1 00:00:04,440 --> 00:00:12,280 Speaker 1: Welcome to tech Stuff, a production from iHeartRadio. Hey there, 2 00:00:12,280 --> 00:00:15,240 Speaker 1: and welcome to tech Stuff. I'm your host, Jonathan Strickland. 3 00:00:15,240 --> 00:00:18,439 Speaker 1: I'm an executive producer with iHeart Podcasts and how the 4 00:00:18,520 --> 00:00:21,000 Speaker 1: tech are you. You might be able to tell from 5 00:00:21,000 --> 00:00:24,880 Speaker 1: my voice that I have a cold, so I apologize 6 00:00:24,880 --> 00:00:28,000 Speaker 1: for that. But we're going to soldier on because I'm 7 00:00:28,040 --> 00:00:30,520 Speaker 1: back from vacation. It's time to get back to work, 8 00:00:30,960 --> 00:00:34,880 Speaker 1: and I love to talk about the intersection of technology 9 00:00:35,200 --> 00:00:38,800 Speaker 1: and music. So in past episodes, I've done shows about 10 00:00:38,840 --> 00:00:42,239 Speaker 1: how electric guitars work, the history of the Moog or 11 00:00:42,400 --> 00:00:47,040 Speaker 1: Mogue synthesizer, the evolution of various kinds of recordable media, 12 00:00:47,520 --> 00:00:51,440 Speaker 1: and much much more. But way back, like back in 13 00:00:51,560 --> 00:00:55,240 Speaker 1: two thousand and nine, my co host at the time, 14 00:00:55,360 --> 00:00:58,080 Speaker 1: Chris Palette, and I did a little episode of tech 15 00:00:58,120 --> 00:01:00,760 Speaker 1: Stuff about auto tune, and I thought it would be 16 00:01:00,840 --> 00:01:03,480 Speaker 1: fun to go back and revisit that topic. So this 17 00:01:03,560 --> 00:01:06,440 Speaker 1: is not a rerun. It's an all new episode about 18 00:01:06,480 --> 00:01:09,520 Speaker 1: the same subject. I haven't even listened to the old episode, 19 00:01:09,600 --> 00:01:12,080 Speaker 1: so I have no idea how much of what I 20 00:01:12,160 --> 00:01:14,440 Speaker 1: have to say is going to be a repeat. I 21 00:01:14,440 --> 00:01:16,560 Speaker 1: imagine a lot of it will be new, but I 22 00:01:16,640 --> 00:01:18,840 Speaker 1: don't know for sure. I figure there will be fewer 23 00:01:19,000 --> 00:01:22,319 Speaker 1: puns in this version compared to the last one, because, 24 00:01:22,760 --> 00:01:26,400 Speaker 1: contrary to popular belief, it was actually Chris Pillette who 25 00:01:26,440 --> 00:01:29,680 Speaker 1: made the most puns on tech stuff back in the day. 26 00:01:29,760 --> 00:01:32,360 Speaker 1: I got a reputation for it, and I don't get 27 00:01:32,360 --> 00:01:34,839 Speaker 1: me wrong, I won't shy away from a good pun, 28 00:01:35,160 --> 00:01:37,920 Speaker 1: and by that I mean a terrible pun. I love them, 29 00:01:38,319 --> 00:01:42,720 Speaker 1: but Chris like he loved them the way I love, 30 00:01:43,440 --> 00:01:48,080 Speaker 1: you know, rich Indian food, he dined upon puns, So 31 00:01:48,160 --> 00:01:50,720 Speaker 1: probably not as many in this one. But let's talk 32 00:01:50,760 --> 00:01:54,600 Speaker 1: about auto tune now. I think just about everyone knows, 33 00:01:54,880 --> 00:01:57,240 Speaker 1: or I think a lot of people know that Shares 34 00:01:57,480 --> 00:02:01,600 Speaker 1: song Believe, which came out nineteen ninety eight, was the 35 00:02:01,640 --> 00:02:05,920 Speaker 1: first major song to prominently use auto tune in an 36 00:02:05,960 --> 00:02:10,280 Speaker 1: effort to achieve a particular artistic effect, but the technology 37 00:02:10,280 --> 00:02:13,200 Speaker 1: had been around for more than a year at that point, 38 00:02:13,360 --> 00:02:16,560 Speaker 1: and the original intention wasn't to make a tool though 39 00:02:16,560 --> 00:02:20,119 Speaker 1: it actually draw attention to itself. Rather as the name 40 00:02:20,200 --> 00:02:24,799 Speaker 1: autotune suggests, it was intended to automatically nudge the pitch 41 00:02:24,919 --> 00:02:27,640 Speaker 1: of a musical note in the right direction so that 42 00:02:27,840 --> 00:02:31,160 Speaker 1: it would be in tune. That way, the occasional wrong 43 00:02:31,240 --> 00:02:35,480 Speaker 1: note could be subtly pushed into place, and it wouldn't 44 00:02:35,520 --> 00:02:38,320 Speaker 1: require you to do another take and then try to 45 00:02:38,400 --> 00:02:42,000 Speaker 1: splice together a great master recording. But even all of 46 00:02:42,040 --> 00:02:45,560 Speaker 1: that is getting way ahead of ourselves. To understand the 47 00:02:45,639 --> 00:02:50,720 Speaker 1: history of autotune, we must first learn about reflection seismology 48 00:02:51,080 --> 00:02:54,080 Speaker 1: as well as the oil industry. And I am being serious, 49 00:02:54,120 --> 00:02:56,480 Speaker 1: I'm not making a joke about this. As it turns out, 50 00:02:57,040 --> 00:02:59,960 Speaker 1: reflection seismology has a lot to do with our story 51 00:03:00,200 --> 00:03:03,680 Speaker 1: because the man who would go on to found the 52 00:03:03,760 --> 00:03:07,560 Speaker 1: company that would create autotune was a doctor, Andy Hildebrand, 53 00:03:07,919 --> 00:03:11,760 Speaker 1: who had made a career in using sound and complex 54 00:03:11,840 --> 00:03:17,440 Speaker 1: mathematical calculations to help oil companies, namely Exon, locate oil 55 00:03:17,520 --> 00:03:23,440 Speaker 1: deposits underground. So reflection seismology is in some ways similar 56 00:03:23,560 --> 00:03:27,919 Speaker 1: to sonar. So with a sonar system, you would beam 57 00:03:28,000 --> 00:03:30,560 Speaker 1: out pulses of sound waves. Typically we talk about this 58 00:03:30,760 --> 00:03:34,440 Speaker 1: in water, right, Like using sonar on a boat or 59 00:03:34,480 --> 00:03:37,200 Speaker 1: on a submarine, that kind of thing. You would pulse 60 00:03:37,240 --> 00:03:40,960 Speaker 1: out these sound waves and those soundwaves travel outward from 61 00:03:41,040 --> 00:03:45,080 Speaker 1: the source from the speaker. Essentially the transmitter, and if 62 00:03:45,080 --> 00:03:48,200 Speaker 1: there's something solid in the way of those sound waves, well, 63 00:03:48,480 --> 00:03:50,720 Speaker 1: the sound waves that hit that solid object, they're going 64 00:03:50,800 --> 00:03:54,040 Speaker 1: to reflect back toward the source. They'll become an echo. 65 00:03:54,440 --> 00:03:56,320 Speaker 1: This is what we get when we hear an echo. 66 00:03:56,560 --> 00:03:58,680 Speaker 1: If you are ever in a place where you make 67 00:03:58,920 --> 00:04:01,120 Speaker 1: a loud noise, then you hear the echo. It's because 68 00:04:01,160 --> 00:04:04,160 Speaker 1: the sound waves have traveled out from you, bounced off 69 00:04:04,160 --> 00:04:07,560 Speaker 1: something and came back to you. Well, if you if 70 00:04:07,560 --> 00:04:10,280 Speaker 1: you measure the amount of time it took for a 71 00:04:10,360 --> 00:04:13,760 Speaker 1: sound to leave you and then reflect off something else 72 00:04:14,000 --> 00:04:16,760 Speaker 1: and come back to you, you can figure out how 73 00:04:16,760 --> 00:04:20,320 Speaker 1: far away you are from that thing, right, because sound 74 00:04:20,360 --> 00:04:23,360 Speaker 1: is going to travel a specific speed away from you 75 00:04:23,920 --> 00:04:26,080 Speaker 1: and then hit the thing and then travel back to you. 76 00:04:26,320 --> 00:04:28,400 Speaker 1: So if you know how long it took, you can 77 00:04:28,480 --> 00:04:31,640 Speaker 1: do some very simple math and figure out how far 78 00:04:31,720 --> 00:04:34,600 Speaker 1: away that object is. So, for example, if you're on 79 00:04:34,640 --> 00:04:37,320 Speaker 1: a ship and use sonar to measure the distance between 80 00:04:37,360 --> 00:04:40,400 Speaker 1: you and the sea floor, you do a little math. Right, 81 00:04:40,440 --> 00:04:43,640 Speaker 1: you have to divide by two because it took a 82 00:04:43,640 --> 00:04:46,000 Speaker 1: certain amount of time to travel down then back up, 83 00:04:46,120 --> 00:04:49,159 Speaker 1: And you have to know how fast sound travels through water. 84 00:04:49,320 --> 00:04:52,080 Speaker 1: You have to have all these important bits of information 85 00:04:52,279 --> 00:04:53,960 Speaker 1: in your mind when you do this, but then you 86 00:04:53,960 --> 00:04:57,080 Speaker 1: can suss that out. You can say how deep the 87 00:04:57,080 --> 00:05:00,160 Speaker 1: ocean floor is from the surface. That saves you the 88 00:05:00,160 --> 00:05:02,320 Speaker 1: trouble of having to do it the really old fashioned way, 89 00:05:02,320 --> 00:05:05,880 Speaker 1: which typically involved lowering a weight on the end of 90 00:05:05,920 --> 00:05:10,040 Speaker 1: a knotted line, a knotted rope, and then you use 91 00:05:10,080 --> 00:05:12,400 Speaker 1: the knots to keep count of how deep in the 92 00:05:12,400 --> 00:05:15,320 Speaker 1: ocean you were. That's a sounding line. That's the other 93 00:05:15,360 --> 00:05:18,440 Speaker 1: way to see how how far down the ocean floor is. 94 00:05:18,640 --> 00:05:22,080 Speaker 1: But sonar made it way simpler, especially once you were 95 00:05:22,120 --> 00:05:26,680 Speaker 1: able to build that math into the sonar workstations. Reflection 96 00:05:26,839 --> 00:05:32,040 Speaker 1: seismology does something similar, but with seismic waves, and those 97 00:05:32,080 --> 00:05:35,240 Speaker 1: are waves that pass through the earth, and we typically 98 00:05:35,360 --> 00:05:39,479 Speaker 1: talk about seismic waves in connection with earthquakes or like 99 00:05:39,560 --> 00:05:42,440 Speaker 1: volcanic eruptions that kind of thing, and in fact earthquakes 100 00:05:42,440 --> 00:05:46,120 Speaker 1: were what inspired smarting Pants is to say, hey, if 101 00:05:46,160 --> 00:05:50,120 Speaker 1: we made something that could you know, create a huge 102 00:05:50,200 --> 00:05:53,040 Speaker 1: vibration through the earth, and something else that could detect 103 00:05:53,200 --> 00:05:58,440 Speaker 1: those vibrations, and we were able to calculate how long 104 00:05:58,480 --> 00:06:01,840 Speaker 1: it took for the instrumentation to pick up on the 105 00:06:01,920 --> 00:06:06,400 Speaker 1: echoes of that initial vibration event, we might be able 106 00:06:06,440 --> 00:06:10,920 Speaker 1: to figure out stuff that's actually underground. We could figure 107 00:06:10,920 --> 00:06:13,600 Speaker 1: out what is underground without having to dig it up 108 00:06:13,640 --> 00:06:17,920 Speaker 1: and see. Now, that's because the seismic waves will travel 109 00:06:18,000 --> 00:06:21,360 Speaker 1: at different speeds depending upon the density of the material 110 00:06:21,440 --> 00:06:24,680 Speaker 1: that they travel through. You've probably heard things like, you know, 111 00:06:24,800 --> 00:06:27,599 Speaker 1: sound travels at a consistent speed. That's true, but that 112 00:06:27,680 --> 00:06:31,279 Speaker 1: consistent speed is dependent upon the medium through which the 113 00:06:31,320 --> 00:06:34,479 Speaker 1: sound is traveling. So sound travels at a different speed 114 00:06:34,480 --> 00:06:37,120 Speaker 1: through the water than it does through the air or 115 00:06:37,160 --> 00:06:40,960 Speaker 1: through solid objects. You know, vibrations travel at different speeds 116 00:06:40,960 --> 00:06:43,719 Speaker 1: depending upon the medium. So at a very basic level, 117 00:06:43,760 --> 00:06:46,720 Speaker 1: a seismic wave will travel at a constant rate through 118 00:06:46,839 --> 00:06:51,200 Speaker 1: one kind of say rocky soil. But let's say there's 119 00:06:51,360 --> 00:06:57,000 Speaker 1: a place underground where that rocky soil gives way to 120 00:06:57,279 --> 00:07:03,600 Speaker 1: a different material says petroleum for example. Well, then the 121 00:07:03,720 --> 00:07:07,000 Speaker 1: speed of those sound waves is going to change. Moreover, 122 00:07:07,480 --> 00:07:10,640 Speaker 1: as the sound waves hit that barrier between one type 123 00:07:10,760 --> 00:07:13,560 Speaker 1: of material and another, some of the sound waves are 124 00:07:13,560 --> 00:07:16,120 Speaker 1: going to reflect off of that and become an echo. 125 00:07:16,960 --> 00:07:19,240 Speaker 1: Some of the sound waves will continue to penetrate through 126 00:07:19,440 --> 00:07:23,880 Speaker 1: the new material and through lots of observations. We gradually 127 00:07:23,920 --> 00:07:27,040 Speaker 1: began to learn about the different rates at which a 128 00:07:27,120 --> 00:07:31,160 Speaker 1: seismic wave will travel depending upon the medium it's traveling through, 129 00:07:31,720 --> 00:07:34,520 Speaker 1: and if it hits something really solid like bedrock, it 130 00:07:34,560 --> 00:07:39,760 Speaker 1: pretty much just echoes back. So here's how reflection seismology works. 131 00:07:39,880 --> 00:07:43,160 Speaker 1: From a very high level. You set up sensitive equipment 132 00:07:43,320 --> 00:07:47,040 Speaker 1: at different distances from a blast sight, and yeah, you're 133 00:07:47,160 --> 00:07:49,920 Speaker 1: likely to use something like explosives or maybe a really 134 00:07:50,000 --> 00:07:52,600 Speaker 1: powerful air gun. It has to be something that's going 135 00:07:52,680 --> 00:07:56,040 Speaker 1: to give a real jolt to the ground in order 136 00:07:56,080 --> 00:07:58,240 Speaker 1: to do this, because that's essentially what you're doing is 137 00:07:58,280 --> 00:08:02,400 Speaker 1: creating like a very localized earthquake. So this vibration travels 138 00:08:02,440 --> 00:08:04,960 Speaker 1: through the earth, and because you know how far away 139 00:08:04,960 --> 00:08:08,120 Speaker 1: you've set up your measuring equipment from that blast site, 140 00:08:08,280 --> 00:08:10,560 Speaker 1: you already have distance figured out, right. You know how 141 00:08:10,600 --> 00:08:14,960 Speaker 1: far away it is from the original source of the vibration, 142 00:08:15,400 --> 00:08:17,680 Speaker 1: and you measure the time it takes for your equipment 143 00:08:17,720 --> 00:08:22,360 Speaker 1: to pick up the echoes from that particular vibration event. 144 00:08:22,800 --> 00:08:25,680 Speaker 1: So you've got distance and now you have time. Now 145 00:08:25,720 --> 00:08:28,560 Speaker 1: you've got those variables sorted, so you can start to 146 00:08:28,600 --> 00:08:33,040 Speaker 1: work out what material is actually under the ground that 147 00:08:33,240 --> 00:08:36,679 Speaker 1: produces this particular result. And by doing that, you're kind 148 00:08:36,720 --> 00:08:39,360 Speaker 1: of like working backwards. You're using this information to draw 149 00:08:39,440 --> 00:08:42,600 Speaker 1: conclusions about what's under there, and that's where you can 150 00:08:42,679 --> 00:08:46,360 Speaker 1: start to make a determination as to whether or not 151 00:08:46,400 --> 00:08:48,720 Speaker 1: you're standing on top of a Beverly hillbilly is like 152 00:08:48,800 --> 00:08:52,080 Speaker 1: oil deposit, or maybe you're just on top of a 153 00:08:52,120 --> 00:08:54,800 Speaker 1: bunch of rocks or whatever. Now, in order to do 154 00:08:54,920 --> 00:08:58,840 Speaker 1: that what I just described, it's actually incredibly complicated. It 155 00:08:58,880 --> 00:09:02,800 Speaker 1: involves an awful lot of calculations in math, and it's 156 00:09:02,800 --> 00:09:04,560 Speaker 1: a lot of work. But then you have to think 157 00:09:04,559 --> 00:09:07,600 Speaker 1: that drilling for oil is even more work. That's a 158 00:09:07,679 --> 00:09:11,440 Speaker 1: huge endeavor. It costs a lot of time and money 159 00:09:11,480 --> 00:09:14,120 Speaker 1: and effort to do it, and like if you drill 160 00:09:14,120 --> 00:09:16,920 Speaker 1: in the wrong place, like that's a huge loss. So 161 00:09:17,200 --> 00:09:19,880 Speaker 1: you want the best possible information before you select a 162 00:09:19,960 --> 00:09:23,880 Speaker 1: drilling site, and reflection seismology is one way to obtain 163 00:09:23,960 --> 00:09:27,640 Speaker 1: information and to help make a decision. So doctor Hildebrand 164 00:09:27,840 --> 00:09:30,120 Speaker 1: was making a really good living out of this work, 165 00:09:30,679 --> 00:09:34,520 Speaker 1: but companies like Exon were saving hundreds of millions of 166 00:09:34,559 --> 00:09:39,040 Speaker 1: dollars through Hildebrand's approach of narrowing down potential drill sites, 167 00:09:39,240 --> 00:09:42,679 Speaker 1: and Hildebrand thought, you know, I'm not doing badly. I'm 168 00:09:42,679 --> 00:09:46,040 Speaker 1: making a decent living. But you know, Exon is making 169 00:09:46,080 --> 00:09:49,040 Speaker 1: out like a bandit. They're saving like half a billion 170 00:09:49,120 --> 00:09:52,840 Speaker 1: dollars a year or whatever using this technology. Maybe if 171 00:09:52,920 --> 00:09:58,880 Speaker 1: I apply my knowledge and skill set in a company 172 00:09:59,080 --> 00:10:02,800 Speaker 1: that I own, I might actually, you know, do better 173 00:10:02,880 --> 00:10:06,599 Speaker 1: than just working for Exon. So Hildebrand left Exon in 174 00:10:06,679 --> 00:10:10,640 Speaker 1: nineteen seventy nine and he founded a company called Landmark Graphics, 175 00:10:10,920 --> 00:10:13,760 Speaker 1: which at first sounds like, you know, it's a company 176 00:10:13,760 --> 00:10:17,839 Speaker 1: that makes computer graphics, which is not untrue, but that 177 00:10:17,960 --> 00:10:21,480 Speaker 1: wasn't It wasn't just general graphics. This company was still 178 00:10:21,600 --> 00:10:25,360 Speaker 1: rooted in the oil industry. Hildebrand's team developed and produced 179 00:10:25,400 --> 00:10:30,200 Speaker 1: workstations that could take incoming seismic information from these these 180 00:10:30,280 --> 00:10:33,560 Speaker 1: you know, soundings that they do and generate three dimensional 181 00:10:33,679 --> 00:10:37,840 Speaker 1: seismic maps based upon the data. And again, it was 182 00:10:37,880 --> 00:10:41,960 Speaker 1: incredibly complicated. You had to analyze so many different points 183 00:10:41,960 --> 00:10:46,280 Speaker 1: of information in order to create this three dimensional representation 184 00:10:46,360 --> 00:10:49,520 Speaker 1: of what's under the ground. But it worked and it 185 00:10:49,520 --> 00:10:52,440 Speaker 1: made Hildebrand very successful. He stuck with it for a 186 00:10:52,520 --> 00:10:56,960 Speaker 1: decade until nineteen eighty nine, whereupon he retired and he 187 00:10:57,040 --> 00:11:00,840 Speaker 1: decided to return his attention to a different passion he 188 00:11:00,920 --> 00:11:04,800 Speaker 1: had had since he was a kid, which was music. 189 00:11:05,200 --> 00:11:08,840 Speaker 1: Now Hildebrand wasn't just a music fan, he was a musician. 190 00:11:08,960 --> 00:11:12,680 Speaker 1: He had played flute professionally. He had been a studio 191 00:11:12,800 --> 00:11:15,520 Speaker 1: musician for some time. He had paid his way through 192 00:11:15,520 --> 00:11:20,520 Speaker 1: college partly by giving flute lessons to musicians, So he 193 00:11:20,640 --> 00:11:23,760 Speaker 1: decided he would go back to school as a retiree 194 00:11:24,120 --> 00:11:27,720 Speaker 1: and study composition and techniques. He attended Rice University to 195 00:11:27,760 --> 00:11:31,400 Speaker 1: do this. While he was back in college, he encountered 196 00:11:31,920 --> 00:11:36,200 Speaker 1: some newer technologies in the music space, like music samplers 197 00:11:36,240 --> 00:11:39,560 Speaker 1: and synthesizers. So these were machines designed to take a 198 00:11:39,720 --> 00:11:43,720 Speaker 1: sample of a sound like a flute, and then allow 199 00:11:43,840 --> 00:11:47,120 Speaker 1: a keyboard musician to recreate those sounds on a synthesizer. 200 00:11:47,480 --> 00:11:50,960 Speaker 1: The only thing is that Hildebrand thought they sounded terrible, 201 00:11:51,600 --> 00:11:54,760 Speaker 1: and partly it was because there was a limitation on 202 00:11:54,920 --> 00:11:58,360 Speaker 1: how much data a synthesizer could actually handle, so it 203 00:11:58,400 --> 00:12:04,160 Speaker 1: couldn't really replicate sound naturally. The sound it replicated would 204 00:12:04,200 --> 00:12:07,440 Speaker 1: be like a gross approximation of the original sound, So 205 00:12:07,520 --> 00:12:10,559 Speaker 1: Hildebrand wasn't really impressed, but he thought that there was 206 00:12:10,640 --> 00:12:13,960 Speaker 1: room for improvement, and he developed a technique to compress 207 00:12:14,080 --> 00:12:18,640 Speaker 1: audio data so that synthesizers could more effectively handle information 208 00:12:19,200 --> 00:12:23,240 Speaker 1: and make notes, to produce notes that sounded more natural 209 00:12:23,320 --> 00:12:27,360 Speaker 1: and less synthetic. He released his software as a product 210 00:12:27,360 --> 00:12:32,320 Speaker 1: called Infinity, and while this tool would revolutionize the orchestration 211 00:12:32,400 --> 00:12:35,199 Speaker 1: process for stuff like film and television, it did not 212 00:12:36,240 --> 00:12:40,240 Speaker 1: revolutionize doctor Hildebrand's bank account. He didn't actually see much 213 00:12:40,320 --> 00:12:43,439 Speaker 1: of that success himself because what actually happened was other 214 00:12:43,520 --> 00:12:47,040 Speaker 1: companies purchased copies of Infinity and then bundled it with 215 00:12:47,120 --> 00:12:50,760 Speaker 1: their own audio processing tools, and then sold those audio 216 00:12:50,880 --> 00:12:54,480 Speaker 1: processing packages to other people and companies, and it kind 217 00:12:54,480 --> 00:12:58,520 Speaker 1: of cut Hildebrand out of the picture. So while others 218 00:12:58,520 --> 00:13:02,640 Speaker 1: were benefiting from his work, he did not see that 219 00:13:02,760 --> 00:13:07,320 Speaker 1: much success. It did, however, again have an enormous impact 220 00:13:07,520 --> 00:13:12,400 Speaker 1: on orchestrations, like According to doctor Hildebrand, he was the 221 00:13:12,440 --> 00:13:16,040 Speaker 1: reason why the Los Angeles Orchestra hit real hard times 222 00:13:16,080 --> 00:13:20,360 Speaker 1: in the nineteen nineties because his tools allowed composers to 223 00:13:20,559 --> 00:13:25,640 Speaker 1: sample various musical instruments and create a natural enough representation 224 00:13:26,240 --> 00:13:28,880 Speaker 1: of those sounds to be able to create a synthetic 225 00:13:29,040 --> 00:13:32,960 Speaker 1: orchestra that sounded more or less like a real one. 226 00:13:33,040 --> 00:13:34,720 Speaker 1: So there was no need to go and hire a 227 00:13:34,720 --> 00:13:38,199 Speaker 1: real orchestra to orchestrate your film or TV project. You 228 00:13:38,240 --> 00:13:41,360 Speaker 1: could do it yourself. I've actually heard some some of 229 00:13:41,400 --> 00:13:45,040 Speaker 1: my favorite music scores. When I listened closely, I can 230 00:13:45,120 --> 00:13:49,800 Speaker 1: tell like, oh, that's not a real cellist. That's a 231 00:13:49,880 --> 00:13:54,800 Speaker 1: synthesizer playing a sample of a cello that sounds almost, 232 00:13:54,800 --> 00:13:58,400 Speaker 1: but not quite like the real thing. Anyway, we can 233 00:13:58,480 --> 00:14:01,360 Speaker 1: thank doctor Hildebrand for that. I'll talk more about what 234 00:14:01,440 --> 00:14:05,640 Speaker 1: we could thank doctor Hildebrand for, specifically auto tune, but 235 00:14:05,720 --> 00:14:07,960 Speaker 1: first let's take a quick break so we could thank 236 00:14:08,000 --> 00:14:21,160 Speaker 1: some other people, namely our sponsors. Will be right back. Okay, 237 00:14:21,320 --> 00:14:24,000 Speaker 1: So before we left off, I was talking about how 238 00:14:24,040 --> 00:14:29,000 Speaker 1: doctor Hildebrand had released a program called Infinity that improved 239 00:14:29,160 --> 00:14:34,440 Speaker 1: the performance of synthesizers and samplers. But in nineteen ninety 240 00:14:34,440 --> 00:14:37,040 Speaker 1: he decided to take an extra step. He founded a 241 00:14:37,240 --> 00:14:41,840 Speaker 1: new company. He called it Antare's Audio Technology, and this 242 00:14:41,880 --> 00:14:46,440 Speaker 1: would be his music company, his music technology company that 243 00:14:46,560 --> 00:14:50,760 Speaker 1: would ultimately produce autotune. And he knew that technology was 244 00:14:50,840 --> 00:14:53,480 Speaker 1: poised to make a huge impact on the music industry 245 00:14:53,520 --> 00:14:56,240 Speaker 1: and already had been like, that's kind of the history 246 00:14:56,240 --> 00:14:58,960 Speaker 1: of modern music is how technology has shaped it. But 247 00:14:59,000 --> 00:15:01,440 Speaker 1: he knew we were on the brink of another revolution. 248 00:15:01,520 --> 00:15:03,880 Speaker 1: He just wasn't exactly sure how that was going to 249 00:15:03,960 --> 00:15:08,280 Speaker 1: manifest now. According to an article by Simon Reynolds, it's 250 00:15:08,280 --> 00:15:12,320 Speaker 1: titled How Autotune Revolutionized the Sound of Popular Music, and 251 00:15:12,360 --> 00:15:16,760 Speaker 1: it was published in Pitchfork, the actual birth of Hildebrand's 252 00:15:16,800 --> 00:15:20,560 Speaker 1: idea for autotune grew out of a casual lunch with 253 00:15:20,640 --> 00:15:24,160 Speaker 1: some of his friends and peers back in nineteen ninety 254 00:15:24,160 --> 00:15:29,520 Speaker 1: five during a National Association of Music Merchants conference. So 255 00:15:29,600 --> 00:15:32,720 Speaker 1: he's at this conference, he's meeting with other people in 256 00:15:32,760 --> 00:15:36,360 Speaker 1: the music and technology spheres, and at this lunch, one 257 00:15:36,400 --> 00:15:41,320 Speaker 1: of the attendees jokingly suggested that what Hildebrand should do 258 00:15:41,440 --> 00:15:43,960 Speaker 1: next is develop a technology that would allow her to 259 00:15:44,040 --> 00:15:47,000 Speaker 1: sing on key, like, can you make a box that 260 00:15:47,120 --> 00:15:51,120 Speaker 1: lets me sing well? And while this was presented as 261 00:15:51,160 --> 00:15:55,600 Speaker 1: a joke, ultimately Hildebrand would think, huh, could I do 262 00:15:55,760 --> 00:16:00,400 Speaker 1: that now? According to Zachary Crockett's article, which is the 263 00:16:00,440 --> 00:16:04,080 Speaker 1: Mathematical Genius of auto Tune, this one in price Anomics 264 00:16:04,480 --> 00:16:07,760 Speaker 1: This wasn't like a light bulb moment where the moment 265 00:16:07,840 --> 00:16:11,600 Speaker 1: this woman says the thing, Hildebrand immediately thinks, ah, that's 266 00:16:11,640 --> 00:16:14,480 Speaker 1: what I shall do. Actually, it took like another six 267 00:16:14,560 --> 00:16:18,560 Speaker 1: months before Hildebrand really kind of revisited the concept and thought, 268 00:16:18,920 --> 00:16:21,880 Speaker 1: maybe there's something here. But in order to do that, 269 00:16:22,600 --> 00:16:24,800 Speaker 1: he would have to develop a technology that could do 270 00:16:24,920 --> 00:16:27,680 Speaker 1: a few things really well, all of which are a 271 00:16:27,760 --> 00:16:30,640 Speaker 1: bit tricky. One is it would need to detect the 272 00:16:30,720 --> 00:16:33,200 Speaker 1: pitch that someone was singing in. For example, if you're 273 00:16:33,280 --> 00:16:36,160 Speaker 1: using it for vocals, and so you would need to 274 00:16:36,160 --> 00:16:40,440 Speaker 1: be able to detect exactly the frequency that was being sung. 275 00:16:40,840 --> 00:16:44,640 Speaker 1: You would need to then also be able to have 276 00:16:44,880 --> 00:16:49,800 Speaker 1: a list of tones that were in the whatever key 277 00:16:49,880 --> 00:16:52,120 Speaker 1: you were supposed to be singing it. So I don't 278 00:16:52,120 --> 00:16:54,520 Speaker 1: want to get into music theory, because goodness knows, I 279 00:16:54,520 --> 00:16:56,720 Speaker 1: don't know that much about it myself, and I would 280 00:16:56,720 --> 00:16:59,240 Speaker 1: just mess things up. But you know, if you're singing 281 00:16:59,240 --> 00:17:02,160 Speaker 1: in a specific key, there are particular tones that belong 282 00:17:02,240 --> 00:17:04,920 Speaker 1: to that key. And often when we sing and we're 283 00:17:05,119 --> 00:17:07,480 Speaker 1: a little off pitch, what we need is to be 284 00:17:07,920 --> 00:17:11,560 Speaker 1: gently nudged a little up or a little down, a 285 00:17:11,600 --> 00:17:14,480 Speaker 1: little sharp or a little flat in order to hit 286 00:17:14,600 --> 00:17:18,080 Speaker 1: a semitone that belongs in that key. So it needs 287 00:17:18,119 --> 00:17:22,600 Speaker 1: to also quote unquote know which tones are appropriate, and 288 00:17:22,600 --> 00:17:25,480 Speaker 1: then it has to be able to digitally alter the 289 00:17:25,600 --> 00:17:30,440 Speaker 1: incoming pitch the actual sung note, and then guide it 290 00:17:30,720 --> 00:17:34,119 Speaker 1: to match that of a target note. Now, ultimately that 291 00:17:34,200 --> 00:17:37,080 Speaker 1: all sounds like a pretty simple idea, but in reality 292 00:17:37,560 --> 00:17:42,199 Speaker 1: to achieve this it was incredibly complex. Ultimately, also, the 293 00:17:42,240 --> 00:17:45,159 Speaker 1: toolould need to work in real time for live performances. 294 00:17:45,240 --> 00:17:47,840 Speaker 1: Like it's one thing to have this for the studio, right, 295 00:17:47,880 --> 00:17:50,800 Speaker 1: because even if you don't have an automatic, you could 296 00:17:50,840 --> 00:17:53,800 Speaker 1: have a tool where an engineer could fiddle with some 297 00:17:53,920 --> 00:17:57,919 Speaker 1: controls and gently alter the pitch of a performance to 298 00:17:57,960 --> 00:18:00,520 Speaker 1: get it closer to being where it needs to be. 299 00:18:00,880 --> 00:18:03,119 Speaker 1: It would be preferable to have that automated so that 300 00:18:03,160 --> 00:18:04,800 Speaker 1: you don't have to go through there and do the 301 00:18:04,800 --> 00:18:08,800 Speaker 1: manual process. But even so, like in a recording setting, 302 00:18:08,880 --> 00:18:11,199 Speaker 1: you don't have to have it be real time necessarily, 303 00:18:11,280 --> 00:18:13,119 Speaker 1: but if you're doing a live performance, you do have 304 00:18:13,160 --> 00:18:15,560 Speaker 1: to have a real time. If someone's up there singing 305 00:18:15,920 --> 00:18:19,280 Speaker 1: and they just hit a flat note when they're not 306 00:18:19,320 --> 00:18:23,280 Speaker 1: supposed to, that could really be a memorable moment and 307 00:18:23,320 --> 00:18:25,439 Speaker 1: not in a great way. So having a tool that 308 00:18:25,520 --> 00:18:28,920 Speaker 1: could gently account for that and fix it in real 309 00:18:29,000 --> 00:18:32,880 Speaker 1: time would be really helpful. But this would mean that 310 00:18:33,200 --> 00:18:35,439 Speaker 1: this tool would have to be able to process a 311 00:18:35,560 --> 00:18:40,920 Speaker 1: huge amount of sound data extremely quickly to make millisecond 312 00:18:40,920 --> 00:18:45,880 Speaker 1: decisions like split millisecond decisions relating to how to shape 313 00:18:45,920 --> 00:18:49,879 Speaker 1: a note moment by moment. Now it does help if 314 00:18:49,920 --> 00:18:53,320 Speaker 1: we also think of sound in terms of mathematics. We 315 00:18:53,480 --> 00:18:56,080 Speaker 1: describe sound in different ways, right, But some of those 316 00:18:56,400 --> 00:19:00,280 Speaker 1: relate specifically to how sound looks to us. If we 317 00:19:00,400 --> 00:19:05,280 Speaker 1: plot sound on like a wave chart, right. For example, 318 00:19:05,400 --> 00:19:07,760 Speaker 1: sounds can be really loud or they can be really quiet, 319 00:19:08,119 --> 00:19:12,040 Speaker 1: and that is volume, But it can also relate to amplitude. 320 00:19:12,440 --> 00:19:14,920 Speaker 1: When you think of a sound wave. The amplitude of 321 00:19:14,960 --> 00:19:18,240 Speaker 1: a sound wave describes how tall those peaks are or 322 00:19:18,280 --> 00:19:22,720 Speaker 1: how low the valleys are. The distance between the furthest 323 00:19:23,080 --> 00:19:26,440 Speaker 1: point of a peak or valley and the zero line. 324 00:19:26,800 --> 00:19:30,320 Speaker 1: That's your amplitude. But we also describe sound in terms 325 00:19:30,320 --> 00:19:34,880 Speaker 1: of pitch or frequencies. Higher frequencies correspond to higher pitches, 326 00:19:35,200 --> 00:19:37,560 Speaker 1: And if we plot a sound wave, let's say that 327 00:19:37,600 --> 00:19:40,919 Speaker 1: we plot it so that the x axis is a 328 00:19:41,000 --> 00:19:46,639 Speaker 1: demarcation of time, so we have one second listed there, 329 00:19:46,880 --> 00:19:49,640 Speaker 1: like the x axis is one second. If there's one 330 00:19:49,760 --> 00:19:52,399 Speaker 1: wave that we draw so that the wave begins at 331 00:19:52,440 --> 00:19:54,880 Speaker 1: the zero point and ends at the one second point, 332 00:19:55,119 --> 00:19:58,760 Speaker 1: then we have a one hurtz sound wave. A hurtz 333 00:19:59,200 --> 00:20:03,040 Speaker 1: is just a measurement a frequency. It refers to one 334 00:20:03,240 --> 00:20:06,960 Speaker 1: cycle per second. So if a wave is one hurts, 335 00:20:06,960 --> 00:20:09,240 Speaker 1: it means it takes one second for one of those 336 00:20:09,240 --> 00:20:13,120 Speaker 1: sound waves to fully pass a given point where you're 337 00:20:13,160 --> 00:20:17,920 Speaker 1: measuring the sound waves, right, If two waves pass that 338 00:20:18,040 --> 00:20:20,679 Speaker 1: point within one second, then you're talking about two hurts, 339 00:20:21,040 --> 00:20:23,440 Speaker 1: you know. Just so that we know, the typical human 340 00:20:23,480 --> 00:20:28,080 Speaker 1: hearing range is anywhere between twenty and twenty thousand hurts. 341 00:20:28,320 --> 00:20:31,359 Speaker 1: So one or two hurts sound We wouldn't even perceive it, 342 00:20:31,359 --> 00:20:33,720 Speaker 1: at least not as sound. If it was a great 343 00:20:33,800 --> 00:20:36,840 Speaker 1: enough amplitude, you could potentially perceive it as vibration, but 344 00:20:36,880 --> 00:20:40,359 Speaker 1: you wouldn't feel it, you wouldn't hear it. But between 345 00:20:40,359 --> 00:20:43,960 Speaker 1: twenty and twenty thousand hurts, that falls into the typical 346 00:20:44,040 --> 00:20:46,120 Speaker 1: range of human hearing. Of course, as we get older, 347 00:20:46,119 --> 00:20:49,440 Speaker 1: we start to lose the ability to hear those higher frequencies. 348 00:20:50,040 --> 00:20:53,600 Speaker 1: These days, I think my hearing tops out around sixteen 349 00:20:53,640 --> 00:20:57,040 Speaker 1: to seventeen thousand hurts somewhere around there. Like once you 350 00:20:57,080 --> 00:21:00,320 Speaker 1: get beyond that, I don't hear anything, whereas younger people 351 00:21:00,320 --> 00:21:04,120 Speaker 1: could hear it. Anyway, Hildebrand was working with music on 352 00:21:04,160 --> 00:21:07,840 Speaker 1: this mathematical level. He was analyzing music to recognize where 353 00:21:07,840 --> 00:21:11,520 Speaker 1: the frequencies were and where they should be, and to 354 00:21:11,600 --> 00:21:14,760 Speaker 1: then shape the sound wave so that it would fit 355 00:21:15,440 --> 00:21:18,879 Speaker 1: what the ideal would be where it would be on key. 356 00:21:19,520 --> 00:21:22,840 Speaker 1: He was not the first person to attempt to do this, however, 357 00:21:22,920 --> 00:21:27,400 Speaker 1: Earlier engineers had largely abandoned the quest because the signal 358 00:21:27,480 --> 00:21:32,800 Speaker 1: processing and statistical analysis needs were so high. They were 359 00:21:32,840 --> 00:21:37,000 Speaker 1: so extreme that you would need a supercomputer dedicated to 360 00:21:37,040 --> 00:21:39,119 Speaker 1: the task to be able to do it. There's just 361 00:21:39,200 --> 00:21:43,399 Speaker 1: too much data to process in too little time to 362 00:21:43,480 --> 00:21:46,720 Speaker 1: be able to do anything meaningful with it. Hildebrand determined 363 00:21:46,720 --> 00:21:49,639 Speaker 1: that yeah, to fully analyze music, you would have to 364 00:21:49,720 --> 00:21:54,280 Speaker 1: run thousands or millions of calculations, but many of those 365 00:21:54,320 --> 00:21:57,280 Speaker 1: calculations were actually redundant at the end of the day, 366 00:21:57,440 --> 00:22:00,919 Speaker 1: and eliminating the redundancy would not affect the quality of 367 00:22:00,960 --> 00:22:04,240 Speaker 1: the outcome, and so in his words he quote changed 368 00:22:04,400 --> 00:22:09,560 Speaker 1: a million multiply ads into just four. It was a trick, 369 00:22:10,080 --> 00:22:14,040 Speaker 1: a mathematical trick. End quote. That's ron the article I 370 00:22:14,119 --> 00:22:19,479 Speaker 1: mentioned earlier by Zachary Crockett. So yeah, in prisonomics, pretty 371 00:22:20,119 --> 00:22:24,520 Speaker 1: phenomenal that he was able to recognize that ultimately he 372 00:22:24,600 --> 00:22:28,879 Speaker 1: just needed these four processes to really be able to 373 00:22:29,240 --> 00:22:33,560 Speaker 1: zero in on pitch correction. So Hildebrand developed the autotune 374 00:22:33,600 --> 00:22:36,760 Speaker 1: technology in nineteen ninety six. He actually used to customized 375 00:22:36,840 --> 00:22:40,080 Speaker 1: Mac computer or specialized Mac computer as the way I've 376 00:22:40,119 --> 00:22:43,119 Speaker 1: seen it explained. I don't know in what way it 377 00:22:43,160 --> 00:22:46,000 Speaker 1: was specialized. I just know it was a Mac. And 378 00:22:46,200 --> 00:22:50,000 Speaker 1: he brought his software to the next National Association of 379 00:22:50,119 --> 00:22:52,840 Speaker 1: Music Merchant's conference, if you remember, that was the same 380 00:22:52,920 --> 00:22:56,639 Speaker 1: conference where one of his lunch companions had inspired the 381 00:22:56,880 --> 00:22:59,639 Speaker 1: idea for autotune in the first place. To say that 382 00:22:59,680 --> 00:23:03,040 Speaker 1: he felt interest in his product at this conference is 383 00:23:03,119 --> 00:23:07,639 Speaker 1: really under selling it, and it's understandable why. So let's 384 00:23:07,760 --> 00:23:12,119 Speaker 1: talk about the process of creating a master recording for 385 00:23:12,200 --> 00:23:15,520 Speaker 1: a song. If you want to get a perfect take 386 00:23:16,040 --> 00:23:20,800 Speaker 1: of a song, where this is the master recording, this 387 00:23:20,840 --> 00:23:23,879 Speaker 1: is what you want to use in order to you know, 388 00:23:23,960 --> 00:23:28,560 Speaker 1: create your album. You can't just hope that everything lines 389 00:23:28,640 --> 00:23:32,560 Speaker 1: up when you hit record and that everyone is playing 390 00:23:32,640 --> 00:23:37,080 Speaker 1: seamlessly together and no one makes a mistake. Invariably something 391 00:23:37,359 --> 00:23:40,159 Speaker 1: is going to be off. Maybe one of the musicians 392 00:23:40,200 --> 00:23:42,560 Speaker 1: is lagging behind the others and it might not even 393 00:23:42,600 --> 00:23:46,159 Speaker 1: be detectable at first, but upon closer examination you're like, ooh, 394 00:23:46,200 --> 00:23:49,280 Speaker 1: you came in late, or you came in too early 395 00:23:49,359 --> 00:23:52,960 Speaker 1: or whatever. Or the drummer is not keeping perfect time, 396 00:23:53,040 --> 00:23:55,680 Speaker 1: whatever it may be. Maybe someone hits a wrong note, 397 00:23:56,000 --> 00:23:59,280 Speaker 1: either while playing an instrument or while singing, or maybe both. 398 00:23:59,760 --> 00:24:02,760 Speaker 1: But what it means for engineers is that they'll need 399 00:24:02,800 --> 00:24:06,760 Speaker 1: to get another take where that mistake isn't there, and 400 00:24:06,800 --> 00:24:09,840 Speaker 1: they'll probably need another take and another take, And if 401 00:24:09,840 --> 00:24:12,800 Speaker 1: you want the perfect performance, this could mean recording the 402 00:24:12,800 --> 00:24:16,879 Speaker 1: same track dozens or gosh even hundreds of times and 403 00:24:16,920 --> 00:24:20,840 Speaker 1: then slowly picking apart each recording in order to piece 404 00:24:20,920 --> 00:24:24,919 Speaker 1: together a perfect edit. And that alone is hard because 405 00:24:24,960 --> 00:24:29,520 Speaker 1: just lining up the different takes isn't always the easiest 406 00:24:29,520 --> 00:24:32,160 Speaker 1: thing to do. You don't always have a seamless point 407 00:24:32,200 --> 00:24:34,520 Speaker 1: where you could line up take one would take two. 408 00:24:34,600 --> 00:24:37,720 Speaker 1: Like again, if the band is playing at a slightly 409 00:24:37,840 --> 00:24:41,200 Speaker 1: different pace in the second take. You can't easily line 410 00:24:41,280 --> 00:24:43,680 Speaker 1: up the two different ones to you know, even if 411 00:24:43,800 --> 00:24:46,359 Speaker 1: like one had an accident and the other one didn't, 412 00:24:46,760 --> 00:24:49,480 Speaker 1: you can't necessarily put them together to make the perfect recording. 413 00:24:49,520 --> 00:24:55,240 Speaker 1: So this is a really laborious, time consuming and expensive process. 414 00:24:55,640 --> 00:25:02,000 Speaker 1: Expensive because studio time is limited, so expensive. Hildebrand's invention 415 00:25:02,480 --> 00:25:05,359 Speaker 1: would take a ton of that effort off the table, 416 00:25:05,400 --> 00:25:08,439 Speaker 1: at least for vocals, because rather than re recording a 417 00:25:08,520 --> 00:25:12,280 Speaker 1: billion times, you could get maybe just one good take, 418 00:25:12,800 --> 00:25:16,159 Speaker 1: one decent take even and then use pitch correction for 419 00:25:16,200 --> 00:25:18,399 Speaker 1: any little flubs that might have found their way in 420 00:25:18,520 --> 00:25:21,880 Speaker 1: during the recording process. So it was a huge time saver, 421 00:25:22,080 --> 00:25:27,680 Speaker 1: and time is money. So immediately studios recognized the value 422 00:25:28,200 --> 00:25:32,560 Speaker 1: of Hildebrand's product and they rushed to get in on that, 423 00:25:33,040 --> 00:25:37,480 Speaker 1: and the tool absolutely revolutionized the recording industry. Studios that 424 00:25:37,560 --> 00:25:41,760 Speaker 1: incorporated auto tune were able to work much faster than 425 00:25:41,800 --> 00:25:44,639 Speaker 1: their competitors. They were able to cycle clients in and 426 00:25:44,680 --> 00:25:48,080 Speaker 1: out of their studios more quickly. That meant getting more 427 00:25:48,080 --> 00:25:52,120 Speaker 1: work done and more money coming in, and efficiency skyrocketed. 428 00:25:52,200 --> 00:25:55,720 Speaker 1: So studios that were not on the auto tuned train 429 00:25:56,040 --> 00:25:59,560 Speaker 1: soon found themselves getting out competed, and they ended up 430 00:25:59,600 --> 00:26:02,720 Speaker 1: adopting the technology as well, because it was either adopted 431 00:26:03,280 --> 00:26:06,600 Speaker 1: or go out of business. It also wasn't enough to 432 00:26:06,720 --> 00:26:10,240 Speaker 1: just be able to change the pitch of a note. 433 00:26:10,440 --> 00:26:13,040 Speaker 1: Auto tune would also have to be able to adjust 434 00:26:13,080 --> 00:26:17,000 Speaker 1: that pitch on a sliding scale of rapidity. That is, 435 00:26:17,240 --> 00:26:21,399 Speaker 1: the sound would be unnatural if you were to correct 436 00:26:21,440 --> 00:26:24,439 Speaker 1: a note instantaneously. It would be the effect that we 437 00:26:24,480 --> 00:26:27,800 Speaker 1: associate with autotune, that robotic effect. That's if you were 438 00:26:27,840 --> 00:26:32,440 Speaker 1: to change the pitch correction super fast. You don't want 439 00:26:32,440 --> 00:26:36,840 Speaker 1: to do that if you want the tool to remain unnoticed. So, 440 00:26:37,080 --> 00:26:39,800 Speaker 1: particularly for stuff like slow ballads, you would not have 441 00:26:40,000 --> 00:26:44,840 Speaker 1: a more gradual approach to correcting a pitch. So Hildebrand 442 00:26:44,880 --> 00:26:47,640 Speaker 1: wanted a tool that would let users determine how quickly 443 00:26:48,119 --> 00:26:51,000 Speaker 1: the note would get nudged to the correct pitch, and 444 00:26:51,119 --> 00:26:53,879 Speaker 1: the scale essentially went from zero to ten. The higher 445 00:26:53,920 --> 00:26:57,080 Speaker 1: settings would have longer adjustment times, so for a really 446 00:26:57,240 --> 00:27:00,320 Speaker 1: slow song, you might go with a nine or a 447 00:27:00,440 --> 00:27:02,919 Speaker 1: ten to let the note find its way to the 448 00:27:02,960 --> 00:27:06,320 Speaker 1: right pitch more gradually. Faster songs like rock and roll 449 00:27:06,440 --> 00:27:09,679 Speaker 1: type stuff or a rap or R and B. You 450 00:27:09,800 --> 00:27:13,200 Speaker 1: might require a lower setting, like fast rock songs, you 451 00:27:13,280 --> 00:27:15,439 Speaker 1: might need a two, three, or maybe even down to 452 00:27:15,520 --> 00:27:19,600 Speaker 1: a one. The zero setting. Really, Hildebrand just added that 453 00:27:19,640 --> 00:27:22,840 Speaker 1: for kicks, So essentially the software would immediately correct the 454 00:27:22,880 --> 00:27:27,639 Speaker 1: pitch upon detecting an incoming signal. And this sounded weird 455 00:27:27,920 --> 00:27:31,240 Speaker 1: and unnatural, and it was obvious that something was going on. 456 00:27:31,600 --> 00:27:35,000 Speaker 1: So this was more for fun than an intent to 457 00:27:35,040 --> 00:27:37,760 Speaker 1: create a new tool for musicians. But it turned out 458 00:27:38,080 --> 00:27:41,240 Speaker 1: that's exactly what autotune was really destined for, to become 459 00:27:41,280 --> 00:27:45,440 Speaker 1: a tool for a process called pitch quantization. But again, 460 00:27:45,680 --> 00:27:48,200 Speaker 1: that wasn't what Hildebrand set out to do. In fact, 461 00:27:48,200 --> 00:27:51,479 Speaker 1: ecquate to that Pitchfork article I mentioned earlier, the idea 462 00:27:51,560 --> 00:27:54,440 Speaker 1: here was to aim for perfection, at least in terms 463 00:27:54,520 --> 00:27:57,520 Speaker 1: of being in the right key and on pitch. That 464 00:27:57,720 --> 00:28:01,520 Speaker 1: imperfections would somehow interfere within a momotional connection to the music, 465 00:28:01,920 --> 00:28:04,199 Speaker 1: and you want that music to be perfect so that 466 00:28:04,280 --> 00:28:07,879 Speaker 1: you can have that emotional impact. Now, personally, I disagree 467 00:28:07,920 --> 00:28:11,320 Speaker 1: with that take. Some of my favorite recordings are with 468 00:28:11,520 --> 00:28:15,760 Speaker 1: artists who have imperfect voices. They weren't screeching or catterwalling. 469 00:28:15,800 --> 00:28:18,280 Speaker 1: It wasn't like it was unpleasant to listen to them, 470 00:28:18,480 --> 00:28:21,399 Speaker 1: but they aren't pitch perfect either, and to me, that 471 00:28:21,480 --> 00:28:24,520 Speaker 1: adds a lot of character and emotion. So as an example, 472 00:28:24,960 --> 00:28:28,399 Speaker 1: Warren Zevon, who you know did the song where Wolves 473 00:28:28,440 --> 00:28:31,119 Speaker 1: of London and tons of other stuff. I mean, prolific 474 00:28:31,240 --> 00:28:34,560 Speaker 1: musician who tragically passed away several years ago. He has 475 00:28:34,600 --> 00:28:37,280 Speaker 1: a great cover of the song back in the High 476 00:28:37,280 --> 00:28:40,720 Speaker 1: Life Again and which is a pretty cheesy song, but 477 00:28:41,080 --> 00:28:44,040 Speaker 1: Warren Zevon's cover is really emotional. It's great, and it's 478 00:28:44,080 --> 00:28:47,760 Speaker 1: a little bit raw, and to me it resonates far 479 00:28:47,840 --> 00:28:51,320 Speaker 1: more than a note perfect performance would have. But I 480 00:28:51,320 --> 00:28:53,320 Speaker 1: do understand where hill le Brand and his team were 481 00:28:53,360 --> 00:28:55,160 Speaker 1: coming from. You know, if you if you have a 482 00:28:55,240 --> 00:28:58,560 Speaker 1: take from a recording session that is almost but not 483 00:28:58,800 --> 00:29:01,600 Speaker 1: quite right, maybe there was a transition where the wrong 484 00:29:01,640 --> 00:29:03,480 Speaker 1: note came out, or you know, just a moment where 485 00:29:03,520 --> 00:29:06,160 Speaker 1: it took an artist a little bit longer to slide 486 00:29:06,320 --> 00:29:09,200 Speaker 1: to find the right pitch. A tool that could smooth 487 00:29:09,200 --> 00:29:14,000 Speaker 1: things out a little while not remaining you know, noticeable, 488 00:29:14,280 --> 00:29:17,120 Speaker 1: while you know, slipping under the radar. That could prevent 489 00:29:17,200 --> 00:29:20,760 Speaker 1: listeners from being distracted by something that was unintentional. But 490 00:29:20,800 --> 00:29:23,160 Speaker 1: what if you took that tool that was meant to 491 00:29:23,240 --> 00:29:27,680 Speaker 1: fix errors and used it to create unintended effects. That's 492 00:29:27,680 --> 00:29:29,800 Speaker 1: what we're going to talk about when we come back 493 00:29:30,000 --> 00:29:42,640 Speaker 1: from this quick break. So we talked about how auto 494 00:29:42,720 --> 00:29:48,080 Speaker 1: tune was meant to fix little imperfections in music recordings 495 00:29:48,080 --> 00:29:51,600 Speaker 1: and live performance. But as I mentioned, if you had 496 00:29:51,640 --> 00:29:55,720 Speaker 1: that setting set to zero so that it would instantaneously 497 00:29:55,760 --> 00:30:00,440 Speaker 1: attempt to correct pitches, then you could create an almost 498 00:30:00,760 --> 00:30:05,360 Speaker 1: robotic vocalization. So instead of shying away from the artificial 499 00:30:05,440 --> 00:30:07,400 Speaker 1: sounds that could come out if you were to use 500 00:30:07,400 --> 00:30:10,920 Speaker 1: it improperly, you leaned into it. That's what happened in 501 00:30:11,000 --> 00:30:14,600 Speaker 1: nineteen ninety eight with Sher's song Believe, saying that dolls 502 00:30:14,720 --> 00:30:17,280 Speaker 1: zero would create the robotic like effect, which in this 503 00:30:17,360 --> 00:30:19,560 Speaker 1: case was the goal in the first place, and that 504 00:30:19,720 --> 00:30:24,200 Speaker 1: song was a smash success. I couldn't stand it, and 505 00:30:24,280 --> 00:30:26,400 Speaker 1: it was everywhere. I couldn't stand it, not because of 506 00:30:26,440 --> 00:30:29,120 Speaker 1: the auto tune. I just didn't vibe with the song. 507 00:30:29,600 --> 00:30:35,040 Speaker 1: No no shade on chare phenomenal artist, you know, incredibly talented, 508 00:30:35,560 --> 00:30:39,720 Speaker 1: Just that song didn't jibe with me and The interesting 509 00:30:39,760 --> 00:30:42,800 Speaker 1: thing was that this huge success not only pulled the 510 00:30:42,840 --> 00:30:44,880 Speaker 1: curtain back on a tool that was meant to correct 511 00:30:44,880 --> 00:30:49,040 Speaker 1: little mistakes and thus create a whole conversation around whether 512 00:30:49,120 --> 00:30:52,040 Speaker 1: or not artists were quote unquote cheating by using it, 513 00:30:52,400 --> 00:30:54,760 Speaker 1: but it launched a whole new way to create music 514 00:30:54,760 --> 00:30:57,400 Speaker 1: in the first place. I personally think the artist who 515 00:30:57,480 --> 00:31:02,000 Speaker 1: is most associated with auto twoun is one who adopted 516 00:31:02,040 --> 00:31:04,880 Speaker 1: the technology and made it an intrinsic part of his brand. 517 00:31:05,280 --> 00:31:07,600 Speaker 1: That would be te Pain. He came to the party 518 00:31:07,640 --> 00:31:10,320 Speaker 1: a little bit late. He became interested in autotune around 519 00:31:10,320 --> 00:31:13,760 Speaker 1: two thousand and four, and he wasn't looking for something 520 00:31:13,760 --> 00:31:17,000 Speaker 1: to help compensate for his singing ability, because he actually 521 00:31:17,000 --> 00:31:19,920 Speaker 1: sings very well. But he liked the thought of the 522 00:31:19,920 --> 00:31:23,719 Speaker 1: technology that would set him apart from other artists, and 523 00:31:23,760 --> 00:31:27,360 Speaker 1: he could forge a vocal identity using this tool to 524 00:31:27,400 --> 00:31:30,000 Speaker 1: create a sound that no one else was really embracing 525 00:31:30,040 --> 00:31:34,920 Speaker 1: at that point. So he jumped wholeheartedly into autotune, and 526 00:31:34,960 --> 00:31:38,840 Speaker 1: he made liberal use of the technology and achieved tremendous 527 00:31:38,880 --> 00:31:43,360 Speaker 1: success along the way, selling like Platinum records by using 528 00:31:43,400 --> 00:31:46,600 Speaker 1: this technology. His love of the software led to an 529 00:31:46,600 --> 00:31:50,360 Speaker 1: official partnership with Hildebrand's company for a few years, and 530 00:31:50,440 --> 00:31:53,640 Speaker 1: Tarees licensed the technology to tee Pain for an app 531 00:31:53,720 --> 00:31:56,600 Speaker 1: called I Am te Pain, which you could use to 532 00:31:56,960 --> 00:32:01,560 Speaker 1: do autotuned right there on your smartphone three dollars initially, 533 00:32:01,720 --> 00:32:04,360 Speaker 1: and it was downloaded by millions of users that generated 534 00:32:04,680 --> 00:32:07,040 Speaker 1: quite a lot of revenue just on its own. Now, 535 00:32:07,080 --> 00:32:11,480 Speaker 1: eventually t Pain and Antarees parted ways, and t Pain 536 00:32:11,640 --> 00:32:15,640 Speaker 1: ultimately partnered with a different pitch correction company called Isotope. 537 00:32:15,880 --> 00:32:20,360 Speaker 1: The t Pain story also led to a lawsuit against Antarees, 538 00:32:20,560 --> 00:32:24,480 Speaker 1: and Antaries filed a countersuit against te Pain, and ultimately 539 00:32:24,520 --> 00:32:27,240 Speaker 1: the whole thing was settled out of court and everyone 540 00:32:27,360 --> 00:32:31,240 Speaker 1: signed an NDA. So I have no details about, you know, 541 00:32:31,320 --> 00:32:34,880 Speaker 1: how that shook out in the end, but it was 542 00:32:34,920 --> 00:32:37,040 Speaker 1: one of those things where it was kind of a 543 00:32:37,160 --> 00:32:43,400 Speaker 1: smudge on the Antarees name at least it was awkward, right. 544 00:32:43,600 --> 00:32:47,440 Speaker 1: But a much larger threat to Hildebrand's company was Apple. 545 00:32:47,840 --> 00:32:53,000 Speaker 1: Apple had purchased a German company called Emagic. Emagic also 546 00:32:53,480 --> 00:32:56,840 Speaker 1: had a pitch correction tool. In fact, it was a 547 00:32:56,840 --> 00:33:02,880 Speaker 1: pitch correction tool that, according to Hildebrand, essentially copied autotune technology. 548 00:33:03,160 --> 00:33:07,600 Speaker 1: This was possible because Antaris had failed to protect its 549 00:33:07,760 --> 00:33:13,960 Speaker 1: German patent properly, and so Emagic was able to appropriate 550 00:33:14,200 --> 00:33:18,840 Speaker 1: that technology or copy that technology without fear of legal recourse. 551 00:33:19,440 --> 00:33:23,760 Speaker 1: So then Apple acquires Emagic, which means Apple is then 552 00:33:23,800 --> 00:33:28,280 Speaker 1: able to incorporate Emagic's technology into their own products, including 553 00:33:28,480 --> 00:33:32,880 Speaker 1: their own sound editing software. And this meant that autotune 554 00:33:32,920 --> 00:33:39,320 Speaker 1: effectively got incorporated into Apple software without having to license 555 00:33:39,360 --> 00:33:43,560 Speaker 1: the technology from Antaries, because again they got it by 556 00:33:43,560 --> 00:33:47,360 Speaker 1: acquiring this German company. Now, Antaris could technically have still 557 00:33:47,440 --> 00:33:50,160 Speaker 1: sued Apple. There's no guarantee that they would have won, 558 00:33:50,560 --> 00:33:54,240 Speaker 1: but they could have sued them. However, Hildebrand explained that 559 00:33:54,320 --> 00:33:59,920 Speaker 1: they didn't really have that option because Apple has enorm 560 00:34:00,080 --> 00:34:05,400 Speaker 1: mislead deep pockets. Apple is just an incredibly rich company, 561 00:34:05,840 --> 00:34:10,760 Speaker 1: and Apple could easily just outweight Antaries in the legal system, 562 00:34:10,760 --> 00:34:14,840 Speaker 1: while Antaries would drain its resources trying to sue Apple. 563 00:34:15,120 --> 00:34:17,000 Speaker 1: So even if Antaris was in the right of it, 564 00:34:17,200 --> 00:34:20,120 Speaker 1: even if they would have won a judgment against Apple, 565 00:34:20,440 --> 00:34:23,000 Speaker 1: the chance was that Antari's would go out of business 566 00:34:23,080 --> 00:34:25,360 Speaker 1: just trying to pay for all the legal fees for 567 00:34:25,560 --> 00:34:29,719 Speaker 1: the whole battle in the first place. So Ultimately, Antari's 568 00:34:29,760 --> 00:34:32,000 Speaker 1: didn't go after Apple. It just it would have been 569 00:34:32,000 --> 00:34:37,080 Speaker 1: a death sentence. Culturally, autotune began to face resistance in 570 00:34:37,120 --> 00:34:42,239 Speaker 1: the late two thousands. Some artists expressed disdain for the technology, 571 00:34:42,280 --> 00:34:45,160 Speaker 1: going so far as to say it ruined Western music. 572 00:34:45,640 --> 00:34:49,840 Speaker 1: This was partly due to an oversaturation problem. The success 573 00:34:49,840 --> 00:34:53,200 Speaker 1: of Tea Pain, as well as the earlier instances of autotune, 574 00:34:53,600 --> 00:34:58,080 Speaker 1: inspired countless others to embrace the technology while not necessarily 575 00:34:58,120 --> 00:35:01,560 Speaker 1: doing very much else to differentiate themselves from other artists. 576 00:35:01,640 --> 00:35:03,799 Speaker 1: In other words, they were kind of leaning on it 577 00:35:03,840 --> 00:35:06,080 Speaker 1: as a crutch or a gimmick. So there was a 578 00:35:06,160 --> 00:35:09,600 Speaker 1: glut of auto tune robotic voiced vocals and music in 579 00:35:09,640 --> 00:35:11,920 Speaker 1: the early to mid two thousands, and by the late 580 00:35:11,960 --> 00:35:14,600 Speaker 1: two thousands some folks were absolutely fed up with this 581 00:35:14,719 --> 00:35:17,360 Speaker 1: and there was a backlash. It actually kind of reminds 582 00:35:17,360 --> 00:35:20,719 Speaker 1: me about how people began to turn against disco in 583 00:35:20,800 --> 00:35:23,640 Speaker 1: the nineteen seventies, and that in some ways the punk 584 00:35:23,719 --> 00:35:28,400 Speaker 1: rock movement was partly a reaction to disco or a 585 00:35:28,480 --> 00:35:31,799 Speaker 1: rejection of disco. I would only say partly because punk 586 00:35:31,880 --> 00:35:34,000 Speaker 1: rock also has its roots in glam rock, and I 587 00:35:34,000 --> 00:35:37,400 Speaker 1: think glam rock also kind of helped inspire disco. So 588 00:35:37,640 --> 00:35:40,759 Speaker 1: it's a complicated set of relationships, as you might say 589 00:35:40,840 --> 00:35:44,400 Speaker 1: on Facebook. But bands like Death Kem for Cuti actually 590 00:35:44,600 --> 00:35:48,440 Speaker 1: actively spoke out against autotune. So again, some artists were 591 00:35:48,560 --> 00:35:52,759 Speaker 1: arguing that autotune was being used by people to compensate 592 00:35:52,800 --> 00:35:55,400 Speaker 1: for a lack of ability, So they're kind of casting 593 00:35:55,520 --> 00:35:57,960 Speaker 1: shade on fellow artists saying, well, yeah, they have to 594 00:35:58,040 --> 00:36:00,959 Speaker 1: use autotune because they can't sing, or others would say 595 00:36:00,960 --> 00:36:04,360 Speaker 1: like it was making music less genuine and sincere, like 596 00:36:04,440 --> 00:36:08,560 Speaker 1: less human because it was going through this digital processing process. 597 00:36:08,960 --> 00:36:12,920 Speaker 1: Jay Z famously released a song titled Death of Autotune 598 00:36:12,920 --> 00:36:14,839 Speaker 1: in two thousand and nine, the same year when our 599 00:36:14,960 --> 00:36:18,239 Speaker 1: original Tech Stuff episode about autotune came out. As you 600 00:36:18,320 --> 00:36:21,640 Speaker 1: might imagine, jay Z's song had some pretty strong opinions 601 00:36:21,719 --> 00:36:25,279 Speaker 1: about the technology inside of it. It resonated enough to 602 00:36:25,320 --> 00:36:28,120 Speaker 1: win him a Grammy for it, so other people agreed. 603 00:36:28,440 --> 00:36:31,759 Speaker 1: But despite all that backlash, autotune continues to be a thing. 604 00:36:32,040 --> 00:36:35,760 Speaker 1: It did not, in fact die. It's been incorporated into 605 00:36:36,040 --> 00:36:40,680 Speaker 1: software and digital audio workstations. It and similar pitch manipulation 606 00:36:40,760 --> 00:36:44,640 Speaker 1: technologies are often found in everything from professional audio engineering 607 00:36:44,680 --> 00:36:48,279 Speaker 1: software suites to free programs that you can download online. So, 608 00:36:48,360 --> 00:36:52,759 Speaker 1: for example, I sometimes use a program called Audacity, and 609 00:36:53,200 --> 00:36:56,120 Speaker 1: Audacity has an option under its effects where I can 610 00:36:56,280 --> 00:36:59,760 Speaker 1: manually adjust the pitch of a recorded piece of audio. 611 00:37:00,080 --> 00:37:02,480 Speaker 1: I can set what the pitch should be. Now that's 612 00:37:02,480 --> 00:37:06,160 Speaker 1: not autotune, right, because by definition I'm not using an 613 00:37:06,239 --> 00:37:10,520 Speaker 1: auto feature. I'm manually changing the pitch, but it's using 614 00:37:10,600 --> 00:37:14,000 Speaker 1: similar approaches to get an effect. I've actually even made 615 00:37:14,040 --> 00:37:16,200 Speaker 1: use of that tool while I was editing my friend 616 00:37:16,239 --> 00:37:19,680 Speaker 1: Shay's podcast, Kadi Womple with the Shadow People. Shae does 617 00:37:19,760 --> 00:37:22,680 Speaker 1: nearly all the voices on that show. I've actually voiced 618 00:37:22,680 --> 00:37:25,960 Speaker 1: two characters on that show. So if you're eager to 619 00:37:26,000 --> 00:37:30,640 Speaker 1: hear other output from me, that's not a technology podcast, 620 00:37:30,760 --> 00:37:32,920 Speaker 1: go listen to Kadi Womple with the Shadow People. I 621 00:37:33,000 --> 00:37:35,480 Speaker 1: voice a couple of characters on that, but I edit 622 00:37:35,560 --> 00:37:38,839 Speaker 1: the show, so I use pitch adjustment tools in order 623 00:37:38,880 --> 00:37:42,360 Speaker 1: to make some of Shay's voices sound like different people. 624 00:37:42,640 --> 00:37:45,440 Speaker 1: So it's still herb doing the voice, but I digitally 625 00:37:45,560 --> 00:37:50,800 Speaker 1: manipulate the voice to give certain characters their own distinct sound. 626 00:37:51,239 --> 00:37:53,960 Speaker 1: It's pretty neat stuff. I have no idea what it 627 00:37:53,960 --> 00:37:56,800 Speaker 1: would sound like if I actually used an auto tuned tool. 628 00:37:57,200 --> 00:38:00,279 Speaker 1: That probably would sound very different, But I have a 629 00:38:00,360 --> 00:38:04,719 Speaker 1: lot of fun playing with these pitch manipulation tools. Now, 630 00:38:04,760 --> 00:38:09,000 Speaker 1: to get more into the cultural and social impact of autotune, 631 00:38:09,200 --> 00:38:13,520 Speaker 1: I highly recommend that article in Pitchfork by Simon Reynolds. Again, 632 00:38:13,560 --> 00:38:17,680 Speaker 1: that's titled how Autotune Revolutionized the sound of Popular Music. 633 00:38:17,840 --> 00:38:20,680 Speaker 1: It's a long form article, it's well worth your time 634 00:38:20,719 --> 00:38:23,360 Speaker 1: to read it. As a Zachary Crockett's article that I 635 00:38:23,400 --> 00:38:26,680 Speaker 1: mentioned earlier, both of those are great articles about autotune 636 00:38:26,719 --> 00:38:30,200 Speaker 1: and not just the technology, but it's impact on music 637 00:38:30,239 --> 00:38:33,960 Speaker 1: in general and society and culture as well. And Reynolds 638 00:38:33,960 --> 00:38:36,480 Speaker 1: goes into much deeper detail about how the technology has 639 00:38:36,520 --> 00:38:39,200 Speaker 1: had an impact on the recording industry and the backlash 640 00:38:39,239 --> 00:38:41,960 Speaker 1: that came out as a result of that, as well 641 00:38:41,960 --> 00:38:45,799 Speaker 1: as sort of a counter movement against autotune. So check 642 00:38:45,840 --> 00:38:48,759 Speaker 1: those out. They are well worth your time. And I 643 00:38:48,800 --> 00:38:51,719 Speaker 1: could go on, but really I feel like those articles 644 00:38:52,000 --> 00:38:55,480 Speaker 1: do a much better job than I would of describing 645 00:38:55,520 --> 00:38:57,400 Speaker 1: all of that, So check those out when you have 646 00:38:57,480 --> 00:39:00,239 Speaker 1: some time. That's it for today. I hope all of 647 00:39:00,239 --> 00:39:02,799 Speaker 1: you out there are doing well, and I will talk 648 00:39:02,800 --> 00:39:13,560 Speaker 1: to you again really soon. Tech Stuff is an iHeartRadio production. 649 00:39:13,840 --> 00:39:18,879 Speaker 1: For more podcasts from iHeartRadio, visit the iHeartRadio app, Apple Podcasts, 650 00:39:19,000 --> 00:39:21,000 Speaker 1: or wherever you listen to your favorite shows.