/> Well, it's even - If y- if you could still transcribe
    </Segment>
    <Segment StartTime="3198.063" EndTime="3200.335" Participant="fe008">
      Instead of losing the content altogether.
    </Segment>
    <Segment StartTime="3198.091" EndTime="3198.692" Participant="me011">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="3199.210" EndTime="3200.139" Participant="me011">
      Is there a - ?
    </Segment>
    <Segment StartTime="3202.140" EndTime="3206.027" Participant="mn017">
      the words based on the <Pause/> far-field microphone or something, <NonVocalSound Description="mike noises"/>
    </Segment>
    <Segment StartTime="3204.040" EndTime="3204.940" Participant="me011">
      Far-field. Yeah.
    </Segment>
    <Segment StartTime="3206.244" EndTime="3209.983" Participant="mn017">
      you could, uh, still use it for, say, language modeling, <NonVocalSound Description="thumping noise"/>
    </Segment>
    <Segment StartTime="3210.045" EndTime="3214.436" Participant="fe008">
      <VocalSound Description="mouth"/> I guess the troub- trouble in my mind is that it - that it's not a very neat corpus if you say
    </Segment>
    <Segment StartTime="3210.398" EndTime="3211.110" Participant="mn017">
      you know.
    </Segment>
    <Segment StartTime="3213.980" EndTime="3214.840" Participant="fe016">
      Yeah.
    </Segment>
    <Segment StartTime="3214.436" EndTime="3219.660" Participant="fe008">
      "these data are available, but these are imperfect, because of the f- you know, the batteries flaked." <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="3214.490" EndTime="3214.934" Participant="mn017">
      Right.
    </Segment>
    <Segment StartTime="3216.155" EndTime="3217.132" Participant="fe016">
      Yeah.
    </Segment>
    <Segment StartTime="3217.495" EndTime="3220.118" Participant="fe016">
      I'd mu- We're - I'd much rather have,
    </Segment>
    <Segment StartTime="3217.581" EndTime="3219.833" Participant="me011">
      Well, we'll just - We'll - we'll just have to note those.
    </Segment>
    <Segment StartTime="3220.118" EndTime="3226.420" Participant="fe016">
      you know, meetings that have all the channels, even if we had to skip a meeting or something, just for <Pause/> all these other purposes. So.
    </Segment>
    <Segment StartTime="3222.198" EndTime="3224.270" Participant="mn017">
      <VocalSound Description="inbreath"/> Well, if - Right.
    </Segment>
    <Segment StartTime="3224.848" EndTime="3225.976" Participant="mn017">
      <VocalSound Description="outbreath"/> <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="3226.131" EndTime="3228.581" Participant="me001">
      You could put them at the back of the queue or something.
    </Segment>
    <Segment StartTime="3226.255" EndTime="3227.627" Participant="mn017">
      I guess there <Emphasis> is </Emphasis> no -
    </Segment>
    <Segment StartTime="3226.360" EndTime="3227.460" Participant="fe008">
      And we have <Emphasis> so </Emphasis> many.
    </Segment>
    <Segment StartTime="3226.459" EndTime="3231.310" Participant="me011">
      Well, I - I - I think we can't throw away that data, cuz otherwise we'll end up with very few meetings. But <Pause/>
    </Segment>
    <Segment StartTime="3228.416" EndTime="3230.950" Participant="fe016">
      Yeah. Right. But may- just -
    </Segment>
    <Segment StartTime="3229.070" EndTime="3232.340" Participant="mn017">
      there's - there's no shortage of meetings, so <Pause/>
    </Segment>
    <Segment StartTime="3230.066" EndTime="3230.990" Participant="me001">
      Yeah.
    </Segment>
    <Segment StartTime="3231.085" EndTime="3231.814" Participant="mn014">
      Yep.
    </Segment>
    <Segment StartTime="3231.950" EndTime="3235.520" Participant="fe016">
      Right. There - but there's a shortage of trans- transcription power.
    </Segment>
    <Segment StartTime="3232.035" EndTime="3232.431" Participant="me011">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="3232.340" EndTime="3234.762" Participant="mn017">
      we can afford - we can a-
    </Segment>
    <Segment StartTime="3232.710" EndTime="3238.259" Participant="fe008">
      We have <Emphasis> such </Emphasis> a backlog. I mean, to be able to get through the backlog of the - of the <Emphasis> good </Emphasis> meetings, it would be nice to not have
    </Segment>
    <Segment StartTime="3232.870" EndTime="3234.810" Participant="mn014">
      <VocalSound Description="cough"/>
    </Segment>
    <Segment StartTime="3233.511" EndTime="3235.270" Participant="me018">
      <VocalSound Description="cough in background"/>
    </Segment>
    <Segment StartTime="3236.282" EndTime="3237.461" Participant="mn017">
      <VocalSound Description="outbreath"/> Hmm.
    </Segment>
    <Segment StartTime="3236.873" EndTime="3238.020" Participant="fe016">
      Yeah.
    </Segment>
    <Segment StartTime="3237.039" EndTime="3237.336" Participant="me011">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="3237.945" EndTime="3241.650" Participant="me011">
      Do we have a - ? I'm sorry for interrupting. Do we have an E_D_U meeting at four?
    </Segment>
    <Segment StartTime="3239.000" EndTime="3240.399" Participant="fe008">
      energies diverted. That's fine.
    </Segment>
    <Segment StartTime="3241.881" EndTime="3247.120" Participant="fe016">
      You know what? They can- their four o'clock cancelled, and they're starting at four twenty, and I'll set up at four fifteen.
    </Segment>
    <Segment StartTime="3242.049" EndTime="3242.730" Participant="mn017">
      Hmm.
    </Segment>
    <Segment StartTime="3242.170" EndTime="3243.062" Participant="fe008">
      Oh, good question.
    </Segment>
    <Segment StartTime="3242.800" EndTime="3243.920" Participant="mn014">
      I don't think so.
    </Segment>
    <Segment StartTime="3245.930" EndTime="3246.548" Participant="me011">
      O_K. Great.
    </Segment>
    <Segment StartTime="3246.827" EndTime="3247.201" Participant="me011">
      Good.
    </Segment>
    <Segment StartTime="3246.993" EndTime="3247.367" Participant="me013">
      <VocalSound Description="mouth"/>
    </Segment>
    <Segment StartTime="3247.002" EndTime="3249.587" Participant="mn017">
      Hmm. <VocalSound Description="mouth"/> Hey, I bet there's <Emphasis> tea. </Emphasis> <Comment Description="while smiling"/> <Comment Description="indirect request"/>
    </Segment>
    <Segment StartTime="3247.742" EndTime="3248.660" Participant="me013">
      Yeah, but there's -
    </Segment>
    <Segment StartTime="3248.470" EndTime="3250.280" Participant="me011">
      <VocalSound Description="mouth"/> Cuz otherwise I was gonna say we have to cancel.
    </Segment>
    <Segment StartTime="3249.030" EndTime="3252.410" Participant="me013">
      There's <Emphasis> tea, </Emphasis> also, though. So, maybe we should - maybe we should do a, uh,
    </Segment>
    <Segment StartTime="3249.212" EndTime="3249.695" Participant="fe016">
      Yeah.
    </Segment>
    <Segment StartTime="3249.981" EndTime="3251.348" Participant="mn017">
      <Pronounce Pronunciation="&quot;nn-yeah&quot;"> Yeah.  </Pronounce> <VocalSound Description="prolonged vocal squeak"/>
    </Segment>
    <Segment StartTime="3250.200" EndTime="3250.769" Participant="fe008">
      O_K.
    </Segment>
    <Segment StartTime="3250.504" EndTime="3251.753" Participant="fe016">
      Yeah. So you'll have si-
    </Segment>
    <Segment StartTime="3251.348" EndTime="3252.075" Participant="mn017">
      <NonVocalSound Description="squeaky door in the background"/>
    </Segment>
    <Segment StartTime="3252.410" EndTime="3253.320" Participant="me001">
      <VocalSound Description="outbreath"/>
    </Segment>
    <Segment StartTime="3252.410" EndTime="3252.911" Participant="me013">
      <VocalSound Description="breath"/>
    </Segment>
    <Segment StartTime="3252.911" EndTime="3258.153" Participant="me013">
      simultaneous digits - digits read. In the interest of - of getting the snacks.
    </Segment>
    <Segment StartTime="3253.451" EndTime="3255.700" Participant="me011">
      A simultaneous digit. O_K. <Comment Description="while laughing"/>
    </Segment>
    <Segment StartTime="3255.749" EndTime="3257.089" Participant="fe008">
      <Uncertain> Mm-hmm. </Uncertain>
    </Segment>
    <Segment StartTime="3255.897" EndTime="3257.092" Participant="me011">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="3256.126" EndTime="3257.594" Participant="me001">
      <VocalSound Description="clears throat"/>
    </Segment>
    <Segment StartTime="3256.840" EndTime="3261.303" Participant="mn017">
      Hey, I've never done one of those. <Comment Description="with mock enthusiasm"/> <VocalSound Description="laugh"/> No. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="3258.529" EndTime="3260.430" Participant="me013">
      Oh, you never have? Oh, it's a treat.
    </Segment>
    <Segment StartTime="3260.190" EndTime="3261.430" Participant="me051">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="3260.275" EndTime="3261.761" Participant="me011">
      <VocalSound Description="breath-laugh"/>
    </Segment>
    <Segment StartTime="3260.600" EndTime="3262.220" Participant="me001">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="3260.850" EndTime="3264.686" Participant="fe016">
      You have to plug your ears or just prepare not to laugh. <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="3262.662" EndTime="3264.315" Participant="mn017">
      Right.
    </Segment>
    <Segment StartTime="3263.877" EndTime="3264.582" Participant="me013">
      O_K.
    </Segment>
    <Segment StartTime="3263.987" EndTime="3267.840" Participant="me011">
      Everyone ready? S- reading <Pronounce> simultaneous  </Pronounce> digits? Three! Two! One!
    </Segment>
    <Segment StartTime="3264.730" EndTime="3265.281" Participant="fe008">
      Yeah.
    </Segment>
    <Segment StartTime="3265.570" EndTime="3266.429" Participant="fe008">
      Yep. Yep.
    </Segment>
    <Segment StartTime="3267.340" EndTime="3268.170" Participant="me001">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="3267.520" EndTime="3268.210" Participant="fe008">
      <VocalSound Description="inbreath"/>
    </Segment>
    <Segment StartTime="3268.064" EndTime="3270.769" Participant="me013" DigitTask="true">
      Transcript L_ s- dash two hundred. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3268.068" EndTime="3270.559" Participant="me018" DigitTask="true">
      Transcript L_ dash two zero two. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3268.095" EndTime="3269.900" Participant="me011" DigitTask="true">
      L_ two O_ three. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3268.157" EndTime="3270.486" Participant="mn014" DigitTask="true">
      Transcript L_ two hundred and one. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3268.170" EndTime="3269.280" Participant="mn017">
      Sev-
    </Segment>
    <Segment StartTime="3268.170" EndTime="3270.417" Participant="me001" DigitTask="true">
      Transcript L_ dash two zero five. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3268.210" EndTime="3269.675" Participant="fe008" DigitTask="true">
      L_ one nine nine <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3268.360" EndTime="3269.955" Participant="fe016" DigitTask="true">
      Transcript L_ two O_ six. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3268.530" EndTime="3271.720" Participant="me051" DigitTask="true">
      Transcript L_ two O_ seven. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3269.400" EndTime="3271.204" Participant="mn017" DigitTask="true">
      L_ two zero four. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3270.273" EndTime="3273.674" Participant="fe016" DigitTask="true">
      seven two seven, one eight, six two two eight <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3270.420" EndTime="3272.898" Participant="fe008" DigitTask="true">
      two two nine five, eight, seven one five <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3270.832" EndTime="3273.957" Participant="me001" DigitTask="true">
      two two six, five two, one four four eight <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3270.844" EndTime="3274.948" Participant="me018" DigitTask="true">
      five six, eight four, zero nine, nine nine, two six <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3270.870" EndTime="3274.728" Participant="mn014" DigitTask="true">
      eight seven one, one three five, five eight one <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3270.950" EndTime="3273.778" Participant="me011" DigitTask="true">
      three two six eight, seven, two six seven <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3271.526" EndTime="3274.768" Participant="me013" DigitTask="true">
      seven two two nine, eight, one two six <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3271.828" EndTime="3275.900" Participant="mn017" DigitTask="true">
      seven one three five, two, seven three eight <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3272.874" EndTime="3276.653" Participant="me051" DigitTask="true">
      seven zero five, six five, one two nine zero <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3273.038" EndTime="3276.035" Participant="fe008" DigitTask="true">
      six nine six, nine one, nine five O_ one <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3273.889" EndTime="3276.665" Participant="fe016" DigitTask="true">
      one two one, eight four four, two six three five <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3274.090" EndTime="3276.695" Participant="me011" DigitTask="true">
      six eight O_ three, one, six four <Emphasis> six </Emphasis> <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3274.207" EndTime="3278.685" Participant="me001" DigitTask="true">
      five five, three one, seven seven, zero six, five three <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3275.220" EndTime="3279.696" Participant="me013" DigitTask="true">
      four five four one, five six three two, two one one six <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3275.241" EndTime="3278.560" Participant="me018" DigitTask="true">
      five eight five, seven six, seven one zero five <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3275.560" EndTime="3278.990" Participant="mn014" DigitTask="true">
      eight six two, four five zero, one six two <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3276.439" EndTime="3279.523" Participant="fe008" DigitTask="true">
      three zero zero, one three seven, two three eight <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3276.542" EndTime="3281.275" Participant="mn017" DigitTask="true">
      three five nine two, eight nine eight nine, two six one six <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3276.966" EndTime="3280.721" Participant="fe016" DigitTask="true">
      five six six one, one five five nine, nine two seven one <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3277.031" EndTime="3280.814" Participant="me051" DigitTask="true">
      seven four zero seven, two, two nine six <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3277.110" EndTime="3280.048" Participant="me011" DigitTask="true">
      three five seven, five two O_, six five four <Emphasis> two </Emphasis> <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3279.070" EndTime="3282.141" Participant="mn014" DigitTask="true">
      two four, two three, eight one, four six, four one <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3279.090" EndTime="3283.988" Participant="me018" DigitTask="true">
      two seven seven four, three one five seven, one two zero seven <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3279.141" EndTime="3283.017" Participant="me001" DigitTask="true">
      two five five, one eight five, three seven three five <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3279.746" EndTime="3283.902" Participant="fe008" DigitTask="true">
      three, four seven two, six five, five three zero, four <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3280.364" EndTime="3283.762" Participant="me013" DigitTask="true">
      three five six, seven six five, four one four <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3280.400" EndTime="3283.962" Participant="me011" DigitTask="true">
      three, five three six, one O_, eight nine two, nine <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3281.056" EndTime="3284.111" Participant="fe016" DigitTask="true">
      eight nine two, six five, nine eight three six <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3281.416" EndTime="3285.423" Participant="me051" DigitTask="true">
      two two five, eight nine six, five seven two <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3281.672" EndTime="3286.008" Participant="mn017" DigitTask="true">
      six, O_ nine nine, O_ two, nine two four, two <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3282.542" EndTime="3285.448" Participant="mn014" DigitTask="true">
      nine two six nine, one, four nine eight <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3283.442" EndTime="3287.905" Participant="me001" DigitTask="true">
      one, one six six, six eight, seven five five, zero <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3284.251" EndTime="3287.699" Participant="me013" DigitTask="true">
      four three two, three one zero, nine four seven five <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3284.391" EndTime="3287.641" Participant="fe008" DigitTask="true">
      seven nine five four, three three six five, zero zero two six <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3284.450" EndTime="3287.130" Participant="me011" DigitTask="true">
      two four three seven, zero, zero nine <Emphasis> two </Emphasis> <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3284.450" EndTime="3288.796" Participant="me018" DigitTask="true">
      three, six eight two, six five, five two eight, five <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3284.573" EndTime="3287.576" Participant="fe016" DigitTask="true">
      one four eight seven, six, seven O_ one <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3285.939" EndTime="3289.407" Participant="mn014" DigitTask="true">
      six two, two nine, one two, nine two, <Emphasis> six </Emphasis> two <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3286.170" EndTime="3289.754" Participant="me051" DigitTask="true">
      three eight five, eight zero, zero six six two <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3286.811" EndTime="3289.735" Participant="mn017" DigitTask="true">
      two two eight four, five, four three eight <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3287.660" EndTime="3290.366" Participant="me011" DigitTask="true">
      O_ five six, two nine zero, nine five <Emphasis> three </Emphasis> <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3287.845" EndTime="3291.300" Participant="fe008" DigitTask="true">
      eight nine two seven, four nine two eight, seven O_ two six. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3288.086" EndTime="3291.101" Participant="me013" DigitTask="true">
      eight one nine, one one, one three one zero <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3288.305" EndTime="3292.740" Participant="me001" DigitTask="true">
      four one, seven four, three two, one seven, <Emphasis> six </Emphasis> eight <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3288.883" EndTime="3291.657" Participant="fe016" DigitTask="true">
      nine seven three, O_ four, five seven nine two <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3289.101" EndTime="3291.987" Participant="me018" DigitTask="true">
      one one seven nine, eight, three three three <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3289.896" EndTime="3293.815" Participant="mn014" DigitTask="true">
      one, two five nine, zero two, one four three, five <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3290.133" EndTime="3294.399" Participant="mn017" DigitTask="true">
      four, six seven seven, two six, one one six, seven <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3290.348" EndTime="3294.303" Participant="me051" DigitTask="true">
      four <Pause/> eight one, three seven eight, five six four <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3290.588" EndTime="3293.103" Participant="me011" DigitTask="true">
      one five eight, nine one, O_ four three <Emphasis> six </Emphasis> <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3291.599" EndTime="3294.864" Participant="me013" DigitTask="true">
      three two seven, five zero, three six nine one <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3291.873" EndTime="3295.547" Participant="fe008" DigitTask="true">
      nine eight five nine, five three one eight, nine three four one <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3291.915" EndTime="3295.130" Participant="fe016" DigitTask="true">
      four, eight six nine, four six, four eight four, nine <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3292.696" EndTime="3295.890" Participant="me018" DigitTask="true">
      three four seven nine, six, five nine zero <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3293.328" EndTime="3297.036" Participant="me001" DigitTask="true">
      zero seven seven, five two eight, nine zero two <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3293.390" EndTime="3296.588" Participant="me011" DigitTask="true">
      five three three eight, one, one eight five. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3293.988" EndTime="3297.296" Participant="mn014" DigitTask="true">
      seven, nine four eight, seven five, zero eight three, six <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3294.824" EndTime="3298.179" Participant="mn017" DigitTask="true">
      four zero one, four nine, zero two three nine <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3294.920" EndTime="3298.513" Participant="me051" DigitTask="true">
      zero zero six, five zero, seven zero four eight <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3295.402" EndTime="3298.805" Participant="fe016" DigitTask="true">
      two five eight six, five, six eight zero. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3295.576" EndTime="3299.219" Participant="me013" DigitTask="true">
      three three seven, five one one, five zero four two <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3295.931" EndTime="3299.137" Participant="fe008" DigitTask="true">
      seven eight one, eight seven six, five four nine seven. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3296.291" EndTime="3299.630" Participant="me018" DigitTask="true">
      six six six, three two nine, nine three eight three <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3297.463" EndTime="3301.795" Participant="me001" DigitTask="true">
      seven six, seven five, zero one, three three, five six <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3297.824" EndTime="3301.031" Participant="mn014" DigitTask="true">
      nine six three five, seven three four two, eight six one eight. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3298.622" EndTime="3301.873" Participant="mn017" DigitTask="true">
      zero two four, four nine, one two two six <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3299.065" EndTime="3303.311" Participant="me051" DigitTask="true">
      seven four, five one, five three, two four, seven one <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3299.942" EndTime="3303.248" Participant="me013" DigitTask="true">
      eight nine zero eight, seven, zero seven six. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3300.127" EndTime="3303.327" Participant="me018" DigitTask="true">
      nine zero one, three three six, zero nine one four. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3302.195" EndTime="3306.303" Participant="mn017" DigitTask="true">
      six four, zero five, nine eight, one seven, one four. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3302.200" EndTime="3306.023" Participant="me001" DigitTask="true">
      three seven eight, five three, six nine seven six. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3303.740" EndTime="3307.080" Participant="me051" DigitTask="true">
      seven four five, six four eight, one three nine nine. <Comment Description="Digits"/>
    </Segment>
    <Segment StartTime="3308.331" EndTime="3310.027" Participant="mn014">
      <VocalSound Description="laugh"/>
    </Segment>
    <Segment StartTime="3308.988" EndTime="3312.331" Participant="fe016">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="3309.120" EndTime="3311.194" Participant="me013">
      <Pronounce Pronunciation="&quot;kay&quot;"> O_K,  </Pronounce> another successful babble.
    </Segment>
    <Segment StartTime="3311.330" EndTime="3312.331" Participant="me018">
      <NonVocalSound Description="mike noise"/>
    </Segment>
    <Segment StartTime="3311.800" EndTime="3312.331" Participant="me011">
      And -
    </Segment>
    <Segment StartTime="3311.810" EndTime="3312.331" Participant="fe008">
      <NonVocalSound Description="mike noise"/>
    </Segment>

  </Transcript>

</Meeting>