WEBVTT - The Spotify Story 0:00:04.120 --> 0:00:07.160 Get in touch with technology with tech Stuff from how 0:00:07.200 --> 0:00:13.960 stuff Works dot com. Hey there, and welcome to tech Stuff. 0:00:14.000 --> 0:00:17.520 I am your host, Jonathan Strickland. I'm an executive producer 0:00:17.560 --> 0:00:19.439 with How Stuff Works and I heart radio and I 0:00:19.480 --> 0:00:23.079 love all things tech, and today I'm going to cover 0:00:23.320 --> 0:00:27.040 the story of Spotify, really the origin of Spotify, and 0:00:27.080 --> 0:00:29.480 then a bit about what it's been up to recently. 0:00:29.880 --> 0:00:32.440 And just for the interests of full disclosure, I feel 0:00:32.440 --> 0:00:34.680 I should point out that the company I work for, 0:00:34.840 --> 0:00:38.040 I Heart Media, is a competitor to Spotify. But don't worry. 0:00:38.159 --> 0:00:40.080 I'm going to cover this topic just like I would 0:00:40.200 --> 0:00:43.800 any other, as objective as possible, and maybe with some 0:00:43.880 --> 0:00:46.360 bad jokes and puns. And I think this will also 0:00:46.440 --> 0:00:49.680 help lead into a discussion about the way consumer media 0:00:49.800 --> 0:00:52.680 has changed significantly over the last couple of decades. That 0:00:52.760 --> 0:00:55.720 might actually be a second episode that I'll focus on 0:00:55.720 --> 0:00:59.280 a little bit later. The story of Spotify is tied 0:00:59.440 --> 0:01:03.120 very close with its founders, uh mainly Daniel Eck but 0:01:03.160 --> 0:01:06.760 also Martin Lawrenson. Those are two Swedish businessmen who set 0:01:06.760 --> 0:01:09.600 out to change how people access music, and while the 0:01:09.640 --> 0:01:12.480 service would launch in two thousand eight, the real story 0:01:12.520 --> 0:01:16.800 begins much earlier with Daniel Eck and initially in two 0:01:16.840 --> 0:01:19.360 thousand two, but we're going to go even earlier than that. See, 0:01:19.400 --> 0:01:24.440 Daniel K was an enterprising kid, like super enterprising, and 0:01:24.520 --> 0:01:27.040 an interview and the event that was called a Pando 0:01:27.160 --> 0:01:29.560 Monthly in New York City back in two thousand twelve, 0:01:29.600 --> 0:01:32.080 he talked about how as a young teen and I'm 0:01:32.120 --> 0:01:35.600 talking about thirteen years old, fourteen years old, he started 0:01:35.600 --> 0:01:39.800 making money by creating websites for local businesses in Sweden. 0:01:39.880 --> 0:01:42.720 He was interested in coding and he was getting into 0:01:42.760 --> 0:01:46.479 it and he started getting hired to make websites for 0:01:46.480 --> 0:01:49.600 for companies. And we're not talking about small amount of money. 0:01:49.640 --> 0:01:51.440 I mean his first job was about a hundred dollars, 0:01:51.440 --> 0:01:53.120 and then he went up to about two hundred dollars 0:01:53.120 --> 0:01:55.640 and it started going up from there. And according to K, 0:01:55.680 --> 0:01:58.440 at one point he was pulling down about fifty grand 0:01:58.720 --> 0:02:03.160 a month doing business. So imagine that being in high 0:02:03.160 --> 0:02:06.960 school and your side gig is pulling you down fifty 0:02:07.080 --> 0:02:12.080 thousand dollars a month. Yikes. That was going pretty well 0:02:12.680 --> 0:02:15.959 until the Swedish tax Authority notified X that he owed 0:02:16.040 --> 0:02:18.600 the government a substantial amount of money to the tune 0:02:18.600 --> 0:02:21.800 of a couple hundred grand. X early fortune was pretty 0:02:21.840 --> 0:02:24.400 much wiped out, and he learned a valuable lesson about 0:02:24.480 --> 0:02:26.880 running a business, which is that you can't do it 0:02:26.960 --> 0:02:30.280 on the q T forever. Around two thousand two, X 0:02:30.320 --> 0:02:33.720 started thinking about the problem with digital music and piracy, 0:02:34.040 --> 0:02:38.200 and it was definitely a big problem. Services like Napster, 0:02:38.480 --> 0:02:42.720 new Tela, and kaza had been home to huge databases 0:02:42.840 --> 0:02:47.239 of pirated music since the late nineteen nineties, or had 0:02:47.320 --> 0:02:51.959 facilitated music piracy since the ninet Since not all of 0:02:52.000 --> 0:02:54.000 them were really databases, they were more like peer to 0:02:54.000 --> 0:02:56.360 peer networks. More on that in a little bit. So 0:02:56.440 --> 0:03:00.000 governments around the world were responding by creating some pretty 0:03:00.160 --> 0:03:04.680 draconian laws in an effort to curb online piracy. The 0:03:04.720 --> 0:03:08.320 governments were largely responding to big media companies that were 0:03:08.360 --> 0:03:11.440 putting a lot of pressure on elected officials to fix 0:03:11.520 --> 0:03:14.400 this problem. Everyone could see that the Internet was going 0:03:14.440 --> 0:03:17.520 to be an important component in all sorts of industries. 0:03:17.880 --> 0:03:20.320 Apple had already staked its own claim with iTunes in 0:03:20.360 --> 0:03:23.480 two thousand one, but there was a real concern that 0:03:23.600 --> 0:03:28.040 piracy essentially translated into lost sales. Now to that point, 0:03:28.639 --> 0:03:31.960 I want to acknowledge that it's not as simple as that. 0:03:32.520 --> 0:03:35.520 If I walk into an old fashioned record store and 0:03:35.600 --> 0:03:38.920 I shoplift an album, we're talking like old vinyl albums, 0:03:38.960 --> 0:03:43.480 and I grab a copy of uh, Phil Collins Greatest 0:03:43.520 --> 0:03:46.640 Hits because I'm real bad, and I slip it under 0:03:46.680 --> 0:03:49.480 my my jacket and I and I step out the door. 0:03:49.840 --> 0:03:52.720 That's a clear loss for the shop owner of that 0:03:52.840 --> 0:03:57.160 music shop, right, because I've taken one physical album. The 0:03:57.200 --> 0:03:59.760 shop owner has a limited number of physical copies that 0:03:59.800 --> 0:04:03.360 they have ordered from the record label, so they've spent 0:04:03.400 --> 0:04:05.960 a certain amount of money to get a certain number 0:04:06.080 --> 0:04:08.680 of copies of each title. I just took one of those. 0:04:08.880 --> 0:04:11.600 They can no longer sell that particular album, so they 0:04:11.640 --> 0:04:14.440 are out the cost they incurred when they ordered it. 0:04:14.840 --> 0:04:17.960 That's pretty clear, right, you can understand that's a direct loss. 0:04:18.360 --> 0:04:21.960 But with digital media it's different. Maybe I never intended 0:04:22.000 --> 0:04:24.880 to buy a certain song. Maybe I kind of want 0:04:24.960 --> 0:04:27.960 that song, but I don't want it badly enough to 0:04:28.000 --> 0:04:31.360 actually spend my money on it. So I decide I'm 0:04:31.400 --> 0:04:35.240 just gonna pirate a copy. The original song, however, still exists. 0:04:35.480 --> 0:04:37.839 The original copy of the song still exists. I didn't 0:04:37.880 --> 0:04:41.840 take anything physical. Nothing is stopping a vendor from selling 0:04:41.880 --> 0:04:44.640 just as many copies as they would have even if 0:04:44.640 --> 0:04:48.119 I hadn't taken anything. So for that reason, you can't 0:04:48.240 --> 0:04:52.840 really argue that stolen's songs directly translate into lost revenue 0:04:53.080 --> 0:04:56.240 in the digital market, because it could be that the 0:04:56.279 --> 0:05:00.560 people who downloaded songs illegally otherwise would have just never 0:05:00.720 --> 0:05:03.719 bought the music at all, so they may have just 0:05:03.880 --> 0:05:06.359 gone without. It's not a loss sale because it was 0:05:06.400 --> 0:05:09.160 never going to be a sale in the first place. Now, 0:05:09.240 --> 0:05:11.440 I don't want this to be one sided. While the 0:05:11.520 --> 0:05:15.760 people downloading music for free, we're circumventing the legal route, 0:05:16.040 --> 0:05:19.400 and they were effectively stealing music, whether you could argue 0:05:19.440 --> 0:05:22.719 that it was causing direct harm or not. The music 0:05:22.760 --> 0:05:27.680 industry wasn't exactly making a good case for purchasing music legally. 0:05:28.240 --> 0:05:32.000 The music landscape was fractured, with some record labels signing 0:05:32.040 --> 0:05:35.120 with certain online stores, other record labels would sign with 0:05:35.240 --> 0:05:39.840 other stores. Sometimes it was very piecemeal, where certain albums 0:05:39.880 --> 0:05:43.359 would be allowed and others wouldn't. And then there was 0:05:43.400 --> 0:05:46.840 the matter of digital rights management or d r m uh. 0:05:47.240 --> 0:05:52.640 This is a way of protecting digital property. Many companies 0:05:52.640 --> 0:05:56.400 have been putting restrictive DRM on music files. It's meant 0:05:56.440 --> 0:05:59.520 to prevent you from being able to make an infinite 0:05:59.560 --> 0:06:02.480 number of copies of a digital file, because obviously that's 0:06:03.120 --> 0:06:06.880 a downside to the digital format, at least if you 0:06:06.920 --> 0:06:09.960 are in the business of selling stuff, Right, if the 0:06:10.000 --> 0:06:14.360 thing you create it can be replicated infinitely by anyone, 0:06:14.640 --> 0:06:16.840 then all you have to do is sell one of those, 0:06:16.880 --> 0:06:19.919 and then immediately you never sell another copy because people 0:06:20.080 --> 0:06:23.919 just start copying them and then distributing them everywhere. So 0:06:24.040 --> 0:06:26.800 DRM was meant to be a counter to that. You're 0:06:26.839 --> 0:06:31.480 to prevent people from being able to copy indiscriminately in 0:06:31.520 --> 0:06:33.880 some way or another. But one of the nasty side 0:06:33.880 --> 0:06:37.440 effects of DRM is that it can impact the experience 0:06:37.480 --> 0:06:41.000 of legitimate customers. Right, So the purpose of DRM is 0:06:41.040 --> 0:06:43.800 to protect digital media so that people can't just steal it. 0:06:44.040 --> 0:06:46.719 But when that same DRM makes it harder to listen 0:06:46.800 --> 0:06:50.880 to legitimately purchased music, the people who are actually buying 0:06:51.240 --> 0:06:55.640 the stuff feel as though they've been punished for buying it. Right. 0:06:55.680 --> 0:06:58.559 If you if you bought a song and it turns 0:06:58.560 --> 0:07:00.760 out you can't listen to the song the way you 0:07:00.800 --> 0:07:03.280 want to, despite the fact that you've spent money on it. 0:07:03.839 --> 0:07:07.320 You're not, you don't have an incentive to keep doing that, right, 0:07:07.360 --> 0:07:11.000 You're You're You've got a bad experience. Meanwhile, pirates would 0:07:11.040 --> 0:07:14.040 spend some time and effort in order to strip files 0:07:14.400 --> 0:07:18.240 from the DRM, or rather DRM from the files, so 0:07:18.320 --> 0:07:22.040 that the versions that they would upload or allow to 0:07:22.200 --> 0:07:26.520 be downloaded from their machines would be DRM free. So 0:07:26.640 --> 0:07:28.800 you would actually have a better experience listening to the 0:07:28.920 --> 0:07:32.679 pirated music than if you were to purchase the legal stuff. 0:07:33.040 --> 0:07:36.840 So as a consumer, you had two choices. Essentially, you 0:07:36.840 --> 0:07:39.760 could jump through some hoops and deal with a crappy 0:07:39.880 --> 0:07:43.600 listening experience with official music that you pay for and 0:07:43.640 --> 0:07:45.480 it's got DRM and everything, so it's got all this 0:07:45.520 --> 0:07:48.600 baggage attached to it, or you could pop on over 0:07:48.800 --> 0:07:52.200 to a piracy site or service and then download whatever 0:07:52.240 --> 0:07:55.040 you want for free. Now, there always was a risk 0:07:55.440 --> 0:07:58.440 that you could also end up downloading some malware on 0:07:58.480 --> 0:08:00.360 top of the music files or whatever where it was 0:08:00.440 --> 0:08:03.080 you were after, but for many it seemed like it 0:08:03.120 --> 0:08:06.760 was worth the risk. So the whole thing was kind 0:08:06.800 --> 0:08:09.400 of a mess, and X thought there had to be 0:08:09.520 --> 0:08:12.480 a better way. There had to be some method to 0:08:12.600 --> 0:08:16.560 making music available digitally. That would remove the hurdles that 0:08:16.600 --> 0:08:18.760 the public would have to jump over to get to 0:08:18.800 --> 0:08:21.800 what they wanted. And if he could crack that, if 0:08:21.800 --> 0:08:24.160 he could build a real business out of it, then 0:08:24.200 --> 0:08:26.680 he could stand to really make some money. And he 0:08:26.760 --> 0:08:29.520 was thinking, there's gotta be a way to make the 0:08:29.520 --> 0:08:36.920 experience seamless enough where the legal route is preferable to piracy. 0:08:36.960 --> 0:08:39.200 That's the secret. If you can make it just as 0:08:39.240 --> 0:08:42.320 easy or easier to buy music as it would be 0:08:42.400 --> 0:08:45.880 to steal it, then more people are going to buy it. 0:08:46.400 --> 0:08:49.199 And the reasoning was that a lot of people want 0:08:49.240 --> 0:08:52.679 to actually spend money supporting the stuff they love. They 0:08:52.720 --> 0:08:56.440 just if you make it frustrating, they'll go around that, 0:08:56.840 --> 0:08:59.600 because who wants to deal with that kind of nonsense. Now, 0:08:59.640 --> 0:09:01.680 this was back in two thousand two, but would still 0:09:01.679 --> 0:09:04.160 be a few years before he would develop the idea further. 0:09:04.720 --> 0:09:07.079 In the meantime, he worked for a couple of companies, 0:09:07.120 --> 0:09:10.400 and he founded some other companies, including an online advertising 0:09:10.400 --> 0:09:13.720 firm called Advertigo, which he would later sell for a 0:09:13.720 --> 0:09:16.600 big profit. In fact, he ended up founding and selling 0:09:16.720 --> 0:09:20.640 four companies in a relatively short time. This was right 0:09:20.679 --> 0:09:25.040 after the wake of eBay purchasing Skype. It kind of 0:09:25.080 --> 0:09:28.720 opened up the mergers and acquisitions market in Europe at 0:09:28.760 --> 0:09:32.720 that time. And by then he was a millionaire, and 0:09:32.960 --> 0:09:36.560 he was considering retiring at the ripe old age of 0:09:36.679 --> 0:09:41.480 twenty three years old, but instead he thought about that 0:09:41.559 --> 0:09:48.480 music challenge again, also millionaire by twenty three. Way to go, dude, man, 0:09:49.000 --> 0:09:51.960 that that's some hustle right there. So Eck found a 0:09:52.040 --> 0:09:54.800 partner in Martin Lawrence n He was actually the guy 0:09:54.800 --> 0:09:58.320 who had acquired ad Vertigo from Daniel Eck, and that 0:09:58.480 --> 0:10:02.080 was ex advertising plat Form. The two began to brainstorm 0:10:02.160 --> 0:10:05.640 about making a legal music service that removed some of 0:10:05.679 --> 0:10:08.440 the barriers that other official stores had kind of been 0:10:08.480 --> 0:10:11.800 putting up. The reason that those barriers encouraged people to 0:10:11.880 --> 0:10:14.640 resort to piracy, and the music industry as a whole 0:10:14.840 --> 0:10:17.720 was hurting as and so something had to be done. 0:10:18.000 --> 0:10:19.880 But if you just made it easier to get that 0:10:19.960 --> 0:10:22.400 music you wanted legally, they figured people would do it, 0:10:22.400 --> 0:10:25.880 just like I was saying earlier. So they created an 0:10:25.920 --> 0:10:28.480 office out of a three bedroom apartment that was above 0:10:28.480 --> 0:10:31.480 a coffee shop, and they began to work on their design. 0:10:31.600 --> 0:10:34.160 Their plan was to leverage peer to peer networks to 0:10:34.280 --> 0:10:38.160 deliver streaming music. So what exactly is a peer to 0:10:38.160 --> 0:10:40.400 peer network and how does it work. Some of you 0:10:40.520 --> 0:10:43.880 might be familiar with peer to peer. It's something that 0:10:43.960 --> 0:10:46.760 was really big in the nineties. It's still quite big today. 0:10:47.120 --> 0:10:49.000 It's just not talked about as much. A lot of 0:10:49.200 --> 0:10:52.400 the conversation is shifted to cloud computing, but peer to 0:10:52.440 --> 0:10:55.200 peer is still a very important methodology. So here's a 0:10:55.240 --> 0:10:58.880 quick rundown. First, You've got a network of computers all 0:10:58.920 --> 0:11:01.760 right there. They're connected, perhaps over the Internet, but it's 0:11:01.800 --> 0:11:04.480 a it's a network within the Internet, it's not the 0:11:04.640 --> 0:11:08.000 entire Internet. So each of those computers could be owned 0:11:08.000 --> 0:11:11.520 by a regular person like yourself, and each computer in 0:11:11.559 --> 0:11:14.360 that network is what we would call a peer. Each 0:11:14.400 --> 0:11:17.360 computer on the network is running special software, which is 0:11:17.400 --> 0:11:21.120 what makes them peers. It connects them, it creates this 0:11:21.200 --> 0:11:24.760 network among all the peers on that particular connection, and 0:11:24.840 --> 0:11:28.160 peers can share files with one another without having to 0:11:28.160 --> 0:11:31.560 go through an intermediary server. So if I'm on the 0:11:31.600 --> 0:11:34.040 network and you're on the network, we could share files 0:11:34.080 --> 0:11:37.800 directly with one another. And typically the way this would 0:11:37.840 --> 0:11:40.160 work is that you would designate a special folder on 0:11:40.240 --> 0:11:43.440 your computer's hard drive to act as a share folder, 0:11:43.760 --> 0:11:46.240 and anything you would put in that share folder could 0:11:46.240 --> 0:11:49.280 be seen and copied by peers on the network, and 0:11:49.320 --> 0:11:51.920 the more peers who have copies of a file, the 0:11:52.000 --> 0:11:54.840 faster downloads go. As your computer can pull from the 0:11:54.880 --> 0:11:57.600 best connection available across the network, you can get bits 0:11:57.600 --> 0:12:01.160 and pieces from different computers and just download a file 0:12:01.280 --> 0:12:04.480 much faster than if it were dependent upon a single source. 0:12:04.840 --> 0:12:08.120 So the more computers on a peer to peer network 0:12:08.120 --> 0:12:11.640 that are hosting a file, the faster those transfers will go. Now, 0:12:11.720 --> 0:12:15.079 there's nothing illegal about this method of file transfer. It's 0:12:15.120 --> 0:12:18.880 a great way to disseminate large files across a network. 0:12:19.160 --> 0:12:21.840 It's efficient, and it removes the need for more hefty 0:12:21.880 --> 0:12:25.160 network servers. But it also allowed for people to share 0:12:25.200 --> 0:12:29.920 stuff illegally at a pretty alarming rate, at least alarming 0:12:29.960 --> 0:12:33.079 to media companies, and so selling the idea of a 0:12:33.320 --> 0:12:36.080 legal peer to peer network was a bit of an 0:12:36.160 --> 0:12:40.120 uphill battle. However, the actual implementation of a peer to 0:12:40.160 --> 0:12:43.840 peer network using Spotify service was a lot different from 0:12:43.880 --> 0:12:47.360 the napsters and kazas of the piracy era. The plan 0:12:47.640 --> 0:12:50.680 wasn't to allow users to build up an enormous library 0:12:50.720 --> 0:12:54.199 of songs in their personal computers, which they then could 0:12:54.400 --> 0:12:58.079 help distribute across an ever widening network and thus sap 0:12:58.240 --> 0:13:01.200 the music industry of its revenue. Instead, the plan was 0:13:01.240 --> 0:13:04.000 to use peer to peer relationships to create a good 0:13:04.080 --> 0:13:08.080 listening experience for users to sidestep another problem. That problem 0:13:08.160 --> 0:13:11.080 was if you wanted to create a centralized music on 0:13:11.200 --> 0:13:15.120 demand service, you'd be beholden to the Internet connections between 0:13:15.160 --> 0:13:19.920 your server, your giant music database, and all of your users. 0:13:20.480 --> 0:13:24.280 Now this wasn't as huge a problem in Sweden, which 0:13:24.360 --> 0:13:26.800 was a leading and still is a leading country when 0:13:26.800 --> 0:13:29.880 it comes to rolling out high speed broadband internet, but 0:13:30.200 --> 0:13:33.920 it would make it difficult to scale Spotify globally. So 0:13:34.000 --> 0:13:35.839 let's say that you did go with this method, where 0:13:35.840 --> 0:13:38.160 you went with a centralized approach. You have your big 0:13:38.240 --> 0:13:41.720 music database that's sitting in a data center somewhere on 0:13:41.840 --> 0:13:45.200 your property. If a connection were bad between a user 0:13:45.600 --> 0:13:48.640 and this database, the user might have to wait for 0:13:48.679 --> 0:13:50.920 a long time for a song to buffer far enough 0:13:50.920 --> 0:13:54.720 along to start playing. That's irritating, or worse, the song's 0:13:54.800 --> 0:13:57.480 quality might suffer in the middle of the song, or 0:13:57.480 --> 0:14:00.640 it might stop playing. That you know it might midplay 0:14:00.679 --> 0:14:03.400 through while the service tried to get the data from 0:14:03.480 --> 0:14:07.480 the server side to the user's device and restart the song. 0:14:07.880 --> 0:14:10.760 The peer to peer approach could help that. Now explain 0:14:10.800 --> 0:14:13.160 how it works in just a moment, But let's get 0:14:13.160 --> 0:14:15.760 back to the business side for a second. They met 0:14:15.800 --> 0:14:18.920 with a lot of resistance from record labels because they 0:14:18.920 --> 0:14:21.800 were going to have to get licensing agreements from those 0:14:21.800 --> 0:14:24.960 record labels in order to do this. But their arguments 0:14:25.040 --> 0:14:28.640 were persuasive. It took a long time, but they were 0:14:28.680 --> 0:14:32.120 making some pretty sound arguments. They laid out the challenges 0:14:32.160 --> 0:14:34.720 that I mentioned earlier with the older model of music. 0:14:34.960 --> 0:14:37.600 They stayed in their case in terms of business plans, 0:14:38.000 --> 0:14:41.280 and they said, are the licensing fees and royalties would 0:14:41.280 --> 0:14:44.720 provide steady revenue to the record labels and presumably too 0:14:44.800 --> 0:14:47.240 artists more than that in a second, But it was 0:14:47.440 --> 0:14:49.680 challenging to get the record labels to agree. It was 0:14:49.720 --> 0:14:53.720 also challenging to get investors on board because without the 0:14:53.760 --> 0:14:57.640 backing of the music industry, there was no business for Spotify. 0:14:57.440 --> 0:15:01.640 They had nothing to provide anybody, and many investors saw 0:15:01.680 --> 0:15:04.320 the idea as a bad bet, So Lawrenson would end 0:15:04.400 --> 0:15:07.320 up funding operations for a couple of years, just out 0:15:07.320 --> 0:15:10.880 of out of pocket. Now, gradually record labels began to 0:15:10.880 --> 0:15:13.880 come around, and they sort of had to because the 0:15:13.960 --> 0:15:18.000 music industry revenues had peaked in nineteen ninety nine at 0:15:18.000 --> 0:15:21.360 twenty seven billion dollars. That's a lot of cheddar, as 0:15:21.400 --> 0:15:26.320 Ben Bolan would say, but that was the peak. It 0:15:26.400 --> 0:15:30.960 went to decline after that, sales started flagging. This wasn't 0:15:31.120 --> 0:15:33.560 just due to piracy, although to hear the music industry, 0:15:33.640 --> 0:15:36.840 they would argue vehemently that piracy was what took the 0:15:36.920 --> 0:15:40.680 largest chunk out of their, uh, their revenue. I would 0:15:40.720 --> 0:15:43.200 say that there were a lot of different factors. But 0:15:43.560 --> 0:15:47.520 by a decade later, in two thousand and eight, the 0:15:47.520 --> 0:15:52.280 global industry was down to fourteen billion dollars. That's still 0:15:52.320 --> 0:15:54.560 a lot of money, but that's like half of what 0:15:54.640 --> 0:15:58.680 it was in nine and so it took Spotify about 0:15:58.680 --> 0:16:02.520 two years from conception to launch. It was thought of 0:16:02.560 --> 0:16:05.520 in two thousand two, it was seriously started in two 0:16:05.560 --> 0:16:08.560 thousand six, and it was launched in two thousand eight, 0:16:08.600 --> 0:16:11.600 in October two thousand eight, uh and launched in just 0:16:11.680 --> 0:16:14.560 a few European countries at the time. It did get 0:16:14.600 --> 0:16:18.640 the backing of the four major music labels. That would 0:16:18.680 --> 0:16:22.600 be Warner Music, Sony, E and My Music and Universal. Now, 0:16:22.600 --> 0:16:24.640 when we come back, I'll talk more about how the 0:16:24.680 --> 0:16:27.480 service works and how it's evolved since two thousand and eight. 0:16:27.560 --> 0:16:37.720 But first let's take a quick break. Okay, so let's 0:16:37.760 --> 0:16:42.520 talk about the Spotify service. To access Spotify, you first 0:16:42.520 --> 0:16:45.880 need to download the client software for whatever platforms you 0:16:45.960 --> 0:16:49.640 plan to use. Now, originally this was limited to desktop computers, 0:16:49.960 --> 0:16:53.240 so your first step was downloading some software to your 0:16:53.320 --> 0:16:57.119 machine and installing it. This would act as your interface, 0:16:57.200 --> 0:16:59.560 allowing you to search for songs on the service, but 0:16:59.680 --> 0:17:03.000 it also served as the peer to peer software to 0:17:03.080 --> 0:17:06.960 make your machine part of the overall sharing network. This 0:17:07.080 --> 0:17:10.720 client can pull music stored on your computer into your 0:17:10.760 --> 0:17:14.560 Spotify library, assuming the music you have isn't stored in 0:17:14.600 --> 0:17:17.960 a protected M four P file. That's actually a little 0:17:17.960 --> 0:17:20.640 bit redundant because in four P stands were IMPEG four 0:17:20.680 --> 0:17:24.560 protected audio. Uh. This was a file extension that Apple's 0:17:24.680 --> 0:17:28.919 fair Play digital rights management system used. The format restricted 0:17:28.920 --> 0:17:32.159 the devices that could play the associated audio file and 0:17:32.280 --> 0:17:36.359 limited to authorized devices only, So if an device was 0:17:36.400 --> 0:17:38.879 authorized to play the file, it could otherwise it could not. 0:17:39.840 --> 0:17:44.520 Spotify just to sidestep any problems, said we can't handle 0:17:44.600 --> 0:17:47.280 any of those types of files. However, if you did 0:17:47.359 --> 0:17:50.640 happen to have songs in that format, Spotify would look 0:17:50.680 --> 0:17:53.480 to see if it had a license for those tracks 0:17:53.520 --> 0:17:56.960 on its service. So if you happen to have a 0:17:57.200 --> 0:18:00.399 track of an artist that was on Spotify, and that 0:18:00.520 --> 0:18:04.040 song was on Spotify, it would match the track with 0:18:04.119 --> 0:18:06.159 what was in your library and just go ahead and 0:18:06.160 --> 0:18:07.840 put that in your playlist. It's just you wouldn't be 0:18:07.880 --> 0:18:11.000 listening to the version that was on your computer. You 0:18:11.000 --> 0:18:13.480 would actually be listening to one that was on the 0:18:13.520 --> 0:18:16.960 Spotify service itself, and it would be a a non 0:18:17.320 --> 0:18:21.000 uh M four P DRM version. All other songs, whether 0:18:21.000 --> 0:18:24.960 in Spotify's licensing agreements or not, would appear in your library, 0:18:25.040 --> 0:18:29.080 assuming you have DRM free versions on your computer. So 0:18:29.320 --> 0:18:32.399 even if you had a whole mess of songs in 0:18:32.520 --> 0:18:36.400 your computer's memory, if it was in its hard drive 0:18:37.000 --> 0:18:39.560 that were not on Spotify. Let's say that you had 0:18:39.720 --> 0:18:43.000 all the Beatles albums, and Beatles was one of those 0:18:43.160 --> 0:18:46.600 groups that wasn't being carried by Spotify, it would show 0:18:46.680 --> 0:18:48.600 up in your library because you had them on your 0:18:48.600 --> 0:18:51.119 hard drive, but no one else would be able to 0:18:51.840 --> 0:18:55.200 listen to those using Spotify, you know, pulling it over 0:18:55.240 --> 0:18:59.720 the network. The format Spotify uses when streaming music is 0:18:59.800 --> 0:19:04.040 the Vorbis method of audio compression. Vorbis is an open 0:19:04.119 --> 0:19:08.119 source software project, and it uses the file extension dot 0:19:08.160 --> 0:19:12.480 o g g or og for compressed audio files. So 0:19:12.560 --> 0:19:16.080 you often hear people talk about as Aga Vorbis, which 0:19:16.080 --> 0:19:19.680 sounds like a character from Game of Thrones. Now, actually 0:19:20.040 --> 0:19:22.919 it does trace its name to a fantasy series, but 0:19:23.119 --> 0:19:25.639 one that's a little more lighthearted than George R. R. 0:19:25.720 --> 0:19:29.400 Martin's epic Saga because the name Vorbius comes from Terry 0:19:29.400 --> 0:19:34.960 Pratchett's fantasy comedy series Discworld, a fantastic comedy fantasy series. 0:19:34.960 --> 0:19:36.800 If you've never read any of the Discworld books, I 0:19:36.840 --> 0:19:40.800 highly recommend checking them out. But this name was specifically 0:19:40.880 --> 0:19:43.919 from the book Small Gods. Aug is also a name 0:19:43.960 --> 0:19:46.199 that shows up in Terry Pratchett's works, but that was 0:19:46.240 --> 0:19:50.040 just a coincidence. The uh the Age extension was not 0:19:50.280 --> 0:19:54.840 named after the Terry Pratchett character uh Nanny Ogg in 0:19:54.880 --> 0:19:59.080 the Discworld books. I've talked about audio compression before, but 0:19:59.240 --> 0:20:02.679 here's a quick reminder. The purpose for compression is that 0:20:02.800 --> 0:20:05.639 audio files tend to be pretty large, and the larger 0:20:05.680 --> 0:20:08.160 the file size, the more time it takes to send 0:20:08.160 --> 0:20:11.320 over a network, and that time is experienced by the 0:20:11.440 --> 0:20:15.080 user as a lag between hitting the play button and 0:20:15.119 --> 0:20:17.960 hearing any sound, or in the form of having a 0:20:18.000 --> 0:20:20.679 song stop part way through while the data transfer continues 0:20:20.720 --> 0:20:24.240 in the background, or having the quality of the sound 0:20:24.480 --> 0:20:27.960 drop in the middle of play through. So compression means 0:20:28.040 --> 0:20:30.480 you take these files and you make the file size 0:20:30.520 --> 0:20:33.680 is smaller and thus easier to transfer across a network. 0:20:33.920 --> 0:20:37.760 And there are two broadways to do this. There's lossless 0:20:37.800 --> 0:20:41.879 and lossy. Now, lossless compression means that you take a file, 0:20:42.280 --> 0:20:45.119 you squish it down, you compress it, and when you 0:20:45.400 --> 0:20:48.160 un squash it when you when you expand it back, 0:20:48.560 --> 0:20:52.159 you have the bit for bit identical file as the 0:20:52.240 --> 0:20:56.439 pre compressed version. So you have song file A, you 0:20:56.520 --> 0:21:00.320 compress it into compressed format B, and when you take 0:21:00.359 --> 0:21:04.320 it back out of compression is back to song file A. Again, 0:21:05.240 --> 0:21:08.000 this is not easy to do, and typically you only 0:21:08.040 --> 0:21:11.520 reduce the file size by a relatively small amount. Some 0:21:11.600 --> 0:21:14.560 files can only be compressed down maybe five cent, which 0:21:14.560 --> 0:21:16.480 does not save you a whole lot of space, and 0:21:16.560 --> 0:21:18.399 you might be able to get down as much as 0:21:18.440 --> 0:21:22.600 fifty with certain types of lossless compression. But if you're 0:21:22.640 --> 0:21:27.200 really looking at a high quality audio file and it's 0:21:27.200 --> 0:21:32.520 a large audio file, even the reduction is still still 0:21:32.560 --> 0:21:34.640 means you're dealing with a big file. It still might 0:21:34.680 --> 0:21:38.199 be too big for you to easily stream. Lossy compression 0:21:38.320 --> 0:21:41.960 can squish files down much smaller than lossless, but they 0:21:41.960 --> 0:21:48.040 do so at drumroll please a loss. Lossy compression ditches 0:21:48.400 --> 0:21:51.040 some of the data in the audio file during the 0:21:51.040 --> 0:21:56.600 compression process. Now, ideally, the information the compression format ditches 0:21:56.960 --> 0:22:00.119 represents stuff that you wouldn't be able to notice anyway, 0:22:00.240 --> 0:22:04.200 because human hearing can only pick up a range of frequencies, 0:22:04.240 --> 0:22:07.360 and typically we describe it as the average human can 0:22:07.400 --> 0:22:11.440 hear between twenty hurts to twenty kill a hurts. Anything 0:22:11.480 --> 0:22:15.280 that is above or below that range is outside what 0:22:15.359 --> 0:22:19.120 we could directly perceive. So if the audio file has 0:22:19.200 --> 0:22:23.440 any information about sounds beyond human hearing, then it might 0:22:23.480 --> 0:22:26.639 just sweep that stuff under the digital rug before compressing 0:22:26.680 --> 0:22:29.040 the rest. It might say, well, no one could hear 0:22:29.080 --> 0:22:31.280 this anyway, so there's no point in keeping it, and 0:22:31.320 --> 0:22:34.560 I'm just gonna not include that information. There are a 0:22:34.560 --> 0:22:37.920 lot of different rules for compression and different codex to 0:22:37.960 --> 0:22:41.679 follow different philosophies, but they're all based on the psychoacoustic 0:22:41.760 --> 0:22:45.160 model of how we perceive sound. So if it determines 0:22:45.200 --> 0:22:47.639 that we would not have perceived that sound, it can 0:22:47.680 --> 0:22:50.120 go out out the door at the end of the day. 0:22:50.560 --> 0:22:53.520