1
00:00:07,800 --> 00:00:11,799
Speaker 1: How any extraordinaries. Today, we're pulling back the dKu curtin

2
00:00:11,840 --> 00:00:14,200
Speaker 1: a bit. We're gonna see how the sausage is made,

3
00:00:14,280 --> 00:00:17,720
Speaker 1: as it were. So Daniel and I write pretty darn

4
00:00:17,800 --> 00:00:21,440
Speaker 1: detailed outlines before we record an episode, but often something

5
00:00:21,480 --> 00:00:25,560
Speaker 1: happens that totally derails our outline conversation, and that's good.

6
00:00:26,040 --> 00:00:29,520
Speaker 1: Sometimes the derailment happens because something I thought was kind

7
00:00:29,520 --> 00:00:32,120
Speaker 1: of obvious turns out to not be so obvious, and

8
00:00:32,240 --> 00:00:35,000
Speaker 1: Daniel will ask clarifying questions that forced me to come

9
00:00:35,080 --> 00:00:40,000
Speaker 1: up with different, better explanations on the fly. Or Daniel

10
00:00:40,000 --> 00:00:42,080
Speaker 1: will be working through his outline and I'll think of

11
00:00:42,120 --> 00:00:45,680
Speaker 1: some tangentially related topic that I now really want to

12
00:00:45,720 --> 00:00:48,559
Speaker 1: know about, and Daniel will hop online to do a

13
00:00:48,640 --> 00:00:51,520
Speaker 1: quick bit of research, and then we'll record a completely

14
00:00:51,680 --> 00:00:55,760
Speaker 1: unplanned segment. In each of these cases, our answers are

15
00:00:55,920 --> 00:00:59,280
Speaker 1: way less polished than we'd like, Lots of false starts,

16
00:00:59,400 --> 00:01:03,520
Speaker 1: lots of basically lots of stuff that the audience definitely

17
00:01:03,760 --> 00:01:06,880
Speaker 1: doesn't want to have to sit through. And then Daniel

18
00:01:06,880 --> 00:01:09,720
Speaker 1: and I have unexpected noises that intrude upon the audio.

19
00:01:10,120 --> 00:01:13,759
Speaker 1: In today's episode, I learned that my microphone actually does

20
00:01:13,840 --> 00:01:16,199
Speaker 1: pick up the sounds my goats make when they're trying

21
00:01:16,240 --> 00:01:19,520
Speaker 1: to get my attention, and sometimes Daniel apparently records from

22
00:01:19,560 --> 00:01:22,959
Speaker 1: the inside of a grocery store. More on that later. Anyway,

23
00:01:23,280 --> 00:01:25,720
Speaker 1: these are only a handful of the many challenges that

24
00:01:25,760 --> 00:01:28,840
Speaker 1: we throw at Matt Kesselman, the audio wizard who makes

25
00:01:28,880 --> 00:01:32,520
Speaker 1: Daniel and Kelly's Extraordinary Universe sound so good. So today

26
00:01:32,520 --> 00:01:34,720
Speaker 1: we're bringing Matt on the show to talk about sound

27
00:01:35,000 --> 00:01:37,320
Speaker 1: and how he goes about fixing hours so that your

28
00:01:37,400 --> 00:01:40,479
Speaker 1: listening experience is so much better that it would be

29
00:01:40,520 --> 00:01:42,720
Speaker 1: if you were a fly on the wall while Daniel

30
00:01:42,720 --> 00:02:00,480
Speaker 1: and I record. Welcome to Daniel and Kelly's Edited Universe. Hi.

31
00:02:00,600 --> 00:02:03,120
Speaker 2: I'm Daniel. I'm a particle physicist and I like thinking

32
00:02:03,160 --> 00:02:06,000
Speaker 2: about aliens, and I've often been told I have a

33
00:02:06,120 --> 00:02:07,160
Speaker 2: face for radio.

34
00:02:13,080 --> 00:02:16,560
Speaker 1: Hi. I'm Kelly leader Smith. I study parasites in space.

35
00:02:17,600 --> 00:02:20,800
Speaker 1: I often make jokes about having a face for radio.

36
00:02:20,880 --> 00:02:22,960
Speaker 1: But I think actually Daniel and I are but quite

37
00:02:23,000 --> 00:02:27,160
Speaker 1: attractive human beings. Oh yes, and Daniel has a great

38
00:02:27,280 --> 00:02:29,960
Speaker 1: voice for radio, So let's focus on the positive.

39
00:02:29,840 --> 00:02:32,040
Speaker 2: As do you, Kelly. I think our voices work really

40
00:02:32,080 --> 00:02:32,640
Speaker 2: well together.

41
00:02:32,919 --> 00:02:34,919
Speaker 1: I appreciate it. But I will note that when we

42
00:02:35,320 --> 00:02:39,120
Speaker 1: transitioned from Jorge on the show to Kelly on the Show,

43
00:02:39,480 --> 00:02:42,880
Speaker 1: we did get a fair number of folks writing in

44
00:02:42,919 --> 00:02:45,799
Speaker 1: to complain about the sound of my voice. But here

45
00:02:45,880 --> 00:02:49,680
Speaker 1: I am over one hundred episodes later. Kiss my butt, loosers.

46
00:02:49,840 --> 00:02:53,160
Speaker 1: I'm not going anywhere hater's gonna hate.

47
00:02:53,520 --> 00:02:55,400
Speaker 2: But we also got a lot of comments from people

48
00:02:55,440 --> 00:02:57,280
Speaker 2: who love your voice. And I think it's just a

49
00:02:57,320 --> 00:03:00,240
Speaker 2: subjective thing, you know, And it's a big part of

50
00:03:00,320 --> 00:03:02,760
Speaker 2: finding a podcast you'd like listening to. Is are the

51
00:03:02,880 --> 00:03:05,800
Speaker 2: voices pleasant and soothing? And that's very personal?

52
00:03:06,000 --> 00:03:08,440
Speaker 1: You know, That's right. But the good news is you

53
00:03:08,520 --> 00:03:11,840
Speaker 1: and I have Matt Kesselman to help us whenever our

54
00:03:11,919 --> 00:03:14,200
Speaker 1: voices are sounding less than their best.

55
00:03:15,760 --> 00:03:17,440
Speaker 2: That's right. And I want my voice to be smooth,

56
00:03:17,480 --> 00:03:18,880
Speaker 2: because a lot of folks write in and tell me

57
00:03:19,120 --> 00:03:22,120
Speaker 2: they like to fall asleep to my voice, and hey,

58
00:03:22,160 --> 00:03:25,080
Speaker 2: you know, I don't want to disturb anybody's slumber. No.

59
00:03:25,280 --> 00:03:28,040
Speaker 1: I wonder what my Calkuli laughed as to their dreams.

60
00:03:28,560 --> 00:03:32,280
Speaker 1: I dream of witches. But today we decided that we

61
00:03:32,280 --> 00:03:34,760
Speaker 1: were going to invite Matt Kesselman, who is just an

62
00:03:34,800 --> 00:03:38,240
Speaker 1: absolutely amazing audio engineer who, thankfully for us, spends a

63
00:03:38,280 --> 00:03:40,280
Speaker 1: lot of time working on our show. And part of

64
00:03:40,280 --> 00:03:42,480
Speaker 1: why we decided to invite him is because anyone who

65
00:03:42,600 --> 00:03:46,080
Speaker 1: submits a question that ends up on our Listener Questions

66
00:03:46,120 --> 00:03:49,760
Speaker 1: episode gets a raw audio file of Daniel and I

67
00:03:49,880 --> 00:03:52,320
Speaker 1: having our conversation, and we send it to them before

68
00:03:52,440 --> 00:03:55,120
Speaker 1: it goes to Matt, our audio engineer, so that they

69
00:03:55,160 --> 00:03:57,400
Speaker 1: have a chance to record themselves asking the question and

70
00:03:57,440 --> 00:04:02,080
Speaker 1: record themselves responding to our response to their question, and

71
00:04:02,120 --> 00:04:03,480
Speaker 1: then we can send all of that to Matt and

72
00:04:03,520 --> 00:04:05,880
Speaker 1: he can edit it all at once. And people often

73
00:04:05,960 --> 00:04:10,000
Speaker 1: respond to us and say, wow, your audio engineer has

74
00:04:10,040 --> 00:04:10,560
Speaker 1: a lot.

75
00:04:10,400 --> 00:04:14,880
Speaker 2: Of work to do, but not just that. Also they

76
00:04:14,960 --> 00:04:18,520
Speaker 2: say that it's really interesting to hear the unedited version

77
00:04:18,560 --> 00:04:21,680
Speaker 2: of the conversation because they can hear our side conversations

78
00:04:21,760 --> 00:04:24,080
Speaker 2: or when we pause to do research, or when we

79
00:04:24,120 --> 00:04:26,760
Speaker 2: back up to say something again, so they get like

80
00:04:26,760 --> 00:04:29,320
Speaker 2: a little glimpse behind the curtain. And so we want

81
00:04:29,360 --> 00:04:31,080
Speaker 2: to provide that to all of you today.

82
00:04:31,240 --> 00:04:33,800
Speaker 1: That's right, and so in the second segment of the show,

83
00:04:33,839 --> 00:04:35,560
Speaker 1: Matt is going to come on and talk to you

84
00:04:35,640 --> 00:04:38,440
Speaker 1: about the magic that he works on our episode, and

85
00:04:38,480 --> 00:04:41,640
Speaker 1: he'll play a couple raw clips so you'll get a

86
00:04:41,640 --> 00:04:43,919
Speaker 1: peek behind the curtain and see how Daniel and I

87
00:04:44,440 --> 00:04:49,400
Speaker 1: sound before Matt steps in. And Matt sent a question

88
00:04:49,920 --> 00:04:53,039
Speaker 1: for us to share with the listeners, and the question

89
00:04:53,560 --> 00:04:57,719
Speaker 1: was what kind of sound do humans hear best? So

90
00:04:57,800 --> 00:05:00,720
Speaker 1: let's go ahead and hear what the extraordinaries to say.

91
00:05:01,040 --> 00:05:05,040
Speaker 3: Maybe like the middle sounds like not super high or

92
00:05:05,040 --> 00:05:05,560
Speaker 3: super low.

93
00:05:05,640 --> 00:05:07,280
Speaker 1: That's what I was gonna say. But the humans are

94
00:05:07,279 --> 00:05:07,760
Speaker 1: in the middle.

95
00:05:08,080 --> 00:05:11,880
Speaker 2: I don't know what middle of what though, depending on

96
00:05:11,920 --> 00:05:16,520
Speaker 2: their gender, Like women would hear babies crying the best,

97
00:05:16,520 --> 00:05:20,360
Speaker 2: because I've heard stories that they often are moiasily working

98
00:05:20,440 --> 00:05:24,360
Speaker 2: up by that related to our size, Larger animals can

99
00:05:24,480 --> 00:05:28,400
Speaker 2: hear sounds with longer wavelengths. Smaller animals can hear sounds

100
00:05:28,440 --> 00:05:32,720
Speaker 2: with shorter wavelength high beach that sounds. Females tend to

101
00:05:32,760 --> 00:05:37,800
Speaker 2: hear higher frequencies better, possibly to help them hear their

102
00:05:37,839 --> 00:05:41,200
Speaker 2: offspring when they're crying or something similar.

103
00:05:41,560 --> 00:05:45,000
Speaker 1: Sounds that already stand out from the background and from

104
00:05:45,080 --> 00:05:45,560
Speaker 1: each other.

105
00:05:45,920 --> 00:05:48,120
Speaker 2: As a new mom, I want to say a baby crying.

106
00:05:48,520 --> 00:05:51,440
Speaker 2: These are great guesses and hilarious stories. Thank you very

107
00:05:51,520 --> 00:05:54,200
Speaker 2: much exterdinaries, and they're basically what I would have guessed.

108
00:05:54,240 --> 00:05:56,120
Speaker 2: You know that there's some range we can hear and

109
00:05:56,160 --> 00:05:59,400
Speaker 2: obviously ranges we can't hear, and that's going to be

110
00:05:59,480 --> 00:06:00,880
Speaker 2: tuned to our survival.

111
00:06:01,200 --> 00:06:04,560
Speaker 1: Yes, biology is the answer, and that's great. That's how

112
00:06:04,600 --> 00:06:06,080
Speaker 1: we know that. It's an interesting question.

113
00:06:06,279 --> 00:06:09,440
Speaker 2: No, it's a harmonious blend of biology and physics. It's

114
00:06:09,480 --> 00:06:11,480
Speaker 2: biology taking advantage of.

115
00:06:11,520 --> 00:06:16,080
Speaker 1: Physics, biophysics biofirst, but yes, biophysics.

116
00:06:17,920 --> 00:06:20,840
Speaker 2: You know, I think if biophysics means biology, it's modifying physics.

117
00:06:20,880 --> 00:06:22,839
Speaker 2: So it's at its core physics.

118
00:06:23,480 --> 00:06:24,880
Speaker 1: Well, you know, you and I could be at this

119
00:06:24,960 --> 00:06:28,120
Speaker 1: all day, and so we thought before we invited Matt

120
00:06:28,160 --> 00:06:30,040
Speaker 1: onto the show, we would let Daniel do a little

121
00:06:30,040 --> 00:06:33,120
Speaker 1: bit of his physics magic and talk about the physics

122
00:06:33,200 --> 00:06:35,440
Speaker 1: of sound. So we're going to dig into that a

123
00:06:35,440 --> 00:06:39,480
Speaker 1: little bit, and then we are going to bring Matt on. So, Daniel,

124
00:06:39,720 --> 00:06:40,680
Speaker 1: what is sound?

125
00:06:41,360 --> 00:06:44,440
Speaker 2: Yeah, so sound is not a thing. It's like an

126
00:06:44,640 --> 00:06:48,400
Speaker 2: arrangement of stuff. It's a traveling pattern of pressure changes

127
00:06:48,920 --> 00:06:54,080
Speaker 2: in matter. So if something is vibrating like a guitar string,

128
00:06:54,560 --> 00:06:58,080
Speaker 2: it's going to push and pull on nearby molecules creating

129
00:06:58,200 --> 00:07:01,720
Speaker 2: regions of compression and the regions when things are less dense,

130
00:07:02,320 --> 00:07:05,240
Speaker 2: and so those pressure changes then move through the medium.

131
00:07:05,320 --> 00:07:08,520
Speaker 2: Like if you slap your hand on the table, you're

132
00:07:08,520 --> 00:07:12,520
Speaker 2: creating momentary high pressure in the molecules on that table,

133
00:07:12,560 --> 00:07:13,960
Speaker 2: and they push on the ones next to it, and

134
00:07:13,960 --> 00:07:15,680
Speaker 2: they push on the ones next to it, so that

135
00:07:15,680 --> 00:07:19,000
Speaker 2: pressure wave travels right. And so this is a wave

136
00:07:19,080 --> 00:07:22,240
Speaker 2: that needs a medium. It moves through a medium. It's

137
00:07:22,240 --> 00:07:27,160
Speaker 2: a rearrangement through time of the relationship between the molecules

138
00:07:27,200 --> 00:07:27,880
Speaker 2: in the medium.

139
00:07:28,280 --> 00:07:32,160
Speaker 1: So I'm feeling even more amazed now by the microphone

140
00:07:32,200 --> 00:07:37,120
Speaker 1: that's sitting on my desk. How does those like squishiness

141
00:07:37,360 --> 00:07:41,720
Speaker 1: of the air get transferred into something my microphone can

142
00:07:41,880 --> 00:07:42,640
Speaker 1: play back to you.

143
00:07:43,240 --> 00:07:46,800
Speaker 2: Yes, so a speaker and a microphone are basically the

144
00:07:46,800 --> 00:07:49,400
Speaker 2: inverse of each other, right, So let's do a speaker first.

145
00:07:49,920 --> 00:07:53,640
Speaker 2: A speaker creates the sound. So there's something inside the speaker,

146
00:07:53,800 --> 00:07:56,920
Speaker 2: like a cone that moves back and forth. It's literally

147
00:07:56,960 --> 00:08:00,960
Speaker 2: physically mechanically moving at the frequent you want to create.

148
00:08:01,200 --> 00:08:04,160
Speaker 2: So we think about sound in terms of frequencies the

149
00:08:04,160 --> 00:08:06,800
Speaker 2: same way we think about light in terms of frequencies

150
00:08:07,120 --> 00:08:09,240
Speaker 2: you can have like red light and blue light or

151
00:08:09,280 --> 00:08:12,600
Speaker 2: green light. They are all vibrations of the electromagnetic field

152
00:08:12,640 --> 00:08:15,920
Speaker 2: at different frequencies. Sound has frequencies also, you can have

153
00:08:16,000 --> 00:08:20,760
Speaker 2: low frequencies and high frequencies, Daniel frequencies and Kelly frequencies exactly.

154
00:08:21,360 --> 00:08:25,120
Speaker 2: And so a speaker can make these by shaking effectively

155
00:08:25,120 --> 00:08:28,480
Speaker 2: a drum or a cone at that frequency. So if

156
00:08:28,520 --> 00:08:31,720
Speaker 2: you shake the cone at four hundred and forty hertz,

157
00:08:31,760 --> 00:08:34,480
Speaker 2: it's going to create four hundred and forty hertz sound,

158
00:08:34,960 --> 00:08:37,600
Speaker 2: just the same way that like a radio antenna oscillates

159
00:08:37,640 --> 00:08:41,439
Speaker 2: electrons at a certain frequency and creates photons or electromagnetic

160
00:08:41,440 --> 00:08:44,880
Speaker 2: waves at that frequency. So you have mechanical vibration. In

161
00:08:44,920 --> 00:08:47,959
Speaker 2: the case of sound transmits pressure waves to the air.

162
00:08:48,559 --> 00:08:52,000
Speaker 2: The microphone is essentially inverse pressure waves in the air

163
00:08:52,400 --> 00:08:56,040
Speaker 2: vibrates something sensitive, So you're converting pressure waves in the

164
00:08:56,040 --> 00:08:59,400
Speaker 2: air to physical vibrations of something, and then you just

165
00:08:59,440 --> 00:09:02,880
Speaker 2: need a divide that's going to record those vibrations. For example,

166
00:09:02,960 --> 00:09:06,160
Speaker 2: you could transmit them onto vinyl, or you could digitize

167
00:09:06,160 --> 00:09:08,800
Speaker 2: them and store them on a computer. So a microphone

168
00:09:08,880 --> 00:09:10,960
Speaker 2: or a speaker essentially inverses of each other.

169
00:09:11,480 --> 00:09:14,440
Speaker 1: Okay, I'm still amazed that we figured out how to

170
00:09:14,440 --> 00:09:16,040
Speaker 1: do that. That's that is pretty cool.

171
00:09:16,120 --> 00:09:17,520
Speaker 2: It is pretty cool. And there's a lot of really

172
00:09:17,559 --> 00:09:20,880
Speaker 2: fascinating physics of sound, like how the speed of sound

173
00:09:21,320 --> 00:09:24,840
Speaker 2: depends on how stiff and dense the material is. So,

174
00:09:25,040 --> 00:09:27,679
Speaker 2: in air, sound travels about three hundred and forty meters

175
00:09:27,720 --> 00:09:31,520
Speaker 2: per second. In water, it's faster because water is denser

176
00:09:32,000 --> 00:09:34,760
Speaker 2: and so molecules are closer together they could push up

177
00:09:34,760 --> 00:09:38,040
Speaker 2: against each other easier. It's like fifteen hundred meters per second,

178
00:09:38,440 --> 00:09:41,160
Speaker 2: and stel it's several kilometers per second.

179
00:09:41,240 --> 00:09:41,520
Speaker 3: Wow.

180
00:09:41,559 --> 00:09:44,640
Speaker 2: So sound travels faster through the ground. So you know

181
00:09:44,679 --> 00:09:47,360
Speaker 2: those scenes where somebody's like putting their ear on the

182
00:09:47,360 --> 00:09:49,680
Speaker 2: ground to hear if the buffalo are coming, Like, there's

183
00:09:49,720 --> 00:09:50,600
Speaker 2: real physics there.

184
00:09:50,880 --> 00:09:51,840
Speaker 1: WHOA, that's real.

185
00:09:53,040 --> 00:09:57,120
Speaker 2: That's real exactly. And so the things to understand about

186
00:09:57,120 --> 00:09:59,920
Speaker 2: sound is that there's frequencies, right, so hind ones and

187
00:10:00,080 --> 00:10:03,439
Speaker 2: low ones. There's also amplitude, right, is it loud or

188
00:10:03,559 --> 00:10:06,200
Speaker 2: is it quiet? Those are the crucial things to understand.

189
00:10:06,200 --> 00:10:09,600
Speaker 2: But musical sounds or voices are much more complicated than

190
00:10:09,679 --> 00:10:12,800
Speaker 2: just like a single frequency. Right, If you ask somebody

191
00:10:12,840 --> 00:10:15,920
Speaker 2: with a violin to play a certain note that's going

192
00:10:15,960 --> 00:10:18,960
Speaker 2: to correspond to a certain frequency, but a violin doesn't

193
00:10:19,000 --> 00:10:22,760
Speaker 2: produce sound that exactly one frequency. So if you want

194
00:10:22,760 --> 00:10:25,280
Speaker 2: to play four hundred and forty hertz, then a violin's

195
00:10:25,360 --> 00:10:27,839
Speaker 2: also going to play a harmonic which means eight hundred

196
00:10:27,880 --> 00:10:31,320
Speaker 2: and eighty hertz, which sounds like another version of the

197
00:10:31,360 --> 00:10:33,959
Speaker 2: same note. You know how somebody plays like the low

198
00:10:34,000 --> 00:10:36,520
Speaker 2: e and a high e and a guitar. They sound

199
00:10:36,640 --> 00:10:40,679
Speaker 2: nice together because they beat together, they're not shifted, but

200
00:10:40,679 --> 00:10:43,760
Speaker 2: there are different notes, and so a violin doesn't just

201
00:10:43,840 --> 00:10:45,880
Speaker 2: play the four to forty. It also plays the eight

202
00:10:45,960 --> 00:10:49,160
Speaker 2: eighty and the thirteen twenty and the seventeen sixty. And

203
00:10:49,240 --> 00:10:51,840
Speaker 2: if you play a guitar or a trombone with the

204
00:10:51,880 --> 00:10:55,080
Speaker 2: same note, it has a different relationship with those harmonics.

205
00:10:55,480 --> 00:10:57,760
Speaker 2: So you can hear a violin is different from a

206
00:10:57,760 --> 00:11:04,599
Speaker 2: trombone even when they're playing the same note. Why is

207
00:11:04,640 --> 00:11:06,680
Speaker 2: that Because they have a different set of frequencies that

208
00:11:06,720 --> 00:11:09,640
Speaker 2: they're playing. It's not a single pure frequency. And this

209
00:11:09,679 --> 00:11:12,560
Speaker 2: is what we call timber. This is why, like my

210
00:11:12,720 --> 00:11:15,120
Speaker 2: voice sounds different from somebody else's voice, even if it

211
00:11:15,160 --> 00:11:17,360
Speaker 2: was at the same frequency, like if you took Kelly's

212
00:11:17,440 --> 00:11:20,240
Speaker 2: voice and you shifted it down to Daniel range, oh no,

213
00:11:20,280 --> 00:11:22,160
Speaker 2: it would not sound like Daniels.

214
00:11:23,480 --> 00:11:26,560
Speaker 1: And that this was all new to me like forty

215
00:11:26,600 --> 00:11:28,440
Speaker 1: minutes ago when we talked to Matt, and so like

216
00:11:28,480 --> 00:11:30,720
Speaker 1: you know, heads up, Matt is going to mention this

217
00:11:30,760 --> 00:11:32,640
Speaker 1: a little bit and how he can see that in

218
00:11:32,679 --> 00:11:35,480
Speaker 1: the software that he uses, and that totally blew my mind.

219
00:11:35,520 --> 00:11:38,240
Speaker 1: I didn't realize that you were getting like harmonics of

220
00:11:38,280 --> 00:11:40,520
Speaker 1: people's voices and anyway, very cool, but.

221
00:11:40,559 --> 00:11:43,280
Speaker 2: There's also lots of fascinating biology here, like how the

222
00:11:43,360 --> 00:11:46,440
Speaker 2: human voice is actually generated. Right, you don't have like

223
00:11:46,440 --> 00:11:49,240
Speaker 2: a little violin in your throat, but you do have

224
00:11:49,360 --> 00:11:52,760
Speaker 2: things that vibrate. You have the vocal chords, or sometimes

225
00:11:52,840 --> 00:11:56,120
Speaker 2: called the vocal folds. They're not like guitar strings. They're

226
00:11:56,160 --> 00:11:59,680
Speaker 2: more like vibrating valves in an airflow and faster vibration.

227
00:12:00,040 --> 00:12:03,000
Speaker 2: It's a higher pitch, and so like adult male voices

228
00:12:03,040 --> 00:12:06,400
Speaker 2: are lower because they have longer, thicker vocal folds that

229
00:12:06,520 --> 00:12:09,880
Speaker 2: vibrate more slowly. And so this is why also two

230
00:12:09,920 --> 00:12:12,120
Speaker 2: people sing the same note but sound.

231
00:12:11,840 --> 00:12:15,960
Speaker 1: Different, and birds are somehow able to make at least

232
00:12:16,040 --> 00:12:19,240
Speaker 1: two noises at the same time, some birds, not all birds,

233
00:12:19,280 --> 00:12:21,640
Speaker 1: which is just absolutely like if you you should really

234
00:12:21,679 --> 00:12:23,360
Speaker 1: pay attention to the next time you hear a bird

235
00:12:23,440 --> 00:12:25,360
Speaker 1: song and try to figure out, like, wait a minute,

236
00:12:25,360 --> 00:12:28,160
Speaker 1: am I actually hearing that bird make two different notes

237
00:12:28,600 --> 00:12:32,520
Speaker 1: at the same time? And it's able to do that,

238
00:12:32,559 --> 00:12:34,640
Speaker 1: and anyway, sound is incredible and.

239
00:12:34,600 --> 00:12:36,360
Speaker 2: Some humans can do that too, right, this is this

240
00:12:36,440 --> 00:12:39,040
Speaker 2: like Himalayan dual note singing thing.

241
00:12:42,880 --> 00:12:45,360
Speaker 1: Oh, I didn't know that. I know that birds have

242
00:12:45,400 --> 00:12:48,400
Speaker 1: a very different like voice box than we do, so

243
00:12:48,440 --> 00:12:50,760
Speaker 1: that they can do that, and so now I need

244
00:12:50,800 --> 00:12:52,920
Speaker 1: to know what happens with the Himalayan people.

245
00:12:53,280 --> 00:12:55,680
Speaker 2: It's very hard to do, but you can train yourself

246
00:12:55,720 --> 00:12:58,920
Speaker 2: apparently to do that. And sound is much more complicated

247
00:12:58,960 --> 00:13:02,079
Speaker 2: than just like generate sound. Here a sound, you can

248
00:13:02,120 --> 00:13:05,280
Speaker 2: have interference effects because sound is a wave, and so

249
00:13:05,400 --> 00:13:08,560
Speaker 2: places where the pressure waves are in the same direction

250
00:13:08,679 --> 00:13:12,439
Speaker 2: will be louder, that's constructive interference, and places where they're

251
00:13:12,480 --> 00:13:15,240
Speaker 2: pushing in opposite directions it'll be weaker. And this is

252
00:13:15,280 --> 00:13:20,000
Speaker 2: how like noise cancelation works. Right, you play the opposite sound,

253
00:13:20,440 --> 00:13:22,400
Speaker 2: the one that's beating down when the other sound is

254
00:13:22,440 --> 00:13:25,400
Speaker 2: beating up and it cancels it out. So for example,

255
00:13:25,480 --> 00:13:29,680
Speaker 2: my daughter has air pods in and I say, hey, Hazel,

256
00:13:29,679 --> 00:13:32,080
Speaker 2: will you take out the trash? And her AirPod plays

257
00:13:32,120 --> 00:13:34,199
Speaker 2: the opposite of Hey Hazel, will you take out the trash?

258
00:13:34,240 --> 00:13:37,040
Speaker 2: And she hears nothing right, which is why she doesn't

259
00:13:37,040 --> 00:13:40,360
Speaker 2: take up the trash. Perfectly hypothetical example.

260
00:13:41,200 --> 00:13:44,720
Speaker 1: However, I would like to note that Matt mentions later

261
00:13:44,760 --> 00:13:48,040
Speaker 1: in the episode that noise canceling headphones don't necessarily cancel

262
00:13:48,080 --> 00:13:50,280
Speaker 1: out people talking to you. Yeah, I don't want to

263
00:13:50,280 --> 00:13:53,960
Speaker 1: get Hazel in trouble, but I'm wondering if maybe Hazel's

264
00:13:54,040 --> 00:13:54,760
Speaker 1: just not listening.

265
00:13:54,880 --> 00:13:56,960
Speaker 2: Yeah, there may be a few effects going on here.

266
00:13:57,679 --> 00:14:01,000
Speaker 2: But this goes into like the design of concert right,

267
00:14:01,080 --> 00:14:03,600
Speaker 2: because sound reflects and it can cancel out. So you

268
00:14:03,640 --> 00:14:06,679
Speaker 2: can have like dead spots in a room where you

269
00:14:06,720 --> 00:14:09,320
Speaker 2: can't hear someone from across the room, or places where

270
00:14:09,360 --> 00:14:12,520
Speaker 2: the sound all balances really really well. And this makes

271
00:14:12,559 --> 00:14:14,800
Speaker 2: a big difference between like a bad concert hall and

272
00:14:14,840 --> 00:14:17,439
Speaker 2: a good concert hall where like every seat sounds good

273
00:14:17,600 --> 00:14:21,320
Speaker 2: or more seats sound really good. You could also have

274
00:14:21,400 --> 00:14:23,880
Speaker 2: beating sounds, like if you have two sounds, they're very

275
00:14:23,920 --> 00:14:26,560
Speaker 2: similar to each other very close, but not quite on

276
00:14:26,640 --> 00:14:28,880
Speaker 2: top of each other. Like if you play your guitar

277
00:14:28,960 --> 00:14:30,840
Speaker 2: tuner at e and then you play your guitar but

278
00:14:30,880 --> 00:14:40,280
Speaker 2: it's not quite in tune. You'll hear this, Like, those

279
00:14:40,280 --> 00:14:43,600
Speaker 2: are the waves drifting out of alignment, right, you'll hear

280
00:14:43,640 --> 00:14:47,200
Speaker 2: them because they're not beating in sync harmonic like for

281
00:14:47,200 --> 00:14:50,240
Speaker 2: forty and a eighty. They have the down moments at

282
00:14:50,240 --> 00:14:52,280
Speaker 2: the same time, and so they sound nice.

283
00:14:52,920 --> 00:14:54,640
Speaker 1: I miss that. Could you make that sound again?

284
00:14:57,360 --> 00:14:59,080
Speaker 2: Mac can make that your ring toney for you if

285
00:14:59,120 --> 00:14:59,440
Speaker 2: you like.

286
00:15:00,960 --> 00:15:01,680
Speaker 1: That would be great.

287
00:15:02,480 --> 00:15:04,120
Speaker 2: And then there's a lot of work that needs to

288
00:15:04,160 --> 00:15:06,640
Speaker 2: be done when you're like doing noise removal. And Matt's

289
00:15:06,640 --> 00:15:08,880
Speaker 2: gonna tell us all about that when he gets here.

290
00:15:09,080 --> 00:15:12,920
Speaker 1: All right, so let us delay no longer, actually quick

291
00:15:12,960 --> 00:15:16,240
Speaker 1: delay we're gonna do. We're gonna do a commercial break,

292
00:15:16,320 --> 00:15:18,520
Speaker 1: and then when we come back, we will bring Matt

293
00:15:18,560 --> 00:15:19,000
Speaker 1: on the show.

294
00:15:19,080 --> 00:15:22,160
Speaker 2: The Great Matt Kesselman.

295
00:15:41,760 --> 00:15:44,480
Speaker 1: Matt Kesselman is an audio engineer and producer with over

296
00:15:44,560 --> 00:15:48,560
Speaker 1: two decades of multidisciplinary experience in the fields of film production,

297
00:15:49,080 --> 00:15:52,840
Speaker 1: audio post production, and music. Matt has worked with artists

298
00:15:52,880 --> 00:15:57,520
Speaker 1: and broadcasters like Netflix, the BBC, Lionsgate, Walk Off the Earth,

299
00:15:57,600 --> 00:16:01,120
Speaker 1: Ill Scarlet and many others. He developed software tools for

300
00:16:01,240 --> 00:16:05,360
Speaker 1: audio engineers through his company Oslot Audio. He brings over

301
00:16:05,440 --> 00:16:08,160
Speaker 1: a decade of teaching and communication experience to the table

302
00:16:08,200 --> 00:16:11,080
Speaker 1: as a college professor at Metalworks Institute, where he is

303
00:16:11,160 --> 00:16:14,280
Speaker 1: currently and before that at Durham College. He's also the

304
00:16:14,320 --> 00:16:17,880
Speaker 1: amazing audio engineer and sound designer behind dk EU. He's

305
00:16:17,920 --> 00:16:21,040
Speaker 1: the guy who makes us sound not horrible. Thanks for

306
00:16:21,120 --> 00:16:21,920
Speaker 1: joining us, Matt.

307
00:16:22,080 --> 00:16:24,440
Speaker 2: He's also the guy who has a shocking amount of

308
00:16:24,440 --> 00:16:26,000
Speaker 2: blackmail material on both of us.

309
00:16:26,040 --> 00:16:30,080
Speaker 3: That's right, that's all true. I'm a little nervous. I

310
00:16:30,120 --> 00:16:33,120
Speaker 3: feel like am I the first guest that doesn't have

311
00:16:33,160 --> 00:16:36,840
Speaker 3: a PhD? Oh not even in like chemistry.

312
00:16:37,480 --> 00:16:39,600
Speaker 2: Wow, I didn't realize we were gate keeping the guests.

313
00:16:39,640 --> 00:16:41,040
Speaker 1: That's right, Oh my gosh.

314
00:16:41,200 --> 00:16:43,920
Speaker 3: I mean you folks try to find the secrets to

315
00:16:44,000 --> 00:16:47,080
Speaker 3: life and how to better all of humankind. And I

316
00:16:47,200 --> 00:16:50,720
Speaker 3: once produced a song called Candy g String. So we're

317
00:16:50,720 --> 00:16:53,840
Speaker 3: a little different. So I'll try to cold man of

318
00:16:53,840 --> 00:16:55,120
Speaker 3: the bargain and make this entertaining.

319
00:16:55,240 --> 00:16:58,240
Speaker 2: Well, that's what makes us a harmonious collaboration. Everybody brings

320
00:16:58,240 --> 00:16:58,960
Speaker 2: their strengths.

321
00:16:59,200 --> 00:16:59,680
Speaker 3: That's right.

322
00:17:00,640 --> 00:17:03,200
Speaker 2: Well, maybe we should start actually by telling the story

323
00:17:03,280 --> 00:17:05,639
Speaker 2: of how we got connected to you, Matt, how you

324
00:17:05,800 --> 00:17:08,280
Speaker 2: ended up being the audio engineer for the pod.

325
00:17:09,280 --> 00:17:12,960
Speaker 3: Yeah, how indeed. Well, I was a fan for a

326
00:17:13,000 --> 00:17:17,400
Speaker 3: long time, and I think, like all the other Extraordinaries,

327
00:17:17,600 --> 00:17:19,880
Speaker 3: I was listening before we were even called the Extraordinaries.

328
00:17:19,920 --> 00:17:22,000
Speaker 3: But I always have questions and I always want to

329
00:17:22,000 --> 00:17:26,640
Speaker 3: know how things work. And a lot of popular science.

330
00:17:27,160 --> 00:17:30,199
Speaker 3: I can only consume popular science because I am not

331
00:17:30,359 --> 00:17:32,159
Speaker 3: smart enough, but a lot of it, you know, I

332
00:17:32,160 --> 00:17:34,560
Speaker 3: don't believe. That leaves me wanting and I feel like

333
00:17:34,560 --> 00:17:37,560
Speaker 3: there's something missing, and then there's conflicting things in different

334
00:17:37,600 --> 00:17:40,320
Speaker 3: popular science publications. And then I found you guys, and

335
00:17:41,119 --> 00:17:43,520
Speaker 3: you break it down and that was really enjoyable. And

336
00:17:43,520 --> 00:17:45,760
Speaker 3: then I thought, why don't I just tell them that

337
00:17:46,000 --> 00:17:48,480
Speaker 3: I can make them sound even better?

338
00:17:50,440 --> 00:17:52,560
Speaker 2: Yeah, we got an email out of the blue from

339
00:17:52,600 --> 00:17:56,280
Speaker 2: you saying, hey, I have notes on your audio engineering,

340
00:17:56,880 --> 00:17:59,280
Speaker 2: and we dug into it with you, and then eventually

341
00:17:59,280 --> 00:18:01,520
Speaker 2: we were like, hey, why do you come make the

342
00:18:01,560 --> 00:18:02,200
Speaker 2: show better?

343
00:18:02,560 --> 00:18:06,600
Speaker 1: And we haven't looked back yet, and so one of

344
00:18:06,640 --> 00:18:08,600
Speaker 1: the reasons we're super excited to have you on the

345
00:18:08,600 --> 00:18:11,200
Speaker 1: show is because after Daniel and I do a listener

346
00:18:11,280 --> 00:18:15,600
Speaker 1: questions episode, we always send the raw audio file export

347
00:18:15,640 --> 00:18:17,840
Speaker 1: it from riverside out to the listeners so that they

348
00:18:17,880 --> 00:18:20,000
Speaker 1: have plenty of time to listen to our conversation and

349
00:18:20,040 --> 00:18:24,399
Speaker 1: send us an audio file responding. And almost every single

350
00:18:24,400 --> 00:18:29,680
Speaker 1: one of them is like, Wow, this sounds awful. Your

351
00:18:29,840 --> 00:18:35,119
Speaker 1: audio guy must do so much work, and so we

352
00:18:35,240 --> 00:18:36,720
Speaker 1: thought that it would be great to have you on

353
00:18:36,760 --> 00:18:39,200
Speaker 1: the show to like pull back the curtain and talk

354
00:18:39,240 --> 00:18:42,440
Speaker 1: about like when Kelly forgets to turn off her air conditioner,

355
00:18:42,480 --> 00:18:44,679
Speaker 1: how do you remove that sound, and like all of

356
00:18:44,720 --> 00:18:46,240
Speaker 1: the work that you have to do to make us

357
00:18:46,320 --> 00:18:49,240
Speaker 1: sound good. Where does that start?

358
00:18:49,440 --> 00:18:53,200
Speaker 3: Okay? Well, first of all, I want to point out

359
00:18:53,240 --> 00:18:56,399
Speaker 3: that there are different kinds of podcasts that have different needs. Right,

360
00:18:56,440 --> 00:19:01,119
Speaker 3: everybody's familiar with the comedians just talking with no editing,

361
00:19:01,160 --> 00:19:03,359
Speaker 3: with no nothing, and they leave all the mistakes in

362
00:19:03,400 --> 00:19:06,600
Speaker 3: and that's part of the charm. They're also narrative podcast

363
00:19:06,680 --> 00:19:08,440
Speaker 3: that tell a story kind of like a movie, and

364
00:19:08,560 --> 00:19:10,920
Speaker 3: our very own network has a few examples of that,

365
00:19:11,040 --> 00:19:14,880
Speaker 3: like Red Elvis and the buzz Aldrin Show. I don't

366
00:19:14,920 --> 00:19:17,320
Speaker 3: know if you've heard that one, but then there's also

367
00:19:17,359 --> 00:19:22,600
Speaker 3: the informative kind where you're basically enjoying a lecture. And

368
00:19:22,920 --> 00:19:26,000
Speaker 3: I feel like this is kind of like a lecture

369
00:19:26,440 --> 00:19:29,680
Speaker 3: mixed in with some lighthearted comedy. And the thing about

370
00:19:29,720 --> 00:19:32,400
Speaker 3: lectures is that they take a lot of time to prepare,

371
00:19:33,000 --> 00:19:36,080
Speaker 3: and you two do something that very few people can do,

372
00:19:36,119 --> 00:19:39,000
Speaker 3: and that is that you basically deliver two lectures a week,

373
00:19:39,440 --> 00:19:45,160
Speaker 3: often on topics that you're not specialized in. And it's

374
00:19:45,200 --> 00:19:48,160
Speaker 3: no surprise that sometimes you need to write on the fly.

375
00:19:48,920 --> 00:19:52,159
Speaker 3: I mean, otherwise, it's just not possible for any person

376
00:19:52,240 --> 00:19:55,439
Speaker 3: to deliver that much information without having to stop and

377
00:19:55,480 --> 00:19:58,280
Speaker 3: sort of rewind and think back on it. And I

378
00:19:58,280 --> 00:20:00,639
Speaker 3: think we all sort of treat it that way, where

379
00:20:01,000 --> 00:20:03,119
Speaker 3: you have your outline and you have your idea, and

380
00:20:03,160 --> 00:20:05,560
Speaker 3: sometimes you come up with a new idea while we record,

381
00:20:06,160 --> 00:20:09,520
Speaker 3: and then you try it out a few times until

382
00:20:09,520 --> 00:20:11,240
Speaker 3: you get it right, and when you send it off

383
00:20:11,240 --> 00:20:14,840
Speaker 3: and you hope that I put it together, I feel like.

384
00:20:14,800 --> 00:20:17,919
Speaker 1: You really you get me, Matt, you get me. And

385
00:20:17,960 --> 00:20:20,880
Speaker 1: I want to start describing the podcast now as lectures

386
00:20:20,920 --> 00:20:23,880
Speaker 1: with light comedy, because I feel like that's actually that's

387
00:20:23,920 --> 00:20:25,560
Speaker 1: a pretty good description.

388
00:20:25,320 --> 00:20:28,199
Speaker 3: That's really what it is, right, It's like, yeah, weird

389
00:20:28,400 --> 00:20:31,520
Speaker 3: things that are very hard to understand, sometimes broken down

390
00:20:31,560 --> 00:20:37,600
Speaker 3: into small, bite sized, understandable pieces with some pro lapse jokes,

391
00:20:40,440 --> 00:20:42,640
Speaker 3: which is my favorite type of entertainment.

392
00:20:43,720 --> 00:20:46,680
Speaker 2: We need constant goat updates. You know, people are curious.

393
00:20:47,359 --> 00:20:49,760
Speaker 1: Well, she's kept her outsides on the inside recently, so

394
00:20:49,840 --> 00:20:51,480
Speaker 1: that's nice, wonderful. Yeah.

395
00:20:51,520 --> 00:20:54,080
Speaker 2: Well, from my perspective, I lean heavily into this editing

396
00:20:54,520 --> 00:20:56,879
Speaker 2: because I know that we can go back and fix

397
00:20:56,920 --> 00:20:59,879
Speaker 2: something or say something a better way I speak to

398
00:21:00,040 --> 00:21:02,520
Speaker 2: friendly than I do in person, or if I'm like

399
00:21:02,600 --> 00:21:05,080
Speaker 2: lecturing to a class, I'm lecturing to a class. I

400
00:21:05,080 --> 00:21:07,240
Speaker 2: don't often back up and say things another way and

401
00:21:07,359 --> 00:21:09,560
Speaker 2: just go with it. So I think the listeners are

402
00:21:09,560 --> 00:21:13,159
Speaker 2: hearing a very unusual slice of our speaking, you know,

403
00:21:13,240 --> 00:21:17,000
Speaker 2: the intended for editing but unedited version, which is not

404
00:21:17,119 --> 00:21:19,040
Speaker 2: how like we have a conversation.

405
00:21:18,640 --> 00:21:22,320
Speaker 3: Normally, Right, And if the listeners want some inside baseball,

406
00:21:22,359 --> 00:21:25,240
Speaker 3: I always send you the finished episode or a draft

407
00:21:25,240 --> 00:21:28,000
Speaker 3: of the episode, and then you listen intently and go, oh,

408
00:21:28,040 --> 00:21:31,240
Speaker 3: actually that detail is technically right, but there's a better

409
00:21:31,280 --> 00:21:33,000
Speaker 3: way to say it, And then you send me a

410
00:21:33,040 --> 00:21:36,040
Speaker 3: pickup because you're perfectionists and you want everything to be

411
00:21:36,480 --> 00:21:39,560
Speaker 3: correct and true, and that's very admirable as well, and you're.

412
00:21:39,400 --> 00:21:41,600
Speaker 1: Always a really good sport when we're like, oh, I

413
00:21:41,880 --> 00:21:44,520
Speaker 1: actually thank you so much for editing all of that,

414
00:21:44,600 --> 00:21:45,840
Speaker 1: but I'd like to do it again.

415
00:21:46,560 --> 00:21:48,840
Speaker 2: So let's make this concrete. Can you show us some

416
00:21:48,880 --> 00:21:51,560
Speaker 2: examples of what you get and then what you send

417
00:21:51,600 --> 00:21:55,000
Speaker 2: us back. Show us the raw, unedited Daniel and Kelly

418
00:21:55,040 --> 00:21:58,320
Speaker 2: craziness and then the smooth Matt output.

419
00:21:58,640 --> 00:22:02,320
Speaker 3: Sure, I have two separate kinds of examples. Here's a

420
00:22:02,400 --> 00:22:06,400
Speaker 3: couple where the noise reduction was already done, and we're

421
00:22:06,440 --> 00:22:09,760
Speaker 3: going to focus on your writing on the fly, if

422
00:22:09,840 --> 00:22:13,000
Speaker 3: you will, so take a close listen to. We'll start

423
00:22:13,040 --> 00:22:18,760
Speaker 3: with with Kelly trying to communicate this idea, backing up

424
00:22:18,760 --> 00:22:21,000
Speaker 3: a couple of times, a couple of uhs and ums,

425
00:22:21,200 --> 00:22:24,560
Speaker 3: and then after we'll listen to the cut up version.

426
00:22:24,640 --> 00:22:26,800
Speaker 1: And I have no idea what file met pick. So

427
00:22:27,640 --> 00:22:31,840
Speaker 1: I'm on the edge of my seat, and we don't

428
00:22:31,840 --> 00:22:34,320
Speaker 1: really have a good framework yet for predicting. Like if

429
00:22:34,359 --> 00:22:38,439
Speaker 1: you took you know X, and say you decided you

430
00:22:38,440 --> 00:22:41,800
Speaker 1: were going to take what's a say you were going

431
00:22:41,840 --> 00:22:43,560
Speaker 1: to take lions and domesticate them.

432
00:22:44,640 --> 00:22:50,200
Speaker 3: Okay, so there's that. Here's here's the edited version.

433
00:22:51,840 --> 00:22:53,560
Speaker 1: Yeah right, And we don't really have a good framework

434
00:22:53,640 --> 00:22:56,200
Speaker 1: yet for predicting, like say you were going to take

435
00:22:56,280 --> 00:22:59,000
Speaker 1: lions and domesticate them. I'm being ridiculous, right, But like

436
00:22:59,040 --> 00:22:59,879
Speaker 1: say you decided.

437
00:22:59,600 --> 00:23:01,760
Speaker 2: To Really that sounds awesome. I would love to have

438
00:23:01,800 --> 00:23:06,600
Speaker 2: a house lion. Hey, whatever happened to the house lion

439
00:23:06,680 --> 00:23:07,200
Speaker 2: I requested?

440
00:23:07,240 --> 00:23:09,199
Speaker 1: Anyway, I need a few more years.

441
00:23:09,960 --> 00:23:12,800
Speaker 3: Yeah. So, I mean you can see that you got

442
00:23:12,840 --> 00:23:15,920
Speaker 3: this idea of comparing it to lines and it came

443
00:23:15,960 --> 00:23:18,639
Speaker 3: on the fly, so you just needed to work through it,

444
00:23:19,359 --> 00:23:22,080
Speaker 3: which is great because I think everybody would rather hear

445
00:23:22,680 --> 00:23:25,000
Speaker 3: what you intend for them to hear and not for

446
00:23:25,080 --> 00:23:27,359
Speaker 3: you to sort of back away and just say something

447
00:23:27,400 --> 00:23:28,359
Speaker 3: that's easier to say.

448
00:23:28,720 --> 00:23:31,200
Speaker 2: And I think that these on the fly moments are

449
00:23:31,240 --> 00:23:35,000
Speaker 2: what make our conversations really fun because it is a conversation.

450
00:23:35,119 --> 00:23:37,119
Speaker 2: You're thinking on the fly, you're coming up with new ideas,

451
00:23:37,160 --> 00:23:40,400
Speaker 2: you're responding to questions. Also, we could never pull off

452
00:23:40,400 --> 00:23:42,840
Speaker 2: a scripted podcast twice a week. That's just so much

453
00:23:42,880 --> 00:23:46,800
Speaker 2: more work. And on top of that, I'm a terrible actor,

454
00:23:47,240 --> 00:23:49,320
Speaker 2: So if I had to read a scriptive podcast, it

455
00:23:49,359 --> 00:23:52,159
Speaker 2: would sound really stilted. The only way for this to

456
00:23:52,200 --> 00:23:53,960
Speaker 2: work is for it to be a natural conversation.

457
00:23:54,320 --> 00:23:57,119
Speaker 1: I massively appreciate that I can like stumble and start

458
00:23:57,160 --> 00:24:00,280
Speaker 1: over many, many, many times and know Mad is going

459
00:24:00,320 --> 00:24:01,879
Speaker 1: to listen to all of that and pick out the

460
00:24:01,880 --> 00:24:04,840
Speaker 1: bits that make sense. And you're amazing.

461
00:24:05,119 --> 00:24:07,720
Speaker 3: Thank you. Sometimes you do put a lot of faith

462
00:24:07,760 --> 00:24:11,800
Speaker 3: in me. And just to be clear, what I mean

463
00:24:11,800 --> 00:24:13,800
Speaker 3: by that is that you'll have a great first half

464
00:24:13,840 --> 00:24:16,480
Speaker 3: of an idea, and then you might write on the

465
00:24:16,480 --> 00:24:18,959
Speaker 3: fly and work out the second half of the idea,

466
00:24:19,280 --> 00:24:22,640
Speaker 3: and then when I put them together, sometimes they could

467
00:24:22,720 --> 00:24:25,959
Speaker 3: sound a little bit disjointed, kind of like a milder

468
00:24:26,040 --> 00:24:29,560
Speaker 3: version of this real segment that really aired on TV.

469
00:24:30,440 --> 00:24:31,879
Speaker 2: An idea, but what we do, we would not be

470
00:24:31,920 --> 00:24:34,600
Speaker 2: good at what we do, would we We would be sloppy?

471
00:24:35,560 --> 00:24:36,600
Speaker 2: You calling it sloppy.

472
00:24:37,400 --> 00:24:39,520
Speaker 3: I don't want your recordings to sound like that, even

473
00:24:39,560 --> 00:24:44,160
Speaker 3: though it's completely normal for speech cadence to change over time,

474
00:24:44,200 --> 00:24:46,840
Speaker 3: but when you put it all together and remove the

475
00:24:46,880 --> 00:24:49,439
Speaker 3: gaps in the middle, it becomes more noticeable. So sometimes

476
00:24:49,480 --> 00:24:52,199
Speaker 3: I have to do things like tune your voice on

477
00:24:52,440 --> 00:24:56,959
Speaker 3: certain words, or look around and find other times when

478
00:24:56,960 --> 00:24:59,720
Speaker 3: you said that word in the episode or in other episodes. Wow,

479
00:25:00,160 --> 00:25:03,480
Speaker 3: use it to help build one cohesive sentence. And just

480
00:25:03,480 --> 00:25:06,760
Speaker 3: to be clear, I never change what you've said. I

481
00:25:06,840 --> 00:25:09,200
Speaker 3: might just change how it sounds, but the ideas are

482
00:25:09,359 --> 00:25:10,080
Speaker 3: purely yours.

483
00:25:10,440 --> 00:25:12,600
Speaker 1: How often does it do we do that to you?

484
00:25:12,920 --> 00:25:14,760
Speaker 1: I'm feeling really bad right now.

485
00:25:15,160 --> 00:25:17,959
Speaker 3: It's not something you do to me, but it probably

486
00:25:18,000 --> 00:25:19,240
Speaker 3: happens more often than you think.

487
00:25:19,359 --> 00:25:21,600
Speaker 2: Oh no, no.

488
00:25:21,840 --> 00:25:25,160
Speaker 3: Which is completely normal. You're focused on educating the masses.

489
00:25:25,400 --> 00:25:28,280
Speaker 3: I'm focused on making it sound smooth without anybody noticing

490
00:25:28,400 --> 00:25:29,360
Speaker 3: anything was done at all.

491
00:25:29,760 --> 00:25:32,199
Speaker 2: Well, thank you for making this sound so good and

492
00:25:32,240 --> 00:25:34,840
Speaker 2: making it sound like you haven't done anything. That's the

493
00:25:34,840 --> 00:25:35,720
Speaker 2: most flattering part.

494
00:25:35,760 --> 00:25:37,760
Speaker 3: You want to hear one of you, Daniel, Yes and no,

495
00:25:39,320 --> 00:25:41,880
Speaker 3: let's do it. Okay, here we go before and after,

496
00:25:41,920 --> 00:25:42,800
Speaker 3: For Daniel.

497
00:25:43,600 --> 00:25:46,040
Speaker 2: Had that one piece of data argue that the Earth

498
00:25:46,080 --> 00:25:49,240
Speaker 2: could be a flat disc that was perfectly that was

499
00:25:49,240 --> 00:25:55,280
Speaker 2: perfectly aligned to make a circular shadow. Yes, oh boy, that's.

500
00:25:55,200 --> 00:26:00,760
Speaker 3: Real for some reason, like were you at because you

501
00:26:00,800 --> 00:26:02,960
Speaker 3: hear that in the Listen again.

502
00:26:03,200 --> 00:26:05,640
Speaker 2: And that one piece of data argue that the Earth

503
00:26:05,680 --> 00:26:08,760
Speaker 2: could be a flat to hear that where Yeah, I

504
00:26:08,800 --> 00:26:10,719
Speaker 2: was not buying a can of olives or something at

505
00:26:10,760 --> 00:26:13,240
Speaker 2: the same time as recording the podcast. I've no idea

506
00:26:13,280 --> 00:26:17,399
Speaker 2: where that comes from. So here's the after Yeah, that

507
00:26:17,440 --> 00:26:19,840
Speaker 2: one piece of data argue that the Earth could be

508
00:26:19,880 --> 00:26:22,399
Speaker 2: a flat disc that was perfectly aligned to make a

509
00:26:22,400 --> 00:26:27,439
Speaker 2: circular shadow. Yes, I sound great afterwards, thank you.

510
00:26:28,520 --> 00:26:31,840
Speaker 3: Yeah, and again, like you said, you don't talk like

511
00:26:31,880 --> 00:26:34,760
Speaker 3: that in person. It's just that you're trying to communicate

512
00:26:34,920 --> 00:26:37,399
Speaker 3: things that are new to you in some way and

513
00:26:37,440 --> 00:26:39,440
Speaker 3: in a way that fits for broadcast.

514
00:26:39,760 --> 00:26:42,840
Speaker 1: Okay, So when I listened to the audio files after

515
00:26:42,880 --> 00:26:44,879
Speaker 1: I send them to you, I'm like, oh, shoot, I

516
00:26:44,960 --> 00:26:48,000
Speaker 1: left my air conditioner on. Or Daniel was trying to

517
00:26:48,040 --> 00:26:51,480
Speaker 1: explain general relativity at the grocery store, like how do

518
00:26:51,680 --> 00:26:55,800
Speaker 1: you remove like the beeps and the background noises? Like

519
00:26:55,880 --> 00:26:56,720
Speaker 1: how hard is that?

520
00:26:57,480 --> 00:27:00,680
Speaker 3: So there are many different types of noises in many

521
00:27:00,720 --> 00:27:05,120
Speaker 3: different solutions. Some are automated, some are not. So if

522
00:27:05,160 --> 00:27:10,479
Speaker 3: it's here, I have some examples. Here is some unprocessed Daniel,

523
00:27:10,560 --> 00:27:12,280
Speaker 3: let me know if you can hear that background noise.

524
00:27:13,880 --> 00:27:18,120
Speaker 2: So Bell's paradox involves length contraction, the fact that when

525
00:27:18,400 --> 00:27:22,159
Speaker 2: things move fast, they look short. So relativity tells us

526
00:27:22,160 --> 00:27:26,960
Speaker 2: two things, moving clocks run slow and moving objects look short.

527
00:27:28,760 --> 00:27:31,240
Speaker 3: So that kind of noise is probably the easiest because

528
00:27:31,280 --> 00:27:31,960
Speaker 3: it's consistent.

529
00:27:32,359 --> 00:27:34,679
Speaker 2: It sounds like I'm crinkling paper or something.

530
00:27:35,040 --> 00:27:37,399
Speaker 3: That's a different thing. That's mouth clicks, and that's a

531
00:27:37,440 --> 00:27:39,640
Speaker 3: normal thing that happens when you put a microphone really

532
00:27:39,640 --> 00:27:42,879
Speaker 3: close to the human mouth. It's weird how it doesn't

533
00:27:42,880 --> 00:27:45,040
Speaker 3: happen in person, but it does on micro For that

534
00:27:45,160 --> 00:27:49,360
Speaker 3: kind of noise, it's because it's consistent. We have software

535
00:27:49,400 --> 00:27:51,840
Speaker 3: that we basically tell the software look at this, as

536
00:27:51,880 --> 00:27:54,200
Speaker 3: long as there's no talking, look at this piece of noise,

537
00:27:54,520 --> 00:27:57,000
Speaker 3: identify it as noise, and from now on remove any

538
00:27:57,080 --> 00:28:00,760
Speaker 3: time you encounter that. And they be come more and

539
00:28:00,800 --> 00:28:04,080
Speaker 3: more sophisticated over the years to the point where they're

540
00:28:04,440 --> 00:28:07,679
Speaker 3: quite transparent at it. It's still in many cases if

541
00:28:07,720 --> 00:28:10,320
Speaker 3: you go too hard, things start to sound watery and

542
00:28:10,359 --> 00:28:12,200
Speaker 3: sort of weird, so you have to be careful with that.

543
00:28:12,440 --> 00:28:14,560
Speaker 3: But dig into that a little bit more in detail,

544
00:28:14,640 --> 00:28:17,400
Speaker 3: because what you're saying is it's consistent, so you can

545
00:28:17,720 --> 00:28:20,280
Speaker 3: measure it when the person's not speaking, and then you'll

546
00:28:20,240 --> 00:28:23,440
Speaker 3: have a template to remove it. But still it's varying.

547
00:28:23,640 --> 00:28:27,280
Speaker 3: So is an average removal. It's a statistical removal because

548
00:28:27,320 --> 00:28:29,800
Speaker 3: it can't remove the exact noise moment to moment, because

549
00:28:29,840 --> 00:28:30,640
Speaker 3: there is variation.

550
00:28:30,800 --> 00:28:31,239
Speaker 2: We heard it.

551
00:28:31,920 --> 00:28:37,760
Speaker 3: There is variation, but it's fairly predictable. Like an air conditioner,

552
00:28:37,840 --> 00:28:40,560
Speaker 3: unless the air conditioner is changing temperatures, is going to

553
00:28:40,720 --> 00:28:45,280
Speaker 3: give or take make the same droney noise throughout the recording.

554
00:28:46,040 --> 00:28:48,479
Speaker 3: But yes, sometimes I do catch that throughout the recording.

555
00:28:48,520 --> 00:28:52,360
Speaker 3: Sometimes the air conditioner change gears and then it's no

556
00:28:52,440 --> 00:28:54,800
Speaker 3: longer being recognized and I have to compensate for that.

557
00:28:55,400 --> 00:29:00,200
Speaker 3: There are settings where it's supposed to follow the noise itself.

558
00:29:01,080 --> 00:29:03,120
Speaker 3: I don't find those to be very good yet, but

559
00:29:03,440 --> 00:29:05,440
Speaker 3: tuch changes all the time. Maybe tomorrow there'll be a

560
00:29:05,440 --> 00:29:09,120
Speaker 3: good one. And is this noise removable? Because it's at

561
00:29:09,160 --> 00:29:12,360
Speaker 3: a different frequency than the voice, so you can identify

562
00:29:12,400 --> 00:29:14,240
Speaker 3: it when I'm not speaking, and then you know exactly

563
00:29:14,280 --> 00:29:15,880
Speaker 3: how to remove it. And you can remove it without

564
00:29:15,920 --> 00:29:18,880
Speaker 3: removing the voice or does it overlap of the voice

565
00:29:19,240 --> 00:29:22,520
Speaker 3: and so you are affecting the actual voice as well?

566
00:29:22,720 --> 00:29:25,080
Speaker 3: Oh okay, that's a great question, you really want to

567
00:29:25,080 --> 00:29:27,560
Speaker 3: look under the hood. So the way it really works

568
00:29:27,680 --> 00:29:32,400
Speaker 3: is that the recording is divided into smaller bands of frequencies,

569
00:29:32,440 --> 00:29:35,080
Speaker 3: and each frequency is given what's called a gate or

570
00:29:35,080 --> 00:29:39,360
Speaker 3: an expander, where if a lot of volume hits that area,

571
00:29:39,880 --> 00:29:43,200
Speaker 3: it's allowed to pass through, so speech, for example. But

572
00:29:43,480 --> 00:29:47,840
Speaker 3: if there's a quieter moment, it enhances the quiet moment

573
00:29:47,880 --> 00:29:52,360
Speaker 3: and makes it even quieter. So if you really listen closely,

574
00:29:52,960 --> 00:29:55,680
Speaker 3: you may be able to notice that there are moments

575
00:29:55,680 --> 00:29:59,760
Speaker 3: of noise when you talk. But because you're talking and

576
00:29:59,760 --> 00:30:03,600
Speaker 3: then noise is only present as you're talking, it's not perceptible.

577
00:30:04,720 --> 00:30:06,520
Speaker 3: A lot of this is about magic tricks and not

578
00:30:06,640 --> 00:30:08,200
Speaker 3: actually about doing what you think we're doing.

579
00:30:08,480 --> 00:30:12,520
Speaker 2: By magic tricks, you mean distracting people and understanding how works.

580
00:30:12,600 --> 00:30:15,120
Speaker 3: Yeah, I mean I mean real real magic, not magic

581
00:30:15,200 --> 00:30:17,360
Speaker 3: magic were slide of hand and stuff like that.

582
00:30:19,920 --> 00:30:22,120
Speaker 2: I love what you call that real magic. That's awesome.

583
00:30:22,320 --> 00:30:25,320
Speaker 1: Okay, So then how would you remove something like the

584
00:30:25,480 --> 00:30:26,280
Speaker 1: grocery store.

585
00:30:26,840 --> 00:30:28,959
Speaker 2: I was not at a grocery store. I want that

586
00:30:29,040 --> 00:30:29,560
Speaker 2: on their front.

587
00:30:29,680 --> 00:30:30,320
Speaker 3: I think you were.

588
00:30:30,400 --> 00:30:30,880
Speaker 1: I saw it.

589
00:30:30,960 --> 00:30:35,640
Speaker 3: I saw it. So for a noise like that or

590
00:30:35,760 --> 00:30:42,680
Speaker 3: like goats, Kelly oh yeah, their goats. Yeah.

591
00:30:42,720 --> 00:30:43,840
Speaker 1: Oh, I'm so sorry.

592
00:30:43,840 --> 00:30:46,320
Speaker 3: Oh no, it's totally it's not your fault that they're goats.

593
00:30:46,640 --> 00:30:49,560
Speaker 3: They do goaty things, uh, for things like that. For

594
00:30:49,560 --> 00:30:53,960
Speaker 3: for noises that are not consistent or predictable, which, by

595
00:30:54,000 --> 00:30:56,440
Speaker 3: the way, let me back up one second. Noise canceling

596
00:30:56,440 --> 00:31:00,800
Speaker 3: headphones work on the same principle, So noise can headphones

597
00:31:01,160 --> 00:31:05,920
Speaker 3: can cancel predictable noise once they recognize that noise. However,

598
00:31:06,000 --> 00:31:08,600
Speaker 3: if somebody's talking to you or there's a dog barking,

599
00:31:08,800 --> 00:31:11,120
Speaker 3: you may notice that your noise canceling headphones don't work

600
00:31:11,160 --> 00:31:14,120
Speaker 3: for this very same reason. So if there is a

601
00:31:14,120 --> 00:31:17,080
Speaker 3: dog barking or a goat. Luckily, we live at a

602
00:31:17,160 --> 00:31:18,840
Speaker 3: time where I can just open it up in a

603
00:31:18,880 --> 00:31:21,640
Speaker 3: spectrol graph and I can actually see all the sound.

604
00:31:21,680 --> 00:31:25,080
Speaker 3: It looks like splotches, and I can see the goat

605
00:31:25,320 --> 00:31:27,520
Speaker 3: bleed to an extent, and I can dry it out

606
00:31:27,600 --> 00:31:28,280
Speaker 3: like an photoshop.

607
00:31:28,400 --> 00:31:33,719
Speaker 2: WHOA, So what you're looking at is the amplitude versus frequency.

608
00:31:34,160 --> 00:31:38,320
Speaker 3: Yeah, time is the x frequency is the y axis,

609
00:31:38,400 --> 00:31:43,480
Speaker 3: and then amplitude is color. Oh bright orange means it's

610
00:31:43,600 --> 00:31:46,640
Speaker 3: very loud. So I can actually see all the harmonics

611
00:31:46,680 --> 00:31:49,440
Speaker 3: of the goat or a door being slammed, or a

612
00:31:49,480 --> 00:31:52,120
Speaker 3: plane flying by, And sometimes it's hard to draw them

613
00:31:52,120 --> 00:31:54,160
Speaker 3: out because there's other stuff happening there. But in most

614
00:31:54,200 --> 00:31:57,400
Speaker 3: cases it just takes time. You have to stop, capture

615
00:31:57,400 --> 00:32:00,600
Speaker 3: that piece and follow the goat and remove it.

616
00:32:01,880 --> 00:32:04,480
Speaker 1: How does Daniel's voice show up different than mine? Go

617
00:32:04,680 --> 00:32:08,000
Speaker 1: So Daniel has a very deep voice. What does his

618
00:32:08,120 --> 00:32:10,360
Speaker 1: look like relative to mine?

619
00:32:10,560 --> 00:32:13,000
Speaker 3: Well, too bad. This is not a visual show. So

620
00:32:13,240 --> 00:32:17,360
Speaker 3: what you would see for any human is just a

621
00:32:17,440 --> 00:32:21,880
Speaker 3: series of lines on top of each other most sounds, really,

622
00:32:21,960 --> 00:32:25,120
Speaker 3: and the lowest line is the actual note that Daniel

623
00:32:25,360 --> 00:32:28,600
Speaker 3: or Kelly are saying. And then all the lines the

624
00:32:28,600 --> 00:32:31,480
Speaker 3: notes above that are harmonics that we don't perceive as

625
00:32:31,560 --> 00:32:35,000
Speaker 3: individual notes, but they're the sort of fingerprint that make

626
00:32:35,840 --> 00:32:37,880
Speaker 3: Daniel and Kelly sound different. Even if you sing the

627
00:32:37,880 --> 00:32:40,480
Speaker 3: same note, for example, is just a different combination of

628
00:32:40,480 --> 00:32:42,880
Speaker 3: those harmonics, huh.

629
00:32:42,560 --> 00:32:44,360
Speaker 2: Like, for the same reason that like, a guitar and

630
00:32:44,400 --> 00:32:47,360
Speaker 2: a violin sound different when they're playing the same note,

631
00:32:47,400 --> 00:32:50,760
Speaker 2: because there's a range of frequencies that gives you the

632
00:32:50,960 --> 00:32:53,200
Speaker 2: mental fingerprint to identify it exactly.

633
00:32:53,240 --> 00:32:55,640
Speaker 1: That's all it is, all right, Well, let's take a break,

634
00:32:55,680 --> 00:32:57,720
Speaker 1: and when we get back, we're going to chat with

635
00:32:57,800 --> 00:33:01,240
Speaker 1: Matt about how software is making his life easier or

636
00:33:01,360 --> 00:33:01,840
Speaker 1: maybe not.

637
00:33:02,120 --> 00:33:03,640
Speaker 2: And whether or not he's actually an AI.

638
00:33:03,960 --> 00:33:25,840
Speaker 1: And we're back and we've got Mat in the hot

639
00:33:25,880 --> 00:33:28,280
Speaker 1: seat today we are quizzing him about how he makes

640
00:33:28,360 --> 00:33:31,720
Speaker 1: us sound not horrible, because we sound pretty bad in

641
00:33:31,760 --> 00:33:34,680
Speaker 1: the raw audio that we send him at So AI

642
00:33:34,840 --> 00:33:38,360
Speaker 1: is like everywhere now, so could you like tell AI? Hey?

643
00:33:38,960 --> 00:33:42,520
Speaker 1: Here is an interview from two people for Starters. Cut

644
00:33:42,520 --> 00:33:45,000
Speaker 1: out every time they say and remove all the background

645
00:33:45,080 --> 00:33:48,400
Speaker 1: noise like are we there where that could happen? Or yeah,

646
00:33:48,400 --> 00:33:49,280
Speaker 1: where are we with AI?

647
00:33:49,360 --> 00:33:55,120
Speaker 3: Now? Uh, We're not there. We might be there. I

648
00:33:55,160 --> 00:33:57,360
Speaker 3: have no idea what's going to happen. Maybe next month

649
00:33:57,840 --> 00:34:00,280
Speaker 3: none of us will have jobs. I know this. People

650
00:34:00,280 --> 00:34:02,600
Speaker 3: have very strong beliefs about what's going to happen. I

651
00:34:02,640 --> 00:34:05,720
Speaker 3: think nobody knows, and we may hit a wall or

652
00:34:06,080 --> 00:34:08,839
Speaker 3: it may never stop. See. The thing is, there are

653
00:34:08,960 --> 00:34:11,319
Speaker 3: tools that can do things like that, but then they

654
00:34:11,360 --> 00:34:14,080
Speaker 3: require adult supervision where you check whether it did and

655
00:34:14,080 --> 00:34:16,600
Speaker 3: then you find that it deleted a very important question

656
00:34:16,880 --> 00:34:20,399
Speaker 3: or you know, things like that. So it doesn't help

657
00:34:20,520 --> 00:34:23,040
Speaker 3: me It's still faster for me to do it all

658
00:34:23,120 --> 00:34:25,840
Speaker 3: myself because I work pretty fast, and you know, I

659
00:34:25,880 --> 00:34:28,200
Speaker 3: don't have to go back and try to find what's missing.

660
00:34:29,239 --> 00:34:32,600
Speaker 3: The tools I use the most with AI are for

661
00:34:33,400 --> 00:34:36,800
Speaker 3: noise cleanup. That's also another option if I don't draw

662
00:34:37,160 --> 00:34:40,560
Speaker 3: it out with a pencil tool. Is There are AIS

663
00:34:40,600 --> 00:34:44,080
Speaker 3: that are pretty good at being able to tell what

664
00:34:44,200 --> 00:34:47,279
Speaker 3: noise is, and they sort of split into two categories.

665
00:34:47,320 --> 00:34:50,759
Speaker 3: One is it removes the noise on its own using

666
00:34:50,760 --> 00:34:54,640
Speaker 3: all kinds of cool phasing algorithms, and then there's one

667
00:34:54,719 --> 00:34:57,840
Speaker 3: that recognizes your voice and then resynthesizes your voice completely

668
00:34:57,840 --> 00:35:04,720
Speaker 3: from scratch. Wow wow, which is a little weird, scary scary. Yeah.

669
00:35:04,760 --> 00:35:06,680
Speaker 2: Well, I had a listener wright to me once that

670
00:35:06,719 --> 00:35:10,200
Speaker 2: they heard an AI generated podcast and that one of

671
00:35:10,239 --> 00:35:13,120
Speaker 2: the hosts on that podcast sounded a lot like me,

672
00:35:13,560 --> 00:35:15,320
Speaker 2: and I was like what, So then I listened to

673
00:35:15,400 --> 00:35:16,920
Speaker 2: it and I was like, that does kind of sound

674
00:35:16,960 --> 00:35:19,600
Speaker 2: like me, And then I realized there are hundreds of

675
00:35:19,719 --> 00:35:22,839
Speaker 2: hours of my voice out there for anybody to train

676
00:35:22,920 --> 00:35:25,359
Speaker 2: an AI to make an AI Daniel.

677
00:35:25,440 --> 00:35:29,439
Speaker 3: And you only need five seconds now. And by the way, PSA,

678
00:35:29,520 --> 00:35:32,160
Speaker 3: I don't know about your bank but my bank has

679
00:35:32,200 --> 00:35:35,440
Speaker 3: this voice print identification if you call them, Yeah, I

680
00:35:35,480 --> 00:35:38,840
Speaker 3: asked them to turn that off. Yeah, because it's incredibly easy,

681
00:35:38,920 --> 00:35:42,080
Speaker 3: especially for anyone that has their voice out there, for

682
00:35:42,120 --> 00:35:44,720
Speaker 3: anybody to fake your voice. So that is no longer

683
00:35:44,760 --> 00:35:46,760
Speaker 3: safe if you have that enabled at your bank.

684
00:35:46,840 --> 00:35:48,840
Speaker 2: And I'm not endorsing it, but I did find this

685
00:35:48,920 --> 00:35:51,080
Speaker 2: AI generated podcast to be pretty good. And I did

686
00:35:51,080 --> 00:35:54,360
Speaker 2: a little experiment where I took one of Katrina's latest

687
00:35:54,400 --> 00:35:56,560
Speaker 2: papers and I ran it through and I had produced

688
00:35:56,560 --> 00:35:59,120
Speaker 2: a fifteen minute podcast about it. Then I sent it

689
00:35:59,160 --> 00:36:01,600
Speaker 2: to her and didn't tell her it was AI generated.

690
00:36:01,680 --> 00:36:04,320
Speaker 2: I said, hey, I heard this podcast about your latest paper.

691
00:36:04,640 --> 00:36:06,920
Speaker 2: And she listened to it. She was really excited and

692
00:36:06,960 --> 00:36:08,080
Speaker 2: she shared it with her laugh.

693
00:36:10,400 --> 00:36:11,480
Speaker 3: Huh, that is weird.

694
00:36:11,640 --> 00:36:14,000
Speaker 2: It was surprisingly good. It's pretty compelling stuff.

695
00:36:14,080 --> 00:36:16,640
Speaker 1: Unfortunately, Well, so, I think you should be flattered that

696
00:36:16,800 --> 00:36:18,600
Speaker 1: out of all the voices they could have picked, they

697
00:36:18,640 --> 00:36:21,560
Speaker 1: decided to scrape your voice. Daniels a good voice.

698
00:36:21,800 --> 00:36:24,920
Speaker 2: I don't know that they did, you know, just generic

699
00:36:25,040 --> 00:36:29,040
Speaker 2: podcast dude with a deep voice, I guess. But no

700
00:36:29,120 --> 00:36:32,080
Speaker 2: AI is out there making goat jokes, Kelly, So you

701
00:36:32,080 --> 00:36:32,640
Speaker 2: are unique?

702
00:36:32,840 --> 00:36:35,680
Speaker 1: Yes, Oh that's great. I'm so glad, So Matt. Even

703
00:36:35,719 --> 00:36:40,520
Speaker 1: with the help of software, we send you about an hour,

704
00:36:40,719 --> 00:36:43,600
Speaker 1: an hour or twenty minutes of audio, it gets compressed

705
00:36:43,640 --> 00:36:47,840
Speaker 1: to about an hour of podcast. How many hours go

706
00:36:47,960 --> 00:36:50,320
Speaker 1: into editing? Uh?

707
00:36:50,560 --> 00:36:55,920
Speaker 3: Anywhere from five to ten wow, depending depending on physics

708
00:36:56,000 --> 00:36:59,759
Speaker 3: or bio subject matter. Yes, yes, Well, look if it's

709
00:37:00,239 --> 00:37:05,440
Speaker 3: if it's about particle accelerators, Daniel already has all that

710
00:37:05,520 --> 00:37:09,160
Speaker 3: stuff memorized and it's easy to do. Or if we're

711
00:37:09,160 --> 00:37:13,440
Speaker 3: talking about killie fish or something like that. But if

712
00:37:13,480 --> 00:37:15,920
Speaker 3: somebody just throws you a curveball question that you had

713
00:37:15,920 --> 00:37:17,839
Speaker 3: to look up, then it could take a bit longer.

714
00:37:17,840 --> 00:37:21,000
Speaker 3: If there's a guest, it usually takes longer because you

715
00:37:21,040 --> 00:37:23,239
Speaker 3: have a rapport and a banter that you're used to,

716
00:37:23,320 --> 00:37:25,239
Speaker 3: and then if somebody else comes in, they may not

717
00:37:25,640 --> 00:37:27,680
Speaker 3: have that same vibe and it takes a bit of time.

718
00:37:28,440 --> 00:37:30,239
Speaker 3: And I want to point out that five to ten

719
00:37:30,280 --> 00:37:33,000
Speaker 3: hours may sound like a lot, but it's not a

720
00:37:33,040 --> 00:37:35,640
Speaker 3: lot or unusual at all if you think about it.

721
00:37:35,640 --> 00:37:38,040
Speaker 3: It's usually roughly about an hour and a half of

722
00:37:38,280 --> 00:37:40,920
Speaker 3: raw audio that I get. I have to listen to

723
00:37:40,920 --> 00:37:43,040
Speaker 3: it at least one time, so that's already an hour

724
00:37:43,080 --> 00:37:46,160
Speaker 3: and a half. Then I need to do some noise reduction,

725
00:37:46,560 --> 00:37:50,719
Speaker 3: which can take thirty to fifty minutes, depending on the

726
00:37:50,760 --> 00:37:53,399
Speaker 3: type of noise there is. And then from there, every

727
00:37:53,560 --> 00:37:56,440
Speaker 3: edit that I make, I have to first hear it, stop,

728
00:37:56,640 --> 00:37:58,319
Speaker 3: listen to it one more time to make sure there

729
00:37:58,360 --> 00:38:01,120
Speaker 3: is a problem, do the edit, which takes between a

730
00:38:01,120 --> 00:38:03,560
Speaker 3: few seconds and a few minutes depending on the edit,

731
00:38:04,000 --> 00:38:07,160
Speaker 3: and then listen back. And these little things all add up,

732
00:38:07,200 --> 00:38:10,080
Speaker 3: and before you know it, you're easily crossing the five

733
00:38:10,120 --> 00:38:12,880
Speaker 3: hour mark. By the way, one other thing that I

734
00:38:12,920 --> 00:38:17,160
Speaker 3: forgot to mention is there's the editing sort of to

735
00:38:17,200 --> 00:38:19,439
Speaker 3: help everything feel smooth and makes sense. But then there's

736
00:38:19,480 --> 00:38:25,239
Speaker 3: also the issue of comedic timing over web conferencing. There's

737
00:38:25,239 --> 00:38:28,000
Speaker 3: a delay, whether you like it or not, and it's

738
00:38:28,040 --> 00:38:31,080
Speaker 3: not something as easy as shifting one voice back. Because

739
00:38:31,560 --> 00:38:35,479
Speaker 3: Kelly would say something funny, Daniel would accidentally talk over

740
00:38:35,560 --> 00:38:39,040
Speaker 3: Kelly because he's been waiting because there's a delay, then

741
00:38:39,120 --> 00:38:43,200
Speaker 3: realize that was funny, laugh at that. Kelly also hears

742
00:38:43,239 --> 00:38:47,120
Speaker 3: a delay, so she's already trying to damage control Daniel

743
00:38:47,200 --> 00:38:52,799
Speaker 3: not laughing, and it becomes this mess where I'm like,

744
00:38:52,840 --> 00:38:55,279
Speaker 3: there's a very funny joke here, and both laughed at

745
00:38:55,280 --> 00:38:58,799
Speaker 3: the joke, but because of that delay, it sounds a

746
00:38:58,800 --> 00:39:01,719
Speaker 3: little bit train wreckage. So that's another thing that.

747
00:39:01,680 --> 00:39:03,880
Speaker 2: I know, Wow, what a review?

748
00:39:04,120 --> 00:39:06,399
Speaker 3: Well, because I know that in person you both would

749
00:39:06,560 --> 00:39:08,440
Speaker 3: have this banter and laugh and it would be so

750
00:39:08,520 --> 00:39:11,640
Speaker 3: I'm sort of time traveling in a way to put

751
00:39:11,760 --> 00:39:14,920
Speaker 3: you where you wanted to be, at least where I

752
00:39:14,960 --> 00:39:17,440
Speaker 3: perceive you want it to be, so that that's also

753
00:39:17,480 --> 00:39:17,960
Speaker 3: part of it.

754
00:39:19,480 --> 00:39:21,520
Speaker 1: Thank you for making us funny.

755
00:39:21,920 --> 00:39:24,200
Speaker 3: No, you're the funny ones. I just undo the damage

756
00:39:24,239 --> 00:39:25,200
Speaker 3: the internet causes.

757
00:39:26,440 --> 00:39:29,560
Speaker 1: So why does my voice sound so much better in

758
00:39:29,640 --> 00:39:32,319
Speaker 1: my head? So like when I listened to the file

759
00:39:32,400 --> 00:39:35,040
Speaker 1: you send me, I think, wow, this sounds way better

760
00:39:35,360 --> 00:39:38,520
Speaker 1: than when I actually was like talking to Daniel in

761
00:39:38,560 --> 00:39:41,000
Speaker 1: this version, Daniel left at my joke right away, but

762
00:39:41,160 --> 00:39:44,759
Speaker 1: I still I still cringe because I don't love the

763
00:39:44,800 --> 00:39:47,360
Speaker 1: sound of my voice when I hear it back to me.

764
00:39:47,480 --> 00:39:49,799
Speaker 1: But when I'm saying it, it doesn't sound so bad.

765
00:39:49,840 --> 00:39:51,719
Speaker 1: So why is there that disconnect?

766
00:39:51,880 --> 00:39:53,400
Speaker 2: I think you have a great voice, Kelly. I love

767
00:39:53,440 --> 00:39:53,839
Speaker 2: hearing it.

768
00:39:53,920 --> 00:39:55,680
Speaker 3: Thanks Daniel the great great voice.

769
00:39:55,760 --> 00:39:56,279
Speaker 1: Thanks guys.

770
00:39:56,360 --> 00:39:59,000
Speaker 3: Otherwise people wouldn't be listening as interesting as you are.

771
00:39:59,040 --> 00:40:00,920
Speaker 3: If you had bad voice, as people would move on.

772
00:40:02,000 --> 00:40:06,359
Speaker 3: It's a matter of perspective. So we hear everything else

773
00:40:06,400 --> 00:40:10,800
Speaker 3: and everyone else through our ear canals. Right, there's air outside,

774
00:40:10,880 --> 00:40:12,959
Speaker 3: and that air vibrates and that goes into our ear canals.

775
00:40:13,000 --> 00:40:16,359
Speaker 3: Will hear it for ourselves. A big chunk of our

776
00:40:16,480 --> 00:40:20,440
Speaker 3: voice travels through conduction into our ear from the inside,

777
00:40:20,480 --> 00:40:21,960
Speaker 3: and that's what people call the inner ear. There's no

778
00:40:21,960 --> 00:40:23,680
Speaker 3: actual inner ear. I used to think there was when

779
00:40:23,719 --> 00:40:25,560
Speaker 3: I was a kid, like a third ear in your throat.

780
00:40:25,880 --> 00:40:30,200
Speaker 3: There isn't one. You're basically hearing sort of a muffled

781
00:40:30,239 --> 00:40:33,720
Speaker 3: version of yourself because it's traveling through soft tissue and bone,

782
00:40:34,239 --> 00:40:36,640
Speaker 3: and then you're also hearing a little bit of yourself

783
00:40:36,680 --> 00:40:38,759
Speaker 3: through the air going back into your ear, so it's

784
00:40:38,800 --> 00:40:40,920
Speaker 3: a different sort of like I don't know. When I

785
00:40:40,960 --> 00:40:43,960
Speaker 3: look in the mirror, I'm gorgeous. When I look at

786
00:40:44,000 --> 00:40:47,680
Speaker 3: a photo, I look like a crescent moon. I noticed

787
00:40:47,719 --> 00:40:51,600
Speaker 3: that my face is curved, and I never noticed that

788
00:40:51,640 --> 00:40:54,600
Speaker 3: in the mirror because it's just a perspective I'm not

789
00:40:54,719 --> 00:40:56,520
Speaker 3: used to. And I think that's exactly what happens with

790
00:40:56,600 --> 00:40:58,960
Speaker 3: your voices. You're used to hearing it a certain way,

791
00:40:59,280 --> 00:41:01,960
Speaker 3: and no matter how the recording sounds, you'll think it

792
00:41:02,000 --> 00:41:04,000
Speaker 3: sounds worse because you're just not used to it.

793
00:41:04,400 --> 00:41:07,560
Speaker 2: So somebody was to put their cheek up against Kelly

794
00:41:07,600 --> 00:41:10,080
Speaker 2: while she was speaking, so that her voice traveled like

795
00:41:10,160 --> 00:41:13,560
Speaker 2: through their soft tissue, would they hear Kelly's internal version

796
00:41:13,560 --> 00:41:14,200
Speaker 2: of her voice?

797
00:41:14,600 --> 00:41:20,439
Speaker 3: If you were somehow able to model Kelly's innards from

798
00:41:20,520 --> 00:41:23,080
Speaker 3: her throat all the way up to the top of

799
00:41:23,120 --> 00:41:27,399
Speaker 3: her skull and put that between the other person, yes,

800
00:41:27,719 --> 00:41:28,800
Speaker 3: it would be very strange.

801
00:41:28,800 --> 00:41:31,560
Speaker 1: But yes, I'd just like to put on record that

802
00:41:31,600 --> 00:41:33,840
Speaker 1: when I see you, I never think Crescent.

803
00:41:33,520 --> 00:41:36,560
Speaker 3: Moved, but I will now I have to tell myself

804
00:41:36,600 --> 00:41:39,480
Speaker 3: the same thing is that that tiny lack of symmetry

805
00:41:39,480 --> 00:41:42,879
Speaker 3: which every human has is just accentuated because I never

806
00:41:42,920 --> 00:41:46,279
Speaker 3: see it from that angle. And I feel like most

807
00:41:46,280 --> 00:41:48,719
Speaker 3: people when they look at a photo compared to the mirror,

808
00:41:48,719 --> 00:41:50,600
Speaker 3: they're like, oh, what is that? What happened?

809
00:41:50,719 --> 00:41:51,600
Speaker 1: Yeah?

810
00:41:51,680 --> 00:41:54,160
Speaker 2: So you mentioned earlier about real magic and how big

811
00:41:54,200 --> 00:41:56,480
Speaker 2: part of this is knowing how it's going to be

812
00:41:56,560 --> 00:41:59,960
Speaker 2: received by people. Tell us more about that, about how

813
00:42:00,320 --> 00:42:02,719
Speaker 2: understanding the way sound is actually perceived and the way

814
00:42:02,719 --> 00:42:06,040
Speaker 2: people pay attention changes how you do your editing and

815
00:42:06,080 --> 00:42:08,520
Speaker 2: tell us about psychoacoustics.

816
00:42:09,120 --> 00:42:12,239
Speaker 3: So, Daniel, you're gonna be proud of me. Before I

817
00:42:12,400 --> 00:42:15,600
Speaker 3: even knew who you were. I've talked to my students

818
00:42:15,600 --> 00:42:16,600
Speaker 3: about aliens a.

819
00:42:16,560 --> 00:42:19,720
Speaker 2: Lot plus ten points.

820
00:42:21,280 --> 00:42:23,520
Speaker 3: Yeah. The reason I do that is I always in

821
00:42:23,760 --> 00:42:26,440
Speaker 3: the first class. I tell them this is a class

822
00:42:26,920 --> 00:42:33,200
Speaker 3: for human beings in Earth's atmosphere, because, first of all,

823
00:42:33,280 --> 00:42:36,920
Speaker 3: sound behaves differently through different materials, and second, humans have

824
00:42:37,040 --> 00:42:40,400
Speaker 3: evolved to process their senses in ways that are useful

825
00:42:40,400 --> 00:42:43,879
Speaker 3: to them, which is another way of saying that what

826
00:42:44,080 --> 00:42:47,640
Speaker 3: we hear and see and feel is not reality, but

827
00:42:48,040 --> 00:42:52,640
Speaker 3: our useful version of reality, and sound is no exception

828
00:42:52,760 --> 00:42:56,040
Speaker 3: to that. I think we may not realize how much

829
00:42:56,080 --> 00:42:58,719
Speaker 3: of a role it plays. My theory is because there

830
00:42:58,760 --> 00:43:06,640
Speaker 3: are no earlids go on. With sight, you have these

831
00:43:06,680 --> 00:43:09,480
Speaker 3: eyelids where you can close them and not see and

832
00:43:09,560 --> 00:43:11,480
Speaker 3: get an idea of what it's like to not have

833
00:43:11,560 --> 00:43:13,960
Speaker 3: that sense. You also have to directly look at the

834
00:43:14,000 --> 00:43:17,120
Speaker 3: thing that you're looking at to perceive it, and with sound,

835
00:43:17,320 --> 00:43:20,319
Speaker 3: you hear everything around you. You don't have to look

836
00:43:20,360 --> 00:43:23,440
Speaker 3: at it with your ear, and you've never closed your ears.

837
00:43:24,239 --> 00:43:28,080
Speaker 3: Your earballs are always exposed. So it's a sense that's

838
00:43:28,640 --> 00:43:30,960
Speaker 3: so enveloping and so much part of our lives that

839
00:43:31,000 --> 00:43:33,320
Speaker 3: we sort of take it for granted and not understand

840
00:43:33,360 --> 00:43:38,120
Speaker 3: how much information we're constantly getting from our surroundings through sound.

841
00:43:38,080 --> 00:43:40,279
Speaker 2: And we don't even have the language for it, right,

842
00:43:40,360 --> 00:43:43,640
Speaker 2: you have to adapt the language from eyes. Like you said,

843
00:43:43,800 --> 00:43:46,120
Speaker 2: look at it with your ear instead of hear at

844
00:43:46,160 --> 00:43:48,680
Speaker 2: it with your ear, because that makes no sense, right.

845
00:43:49,280 --> 00:43:52,400
Speaker 3: Yeah, So we think there's just this baseline of nothing,

846
00:43:52,480 --> 00:43:55,640
Speaker 3: but in reality, you know the size of the room

847
00:43:55,640 --> 00:43:57,319
Speaker 3: you're in and what the walls are made out of

848
00:43:57,600 --> 00:44:01,120
Speaker 3: before you look and tap them. Walls, you know from

849
00:44:01,160 --> 00:44:02,920
Speaker 3: how sound reflects in that room.

850
00:44:03,000 --> 00:44:04,759
Speaker 2: I have a question about that because I've been told

851
00:44:04,760 --> 00:44:08,000
Speaker 2: many times to record in a room that doesn't have

852
00:44:08,120 --> 00:44:11,920
Speaker 2: auditory reflections, right, like surround yourself with laundry or something,

853
00:44:12,320 --> 00:44:14,520
Speaker 2: because nobody wants to hear the room that you're in.

854
00:44:15,360 --> 00:44:18,279
Speaker 2: So isn't that weird for people to hear audio that

855
00:44:18,320 --> 00:44:21,520
Speaker 2: sounds sort of like disembodied or de rumified.

856
00:44:22,719 --> 00:44:26,280
Speaker 3: No, because it's far from disembodied and de rumified unless

857
00:44:26,320 --> 00:44:28,279
Speaker 3: you've been to an anichoid chamber. You have never heard

858
00:44:28,320 --> 00:44:31,200
Speaker 3: silence in your life. It's a have you been to one?

859
00:44:31,280 --> 00:44:31,360
Speaker 1: No?

860
00:44:31,520 --> 00:44:34,240
Speaker 3: I highly recommend it. It's weird.

861
00:44:34,480 --> 00:44:36,400
Speaker 2: Sounds like a torture device, you know what.

862
00:44:36,520 --> 00:44:39,080
Speaker 3: People say that there's a rumor that no one has

863
00:44:39,120 --> 00:44:41,600
Speaker 3: ever lasted more than forty five minutes. Not true. I've

864
00:44:41,600 --> 00:44:44,640
Speaker 3: been there for hours. But you can hear blood rushing

865
00:44:44,640 --> 00:44:48,880
Speaker 3: through your ears. You can hear everything. And not only that,

866
00:44:49,040 --> 00:44:51,200
Speaker 3: if you're there with a buddy and they turn around

867
00:44:51,400 --> 00:44:53,320
Speaker 3: and talk and speak the other way, you can barely

868
00:44:53,360 --> 00:44:56,799
Speaker 3: hear them. We don't realize, wow, or we take for

869
00:44:56,840 --> 00:44:59,359
Speaker 3: granted how much of what we hear is actually reflections

870
00:44:59,719 --> 00:45:02,799
Speaker 3: of the source and not the source itself. So I

871
00:45:02,800 --> 00:45:07,480
Speaker 3: mean an example of some important psychoacoustics that engineers have

872
00:45:07,520 --> 00:45:10,000
Speaker 3: to keep in mind. And also the answer to the

873
00:45:10,040 --> 00:45:14,760
Speaker 3: listener question, what do you think what is the type

874
00:45:14,760 --> 00:45:16,319
Speaker 3: of sound or sound that humans hear?

875
00:45:16,360 --> 00:45:19,120
Speaker 2: Best? I would have guessed that there's something similar to

876
00:45:19,280 --> 00:45:22,040
Speaker 2: our visual arrange, where there's a whole spectrum of frequencies

877
00:45:22,440 --> 00:45:24,839
Speaker 2: and we can see something in the middle. Where in

878
00:45:24,880 --> 00:45:28,200
Speaker 2: the middle basically means where we can hear, which makes

879
00:45:28,239 --> 00:45:29,080
Speaker 2: it kind of a non.

880
00:45:28,920 --> 00:45:33,840
Speaker 3: Answer, right, Well, that's actually that's a good comparison with sound.

881
00:45:34,000 --> 00:45:36,200
Speaker 3: You can see red just as well as you can

882
00:45:36,239 --> 00:45:40,479
Speaker 3: see blue with sight, but with sound it doesn't work

883
00:45:40,520 --> 00:45:44,120
Speaker 3: that way. With sound, we can hear frequencies from twenty

884
00:45:44,200 --> 00:45:47,440
Speaker 3: vibrations per second twenty hurts to twenty thousand vibrations per

885
00:45:47,480 --> 00:45:49,919
Speaker 3: second twenty killer hurts, but out of those we don't

886
00:45:49,960 --> 00:45:52,680
Speaker 3: hear them all equally. The frequency range we hear best

887
00:45:53,320 --> 00:45:55,680
Speaker 3: is between one and five killer hurts, and that is

888
00:45:55,719 --> 00:46:01,040
Speaker 3: the range of leaves rustling and breaking.

889
00:46:00,880 --> 00:46:02,960
Speaker 2: And jaguar is creeping up on you so that.

890
00:46:02,960 --> 00:46:05,360
Speaker 3: Predators don't creep up on you. It's also the range

891
00:46:05,360 --> 00:46:08,680
Speaker 3: of a baby crying. You can hear a baby cry

892
00:46:08,800 --> 00:46:12,440
Speaker 3: from way farther away than I don't know a car

893
00:46:12,640 --> 00:46:14,960
Speaker 3: engine reving at the same distance, same volume.

894
00:46:15,360 --> 00:46:17,600
Speaker 2: It's also the sound of a candy bar wrapper, isn't

895
00:46:17,600 --> 00:46:18,440
Speaker 2: it That is?

896
00:46:19,120 --> 00:46:23,480
Speaker 3: Yes, yes, it is, and we've been using that. For example.

897
00:46:23,520 --> 00:46:26,680
Speaker 3: Police sirens are tuned to that frequency range, and that's

898
00:46:26,680 --> 00:46:28,759
Speaker 3: why you may have found yourself sitting in your car

899
00:46:28,840 --> 00:46:30,920
Speaker 3: or walking hearing a siren looking around being like, I

900
00:46:30,920 --> 00:46:32,879
Speaker 3: don't see the police anywhere. Where are they? It's because

901
00:46:32,920 --> 00:46:36,759
Speaker 3: we're so good at hearing those frequencies that we can

902
00:46:36,800 --> 00:46:39,120
Speaker 3: hear them from so far away that we can't even

903
00:46:39,160 --> 00:46:40,640
Speaker 3: see the source of the sound yet.

904
00:46:40,680 --> 00:46:40,960
Speaker 1: Wow.

905
00:46:41,000 --> 00:46:43,160
Speaker 2: So then how does all of this shape your editing?

906
00:46:43,840 --> 00:46:45,920
Speaker 2: Are you trying to push my voice up to police

907
00:46:45,960 --> 00:46:46,880
Speaker 2: siren frequencies?

908
00:46:48,080 --> 00:46:52,680
Speaker 3: Well, I definitely have to make sure that intelligibility is

909
00:46:52,719 --> 00:46:55,960
Speaker 3: there and that is in those frequencies. But there's a

910
00:46:56,040 --> 00:46:59,600
Speaker 3: challenge here because we hear those frequencies so well, they

911
00:46:59,640 --> 00:47:04,359
Speaker 3: become annoying very quickly too, because we're acutely tuned to them.

912
00:47:04,880 --> 00:47:07,839
Speaker 3: So there's a balancing act where if there's not enough

913
00:47:08,440 --> 00:47:10,520
Speaker 3: it sounds I can show you. Do you want me

914
00:47:10,560 --> 00:47:11,239
Speaker 3: to show you? Yeah?

915
00:47:11,320 --> 00:47:11,960
Speaker 1: Yeah, please?

916
00:47:12,160 --> 00:47:15,160
Speaker 3: So here let me play process Daniel here.

917
00:47:15,520 --> 00:47:19,759
Speaker 2: So Bell's paradox involves length contraction, the fact that when

918
00:47:20,040 --> 00:47:23,799
Speaker 2: things move fast, they look short. So relativity tells us

919
00:47:23,800 --> 00:47:28,600
Speaker 2: two things, moving clocks run slow and moving objects look short.

920
00:47:29,320 --> 00:47:31,400
Speaker 3: So now I'm going to take out that frequency range

921
00:47:31,800 --> 00:47:33,920
Speaker 3: that I was just talking about, and you'll notice so

922
00:47:33,960 --> 00:47:37,560
Speaker 3: you can still hear Daniel. He's loud, but the part

923
00:47:37,600 --> 00:47:40,279
Speaker 3: that makes him human loud, I mean loud is the

924
00:47:40,400 --> 00:47:42,760
Speaker 3: volume I made him. Not because Daniel is a loud person,

925
00:47:43,560 --> 00:47:46,920
Speaker 3: but the intelligibility is just not there. Take a listen.

926
00:47:47,600 --> 00:47:51,839
Speaker 2: So Bell's paradox involves length contraction, the fact that when

927
00:47:52,120 --> 00:47:55,840
Speaker 2: things move fast, they look short. So relativity tells us

928
00:47:55,880 --> 00:48:00,680
Speaker 2: two things, moving clocks run slow and moving objects look short.

929
00:48:01,239 --> 00:48:02,319
Speaker 1: Huh.

930
00:48:02,360 --> 00:48:04,799
Speaker 3: And now I'll exaggerate those frequencies and you'll see that

931
00:48:04,800 --> 00:48:06,040
Speaker 3: it'll kind of hurt your ears.

932
00:48:06,480 --> 00:48:10,719
Speaker 2: So Bell's paradox involves length contraction, the fact that when

933
00:48:11,000 --> 00:48:13,120
Speaker 2: things move fast they look short.

934
00:48:13,400 --> 00:48:15,760
Speaker 3: Ooh right, So I have to find a happy medium,

935
00:48:15,800 --> 00:48:18,200
Speaker 3: and that is sometimes something that has to be done temporally.

936
00:48:18,280 --> 00:48:21,040
Speaker 3: So I have to make sure that some words don't

937
00:48:21,080 --> 00:48:24,680
Speaker 3: have too much of those frequencies and others do. That's

938
00:48:24,719 --> 00:48:27,440
Speaker 3: just one of many ways in which I have to

939
00:48:27,600 --> 00:48:31,359
Speaker 3: look at the waveforms and understand the science and then

940
00:48:31,480 --> 00:48:36,120
Speaker 3: also understand the psychoacoustics, the psychology of how humans perceive

941
00:48:36,200 --> 00:48:36,640
Speaker 3: this stuff.

942
00:48:36,680 --> 00:48:39,080
Speaker 2: Well, you're doing a great job, because I sometimes meet

943
00:48:39,120 --> 00:48:41,160
Speaker 2: listeners in real life and they'll say to me, oh

944
00:48:41,160 --> 00:48:43,239
Speaker 2: my gosh, it's so weird. You sound just like you

945
00:48:43,320 --> 00:48:46,880
Speaker 2: do on the podcast. So you are reproducing this faithfully.

946
00:48:47,239 --> 00:48:48,399
Speaker 3: I did it, woo.

947
00:48:48,640 --> 00:48:51,360
Speaker 1: So how much physics did you have to learn to

948
00:48:51,480 --> 00:48:52,760
Speaker 1: be this good? At your job.

949
00:48:53,080 --> 00:48:56,360
Speaker 3: I see what you're saying. Yes, audio engineers are real engineers.

950
00:48:57,239 --> 00:49:02,000
Speaker 3: We're like the dentists of the audio profession.

951
00:49:02,120 --> 00:49:03,680
Speaker 2: She's just trying to sus that if you're on the

952
00:49:03,680 --> 00:49:06,640
Speaker 2: physics camp or the biology camp, she's wondering if you're biased.

953
00:49:07,000 --> 00:49:09,239
Speaker 3: See, I'm stuck right in the middle between physics and

954
00:49:09,280 --> 00:49:10,920
Speaker 3: biology exactly.

955
00:49:11,160 --> 00:49:12,040
Speaker 1: It's a good place to be.

956
00:49:12,880 --> 00:49:15,120
Speaker 3: Yeah, it's it's a very fun place to be because

957
00:49:15,120 --> 00:49:17,160
Speaker 3: who cares about physics if it's not for how it

958
00:49:17,200 --> 00:49:19,040
Speaker 3: affects sentient beings?

959
00:49:20,200 --> 00:49:21,040
Speaker 1: Amen?

960
00:49:21,320 --> 00:49:24,440
Speaker 2: Amen, hmmm, I have thoughts of that.

961
00:49:24,719 --> 00:49:26,440
Speaker 1: I'm so glad we got you. You know, Matt, you

962
00:49:26,440 --> 00:49:28,000
Speaker 1: should be on the show every week.

963
00:49:28,200 --> 00:49:30,719
Speaker 2: I see, Kelly just wants to invite people on the

964
00:49:30,719 --> 00:49:33,560
Speaker 2: show who will disagree with me gang up on me.

965
00:49:35,239 --> 00:49:39,520
Speaker 3: So let me tell you another psycho acoustic phenomenon called masking.

966
00:49:40,320 --> 00:49:44,120
Speaker 3: The way this works and this probably like we can't

967
00:49:44,160 --> 00:49:47,680
Speaker 3: ask evolution, right, but it looks like we've developed this

968
00:49:47,800 --> 00:49:53,239
Speaker 3: property to u curb over stimulation. What happens is if

969
00:49:53,239 --> 00:49:55,279
Speaker 3: there are certain and it's very predictable, if there are

970
00:49:55,280 --> 00:49:59,600
Speaker 3: certain frequencies present, we are deaf to certain other frequencies

971
00:50:00,480 --> 00:50:03,279
Speaker 3: in those moments. It's just like a blind spot in

972
00:50:03,320 --> 00:50:05,480
Speaker 3: the car, Like, no matter what mirror you look at,

973
00:50:05,520 --> 00:50:07,200
Speaker 3: you will not see the car right next to you

974
00:50:07,239 --> 00:50:08,839
Speaker 3: if they're in your blind spot. So we have these

975
00:50:08,880 --> 00:50:12,800
Speaker 3: blind spots that are constantly, very quickly changing as different

976
00:50:12,840 --> 00:50:18,520
Speaker 3: sounds are present, and very smart programmers mapped out the

977
00:50:18,640 --> 00:50:23,120
Speaker 3: exact way we have these blind spots and figured out,

978
00:50:23,160 --> 00:50:25,080
Speaker 3: you know what, we can save a lot of space

979
00:50:25,080 --> 00:50:27,799
Speaker 3: and audio files if we just throw that stuff away.

980
00:50:27,960 --> 00:50:31,759
Speaker 3: Oh oh wow. And that is how the MP three

981
00:50:31,840 --> 00:50:34,759
Speaker 3: file format was born. It does throw away a lot

982
00:50:34,800 --> 00:50:37,920
Speaker 3: of things, but it throws away only things you would

983
00:50:37,920 --> 00:50:40,560
Speaker 3: not perceive anyway. And so if you've ever had this

984
00:50:41,320 --> 00:50:43,919
Speaker 3: debate with some or just heard someone claim that MP

985
00:50:43,960 --> 00:50:47,600
Speaker 3: three's just don't sound like waves, man, there's just something missing.

986
00:50:48,120 --> 00:50:50,560
Speaker 3: Not necessarily, if it was a well encoded MP three,

987
00:50:51,360 --> 00:50:56,000
Speaker 3: you will never hear what is missing. Wow, Because you're human.

988
00:50:56,680 --> 00:51:01,360
Speaker 3: Our imperfect hearing has really helped us on file sizes. However,

989
00:51:01,480 --> 00:51:04,120
Speaker 3: if an alien came and you went, here's a wavefile,

990
00:51:04,200 --> 00:51:06,200
Speaker 3: like here's a recording, and then here's my MP three

991
00:51:06,239 --> 00:51:09,600
Speaker 3: of that recording to that alien, it might sound completely

992
00:51:09,640 --> 00:51:13,720
Speaker 3: different and upsetting because they don't have our same perception.

993
00:51:14,080 --> 00:51:16,600
Speaker 2: Those aliens who listen to the podcast might be really

994
00:51:16,640 --> 00:51:20,960
Speaker 2: sensitive to those frequencies that our audience, our human audience,

995
00:51:21,040 --> 00:51:21,520
Speaker 2: can't hear.

996
00:51:22,040 --> 00:51:23,759
Speaker 3: As far as we know, when we listen to an

997
00:51:23,840 --> 00:51:28,160
Speaker 3: MP three, our pets are suffering. They don't seem to

998
00:51:28,200 --> 00:51:31,239
Speaker 3: be complaining, but they're definitely not hearing what we hear

999
00:51:31,280 --> 00:51:33,200
Speaker 3: when we listen to the original recording the way that

1000
00:51:33,280 --> 00:51:33,719
Speaker 3: humans do.

1001
00:51:33,840 --> 00:51:36,680
Speaker 2: So this is like the moral equivalent of masking out

1002
00:51:36,680 --> 00:51:40,239
Speaker 2: the ultraviolet off of a painting, and a bee might

1003
00:51:40,239 --> 00:51:42,719
Speaker 2: see a painting very very differently, whereas a human would

1004
00:51:42,719 --> 00:51:44,000
Speaker 2: be totally unaware.

1005
00:51:44,760 --> 00:51:47,720
Speaker 3: Kind of It's like if humans, when you see yellow,

1006
00:51:47,960 --> 00:51:49,800
Speaker 3: you're temporarily blind to purple.

1007
00:51:51,320 --> 00:51:55,360
Speaker 1: Weird, Why why why does that happen?

1008
00:51:56,280 --> 00:51:58,959
Speaker 3: I don't know. I haven't asked God yet, but I

1009
00:51:59,160 --> 00:52:02,480
Speaker 3: think probably because of the overstimulation thing. I think if

1010
00:52:02,520 --> 00:52:07,000
Speaker 3: we were to hear everything all at once, our brains

1011
00:52:07,000 --> 00:52:09,280
Speaker 3: would explode. I think it's just too much information.

1012
00:52:09,880 --> 00:52:13,400
Speaker 1: So people who get over stimulated, do they hear some

1013
00:52:13,520 --> 00:52:15,040
Speaker 1: of that or is that just a different thing?

1014
00:52:15,680 --> 00:52:19,560
Speaker 3: You know, that's a great question. That might be part

1015
00:52:19,600 --> 00:52:21,200
Speaker 3: of it, That might be part of it that they

1016
00:52:21,400 --> 00:52:24,040
Speaker 3: don't experience masking the same way. But the interesting is

1017
00:52:24,040 --> 00:52:27,080
Speaker 3: that it's very predictable, this masking and how it happens.

1018
00:52:27,080 --> 00:52:30,360
Speaker 3: It's very specific, and it's the same for children and

1019
00:52:30,400 --> 00:52:33,239
Speaker 3: for women and for men and for everybody. I don't

1020
00:52:33,280 --> 00:52:35,719
Speaker 3: know that there's a group of people to whom MP

1021
00:52:35,800 --> 00:52:40,560
Speaker 3: three's don't sound right objectively, like a blind test. Yeah,

1022
00:52:40,920 --> 00:52:45,200
Speaker 3: so yeah, I don't know. Maybe some people do experience

1023
00:52:45,280 --> 00:52:46,759
Speaker 3: less masking. Maybe that's part of it.

1024
00:52:47,600 --> 00:52:50,480
Speaker 1: So, Matt, I imagine you have loads of fun stories,

1025
00:52:50,520 --> 00:52:52,840
Speaker 1: in part because like Daniel and I always hit record

1026
00:52:52,880 --> 00:52:55,320
Speaker 1: and then we're like, we totally forgot. We hit record,

1027
00:52:55,360 --> 00:52:58,440
Speaker 1: and we have conversations about the most personal things happening

1028
00:52:58,440 --> 00:53:00,680
Speaker 1: in our lives, and then we send it to you,

1029
00:53:01,160 --> 00:53:04,120
Speaker 1: and so you it's a very asymmetrical relationship. I think

1030
00:53:04,160 --> 00:53:06,040
Speaker 1: you know a lot about the things that are going

1031
00:53:06,040 --> 00:53:08,040
Speaker 1: wrong in my life. I don't know nearly as much

1032
00:53:08,040 --> 00:53:09,520
Speaker 1: about what's going on in your life.

1033
00:53:09,600 --> 00:53:10,960
Speaker 3: I can make you a list if you want to

1034
00:53:11,200 --> 00:53:12,760
Speaker 3: even the playing field.

1035
00:53:13,800 --> 00:53:16,520
Speaker 1: I mean that lay it on me. But but so

1036
00:53:16,600 --> 00:53:18,800
Speaker 1: I imagine that one you know a lot of secrets

1037
00:53:18,800 --> 00:53:22,120
Speaker 1: from people and tell us some weird, funny stories from

1038
00:53:22,160 --> 00:53:24,280
Speaker 1: from life as an audio engineer.

1039
00:53:24,840 --> 00:53:30,280
Speaker 3: Uh, there are many inappropriate stories. Uh there's okay, here's

1040
00:53:30,360 --> 00:53:35,120
Speaker 3: one that is right in the middle between science and

1041
00:53:35,120 --> 00:53:40,279
Speaker 3: and some biology and perception, so mostly physics. So I

1042
00:53:40,440 --> 00:53:43,480
Speaker 3: was in the Santa Monica, which is uh is it

1043
00:53:43,600 --> 00:53:45,279
Speaker 3: a part of LA or is it outside of La

1044
00:53:45,719 --> 00:53:48,840
Speaker 3: boy Daniel, California.

1045
00:53:48,840 --> 00:53:50,799
Speaker 2: You should know this. I don't know. It's well, you know,

1046
00:53:50,840 --> 00:53:53,040
Speaker 2: there's the affective LA which is a big blob of

1047
00:53:53,120 --> 00:53:56,480
Speaker 2: urbanized area, then there's La County, then there's LA City,

1048
00:53:56,520 --> 00:53:58,560
Speaker 2: and I don't know the distinctions between all of them.

1049
00:53:58,840 --> 00:54:01,359
Speaker 3: Yeah, I don't understand how that's a municipality because when

1050
00:54:01,400 --> 00:54:04,040
Speaker 3: I was working there, they were shuttle us around from

1051
00:54:04,200 --> 00:54:07,200
Speaker 3: location to location, and the names of the neighborhoods and

1052
00:54:07,840 --> 00:54:11,120
Speaker 3: how things fit into the municipalities didn't make any sense. Anyway,

1053
00:54:11,160 --> 00:54:17,000
Speaker 3: I was in Santa Monica filming with Smokey Robinson, Wow

1054
00:54:17,840 --> 00:54:21,040
Speaker 3: for a very interesting documentary by the way, called Paid

1055
00:54:21,040 --> 00:54:25,239
Speaker 3: in Full on the BBC and CBC wherever you are,

1056
00:54:25,440 --> 00:54:30,480
Speaker 3: about the unfair treatment of black musicians throughout history. In

1057
00:54:30,520 --> 00:54:34,960
Speaker 3: any case, I was there with Smokey Robinson and they

1058
00:54:34,960 --> 00:54:37,840
Speaker 3: rented a special villa just for him. You know. The

1059
00:54:37,920 --> 00:54:42,360
Speaker 3: VIP treatment, and there's always this air of when a

1060
00:54:42,480 --> 00:54:47,160
Speaker 3: really important person comes in, no one stops the recording

1061
00:54:47,239 --> 00:54:51,040
Speaker 3: for any reason. Right, do your job be quiet. We

1062
00:54:51,120 --> 00:54:53,000
Speaker 3: have a very limited amount of time with this person.

1063
00:54:53,520 --> 00:54:55,759
Speaker 3: So I'm sitting there, I'm set up, I have my

1064
00:54:55,760 --> 00:54:58,399
Speaker 3: computer open, I've miked him up, I have a mic

1065
00:54:58,480 --> 00:55:00,520
Speaker 3: on him, I have a mic above him, and I

1066
00:55:00,560 --> 00:55:03,399
Speaker 3: have a spectrum analyzer open, just because I didn't really

1067
00:55:03,440 --> 00:55:06,120
Speaker 3: need it, but it shows you all the frequencies in

1068
00:55:06,160 --> 00:55:09,680
Speaker 3: real time. And suddenly during the interview, I see these

1069
00:55:09,840 --> 00:55:14,600
Speaker 3: sort of spikes at like five hurts, which is way

1070
00:55:14,640 --> 00:55:18,080
Speaker 3: lower than what humans can hear. And I'm realizing that

1071
00:55:18,800 --> 00:55:21,480
Speaker 3: even though the microphone is suspended and they're supposed to

1072
00:55:21,600 --> 00:55:24,880
Speaker 3: not pick anything up through the floor, I've created a

1073
00:55:24,920 --> 00:55:29,200
Speaker 3: seismometer by accident. Oh and I can see that there's

1074
00:55:29,200 --> 00:55:30,000
Speaker 3: an earthquake coming.

1075
00:55:30,200 --> 00:55:31,600
Speaker 2: Oh wow wow.

1076
00:55:31,800 --> 00:55:35,480
Speaker 3: And so in the middle of this interview, I go, hey,

1077
00:55:35,600 --> 00:55:37,400
Speaker 3: excuse me, and the director looks at me, like what

1078
00:55:37,440 --> 00:55:42,640
Speaker 3: are you doing, And I'm like, I'm very very sorry.

1079
00:55:43,040 --> 00:55:45,839
Speaker 3: I think there's an earthquake coming and we should get

1080
00:55:45,880 --> 00:55:48,319
Speaker 3: ready to evacuate or move away from the walls or

1081
00:55:48,680 --> 00:55:50,600
Speaker 3: get too though I don't remember what you're supposed to do.

1082
00:55:50,640 --> 00:55:53,240
Speaker 3: But we're in California. I expected everyone to know except

1083
00:55:53,280 --> 00:55:56,040
Speaker 3: for me, and they're like, what are you talking about?

1084
00:55:56,040 --> 00:56:00,200
Speaker 3: There's nothing happening, you've this is Smokey Robinson. What are

1085
00:56:00,200 --> 00:56:02,800
Speaker 3: you doing? I mean, everybody was nice, we're all friends.

1086
00:56:02,800 --> 00:56:06,960
Speaker 3: But as that's happening, the lights start to shake. Oh wow,

1087
00:56:07,160 --> 00:56:10,080
Speaker 3: and there's a rumble. And luckily it wasn't a big

1088
00:56:10,080 --> 00:56:13,439
Speaker 3: earthquake or anything. Everything was fine. But I predicted on an

1089
00:56:13,440 --> 00:56:16,360
Speaker 3: earthquake using a microphone and a computer screen.

1090
00:56:16,800 --> 00:56:17,520
Speaker 1: That's amazing.

1091
00:56:17,560 --> 00:56:19,600
Speaker 3: And I interrupted an important recording.

1092
00:56:19,440 --> 00:56:22,000
Speaker 1: And that's magic, right, it's Smokey Robinson is now like

1093
00:56:22,040 --> 00:56:23,960
Speaker 1: Matt Kesselman is magic. Everybody knows.

1094
00:56:24,400 --> 00:56:26,359
Speaker 3: He was like, that's okay, baby, he's got a very

1095
00:56:26,440 --> 00:56:27,239
Speaker 3: very soft voice.

1096
00:56:27,280 --> 00:56:27,799
Speaker 2: I love him.

1097
00:56:29,080 --> 00:56:31,640
Speaker 1: Well, that is a great story. I don't have any

1098
00:56:31,840 --> 00:56:36,640
Speaker 1: comparable stories for warning famous people about impending natural disasters.

1099
00:56:37,160 --> 00:56:38,760
Speaker 3: It's a very niche story.

1100
00:56:38,760 --> 00:56:42,759
Speaker 2: Market all right, well, thank you for coming on the show,

1101
00:56:42,840 --> 00:56:45,120
Speaker 2: and we're telling us all about what happens in the

1102
00:56:45,120 --> 00:56:48,160
Speaker 2: background to make us sound so good. Before you go,

1103
00:56:48,280 --> 00:56:50,359
Speaker 2: tell people where they can find out more about you

1104
00:56:50,440 --> 00:56:51,120
Speaker 2: and what's.

1105
00:56:51,000 --> 00:56:54,759
Speaker 3: Up to you. Can find me for everything at Matt

1106
00:56:55,160 --> 00:56:59,200
Speaker 3: Kesselman dot com. That's m A T K E S

1107
00:56:59,280 --> 00:57:02,880
Speaker 3: E l m N dot com. And if you don't

1108
00:57:03,000 --> 00:57:06,520
Speaker 3: like to spell things that are complicated, I just bought

1109
00:57:06,680 --> 00:57:10,040
Speaker 3: Thegoshfather dot com and dot cost to my website too,

1110
00:57:10,800 --> 00:57:12,680
Speaker 3: so you can use that instead.

1111
00:57:13,480 --> 00:57:16,200
Speaker 1: Matt, I appreciate you so much for so many reasons,

1112
00:57:16,240 --> 00:57:18,040
Speaker 1: and now I appreciate that you came on the show.

1113
00:57:18,160 --> 00:57:18,760
Speaker 1: You're the best.

1114
00:57:18,840 --> 00:57:21,120
Speaker 2: There are lots of reasons why the show sounds good

1115
00:57:21,200 --> 00:57:23,120
Speaker 2: and it's fun to listen to and easy to listen to,

1116
00:57:23,200 --> 00:57:25,600
Speaker 2: and many of those are because of Matt. Thank you

1117
00:57:25,680 --> 00:57:26,120
Speaker 2: very much.

1118
00:57:26,320 --> 00:57:27,960
Speaker 3: Amen, thank you for having me. This was a lot

1119
00:57:27,960 --> 00:57:28,280
Speaker 3: of fun.

1120
00:57:35,120 --> 00:57:37,560
Speaker 2: Thanks everybody for listening. Please go and do us a

1121
00:57:37,560 --> 00:57:40,840
Speaker 2: favor and rate the show on whatever podcast app you're using.

1122
00:57:40,920 --> 00:57:42,520
Speaker 2: It really helps people find us.

1123
00:57:43,040 --> 00:57:46,960
Speaker 1: Daniel and Kelly's Extraordinary Universe is edited by the amazing

1124
00:57:47,000 --> 00:57:47,720
Speaker 1: Matt Kesselman.

1125
00:57:47,960 --> 00:57:51,200
Speaker 2: He really is a wizard. You can also find us

1126
00:57:51,280 --> 00:57:56,400
Speaker 2: online on Blue Sky, Instagram, and x d K Universe.

1127
00:57:56,440 --> 00:57:57,680
Speaker 2: Come engage with us.

1128
00:57:57,920 --> 00:58:00,280
Speaker 1: You can email us I had questions at Dan and

1129
00:58:00,400 --> 00:58:03,120
Speaker 1: Kelly dot org. We really do want to hear from you,

1130
00:58:03,280 --> 00:58:03,520
Speaker 1: and you.

1131
00:58:03,520 --> 00:58:07,520
Speaker 2: Can find our website www dot danieland Kelly dot org,

1132
00:58:07,800 --> 00:58:10,920
Speaker 2: where you'll also find an invitation to join our discord

1133
00:58:11,000 --> 00:58:14,040
Speaker 2: where everybody comes and talks about the amazing.

1134
00:58:13,720 --> 00:58:17,960
Speaker 1: Universe, and we also have the most amazing moderators. This

1135
00:58:18,400 --> 00:58:21,040
Speaker 1: is an iHeart podcast. Thanks for joining us.