Stream of Consciousness
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
Title |
| |
Subtitle |
| |
Title of Series | ||
Number of Parts | 234 | |
Author | ||
License | CC Attribution - ShareAlike 3.0 Germany: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor and the work or content is shared also in adapted form only under the conditions of this | |
Identifiers | 10.5446/33157 (DOI) | |
Publisher | ||
Release Date | ||
Language |
Content Metadata
Subject Area | ||
Genre | ||
Abstract |
|
00:00
Streaming mediaPerspective (visual)State of matterProduct (business)Musical ensembleMereologyProjective planeData conversionLevel (video gaming)Computer animationJSONXMLUMLMeeting/Interview
01:13
Musical ensembleService (economics)Basis <Mathematik>Video gameHand fanData miningGoodness of fitMereologyHypermediaData conversionConnected spaceLecture/Conference
02:12
Musical ensembleObservational studyInequality (mathematics)AlgorithmState of matterDifferent (Kate Ryan album)Streaming mediaProduct (business)Web serviceMeeting/InterviewLecture/Conference
03:24
Web serviceStreaming mediaService (economics)InternetworkingYouTubeMusical ensembleService (economics)Shared memoryPay televisionSurfaceOffice suiteOperator (mathematics)Web servicePerspective (visual)BuildingMathematicsPoint (geometry)System identificationBitBit rateArchaeological field surveyAdditionDistanceDigitizingWeb portalComputer animationXML
05:29
AlgorithmMusical ensembleMultiplication signHeegaard splittingVideo gamePerspective (visual)MassSlide ruleTwitterWeb serviceMoving averageLecture/Conference
07:15
InternetworkingSocial softwareBlogWeb serviceProfil (magazine)AlgorithmHypermediaTwitterElement (mathematics)Bit rateQuantificationLecture/ConferenceMeeting/Interview
08:10
Profil (magazine)Multiplication signUser profileMusical ensembleAlgorithmElement (mathematics)Library (computing)Source codePort scannerLecture/Conference
09:17
InternetworkingSocial softwareUser profileBlogAlgorithmAlgorithmMathematicsCategory of beingNumberConstructor (object-oriented programming)Key (cryptography)Port scannerMatrix (mathematics)Musical ensembleFacebookHypermediaInternetworkingBlogMoment (mathematics)WordDescriptive statisticsInterface (computing)TwitterHost Identity ProtocolComputer animationLecture/Conference
10:46
Computer programDemosceneFreewareBuildingMatrix (mathematics)MassMusical ensembleStandard deviationModemProduct (business)Hasse diagramBeer steinString (computer science)Philips CD-iInterior (topology)CurvatureElement (mathematics)BitMusical ensembleNetwork topologyVirtual machinePattern languageGeometryLink (knot theory)Moment (mathematics)WritingNumerical taxonomyAlgorithmCalculationMeeting/InterviewXML
12:37
InformationVirtual machineMusical ensembleAlgorithmBitLecture/ConferenceMeeting/Interview
14:01
Wage labourMusical ensembleAnalogyCASE <Informatik>Observational studyHypermediaPerspective (visual)Disk read-and-write headFiber (mathematics)SoftwareArmDistribution (mathematics)Wave packetTerm (mathematics)Frame problemStreaming mediaShape (magazine)Lecture/ConferenceMeeting/Interview
15:30
MathematicsArmMusical ensembleLengthRight angleDiameterBoundary value problemMultiplication signCondition numberShape (magazine)2 (number)NeuroinformatikEuler anglesLimit (category theory)Reading (process)TrailMoving averagePlastikkarteLecture/ConferenceMeeting/Interview
17:07
Business model2 (number)Streaming mediaArithmetic meanNumberMultiplication signProjective planeHand fanNeuroinformatikTrailSoftwareEuler anglesMusical ensembleMeeting/InterviewComputer animation
18:26
2 (number)TrailMetropolitan area networkExtension (kinesiology)AnalogyResultantComputer animationMeeting/InterviewLecture/Conference
19:25
2 (number)Arithmetic meanMultiplication signMusical ensembleMeeting/Interview
20:19
Ideal (ethics)Data structureComputing platformArithmetic meanMusical ensembleMathematicsOrder (biology)Sound effectMultiplication sign2 (number)Term (mathematics)QuicksortContent (media)BitBusiness modelMedianResonatorFile formatGravitationMeeting/InterviewComputer animation
21:47
Film editingMoment (mathematics)WeightComputing platformConnected spaceWordMusical ensembleGravitationTerm (mathematics)Endliche ModelltheorieResonatorContent (media)Business modelService (economics)Meeting/Interview
22:39
Hand fanLoop (music)Multiplication signMusical ensembleShared memoryPoint (geometry)Self-organizationTerm (mathematics)Graph coloringTotal S.A.Object (grammar)PhysicalismAdditionComputing platformEndliche ModelltheorieLecture/ConferenceMeeting/Interview
23:50
Endliche ModelltheorieDisk read-and-write headBeta functionStreaming mediaRight anglePay televisionSound effectPlanningPhysical systemPattern languageDifferent (Kate Ryan album)Multiplication signBitChemical equationControl flowLecture/ConferenceMeeting/Interview
24:53
Phase transitionStreaming mediaFreewarePhase transitionLibrary catalogBitHand fanGravitationWell-formed formulaComputer filePoint (geometry)Mobile appMeeting/Interview
25:55
ResultantMultiplication signHand fanRegular graphGravitationMathematicsBusiness modelComputing platformYouTubeBitStreaming media3 (number)Musical ensembleConnected spaceContent (media)Endliche ModelltheorieDatabase transactionMeeting/Interview
27:07
Phase transitionStreaming mediaFreewareBit rateStreaming mediaInformationOrder (biology)Computing platformSlide ruleHand fanMultiplication signMusical ensemblePerspective (visual)Moving averagePoint (geometry)Perfect groupWordMeeting/InterviewLecture/Conference
28:11
Shape (magazine)Musical ensemblePhysical systemWordAlgorithmForm (programming)Different (Kate Ryan album)Multiplication signMathematicsVirtual machineBitNumberTime zoneNetwork topologyLecture/ConferenceMeeting/Interview
30:21
Musical ensembleOpen sourceWorkstation <Musikinstrument>Computing platformPerspective (visual)Sound effectPoint (geometry)Classical physicsLecture/ConferenceMeeting/Interview
31:08
Musical ensembleAdditionData storage deviceRow (database)Axiom of choicePoint (geometry)NumberAlgorithmPersonal digital assistantDirection (geometry)Lecture/Conference
33:24
GravitationData structurePoint (geometry)AlgorithmMusical ensembleDrum memoryPerspective (visual)MathematicsGeneric programmingElement (mathematics)Lecture/ConferenceMeeting/Interview
34:35
Block (periodic table)Musical ensembleData profilingBusiness modelPoint (geometry)Streaming mediaPay televisionElectronic mailing listOpen sourceBlogTrailInteractive televisionResonatorOSI modelAlgorithmLecture/ConferenceMeeting/Interview
36:25
Perspective (visual)Perfect groupRhombusIdeal (ethics)Condition numberLink (knot theory)Musical ensembleTerm (mathematics)Mobile appComputing platformAlgorithmEndliche ModelltheorieOpen setLecture/ConferenceMeeting/Interview
38:31
Data structureShape (magazine)BitOpen setHand fanConnected spaceEqualiser (mathematics)Endliche ModelltheorieMereologyVideo gameTelecommunicationLocal ringComputing platformSoftwareAnalogyArithmetic meanNumberMeeting/Interview
39:55
MereologyLatent heatCache (computing)Local ringMusical ensembleProfil (magazine)Block (periodic table)SoftwareTouchscreenConnected spaceShared memoryArithmetic meanBasis <Mathematik>Hand fanTerm (mathematics)Computing platformAuthorizationScaling (geometry)Service (economics)Functional (mathematics)Message passingTrailTwitterInternetworkingStatisticsMobile appBlogMilitary baseLecture/ConferenceMeeting/Interview
42:38
Information privacyInformationMultiplication signMusical ensembleTerm (mathematics)AreaComputing platformLecture/Conference
43:50
Pattern languageAlgorithmMusical ensembleData structureShape (magazine)PlanningProcess (computing)MassTunisDecision theoryCASE <Informatik>View (database)Point (geometry)Bridging (networking)Sinc functionMathematicsTerm (mathematics)Right angleAnalogyHypermediaLevel (video gaming)Lecture/ConferenceMeeting/Interview
46:30
GravitationData structureShape (magazine)BitTrailGoogolMusical ensembleBusiness modelMereologyLecture/ConferenceMeeting/Interview
47:18
Position operatorDecision theoryTelecommunicationDrop (liquid)Computing platformLibrary catalogBusiness modelMIDIMusical ensembleMathematicsHand fanMobile appVotingLecture/ConferenceMeeting/Interview
48:50
Computing platformProcess (computing)Streaming mediaSpeech synthesisLecture/ConferenceMeeting/Interview
49:35
Multiplication signRow (database)Metropolitan area networkBitPerfect groupComputing platformShared memoryNumberMeeting/InterviewLecture/Conference
50:33
Endliche ModelltheorieTerm (mathematics)Right angleForestProcess (computing)Different (Kate Ryan album)MereologyComa BerenicesWeightDistribution (mathematics)CalculationMultiplication signPiLecture/ConferenceMeeting/Interview
51:26
Data structureGravitationCalculationMultiplication signEstimatorNumberConservation lawWordMereologyService (economics)Musical ensembleFunction (mathematics)Design by contractGoogolHand fanTrailEqualiser (mathematics)Lecture/ConferenceMeeting/Interview
53:02
Data structurePopulation densityEndliche ModelltheorieService (economics)Business modelResonatorMultiplication signWordTrailStreaming mediaCellular automatonPhysical systemStatement (computer science)Roundness (object)Insertion lossSign (mathematics)Meeting/InterviewLecture/Conference
54:52
GravitationMusical ensembleStatement (computer science)Roundness (object)AlgorithmService (economics)Stress (mechanics)Fisher's exact testPlotterAnalytic setFitness functionAriana TVMeeting/InterviewLecture/Conference
57:06
Computer animation
Transcript: English(auto-generated)
00:16
My name is actually Nielle Heise. I am the host of this session, but let's not talk about this.
00:23
We have other things to discuss today. Yeah, welcome to this panel. We're glad to be here and to talk about music streaming with you, hopefully. So before I start with a short introduction, let me explain the Opus Modus Operandi, that is how we call it. We will
00:43
have a short introduction, as I said, and then I will be joined by these three wonderful speakers here on stage and they will give a short explanation of their perspectives on music streaming and their projects, which are very interesting. And one part of that would be, for example, Ariana, Peter
01:01
and Andreas. And after that, or after those short lightning talks, we will hopefully have like 30 minutes for discussion and we want to invite you to join the conversation. So after the talks are finished, please grab a mic if you have questions or comments and
01:22
yeah, just join our discussion. Okay, the question is why I am here. Nielle Heise, University of Hamburg. I'm a media researcher and naturally I'm interested in what people do with music or other stuff online. So this is part of my research,
01:41
is looking at podcasts. This has nothing to do with streaming actually. Maybe the connection is with Andreas, who's a good friend of mine. But I'm consumer of music and I'm a big fan of streaming services. And to give you an impression how this affects my life on an everyday basis, these are three artists, Thundercat, FK Twix and WolfTek,
02:08
who I've discovered via the discover weekly feature on Spotify. And I not only listen to their music regularly, but I also go to the concerts, buy tickets, buy CDs from them.
02:22
So these are just three examples from musicians that came to me. We magically wire the algorithm of Spotify. And as a study which has been published three or four weeks ago,
02:40
research showed that the adoption of streaming leads to very large increases in the quantity and diversity of music consumption. So there are some hints in the research that shows that yeah, streaming music streaming really changes the way people discover music and also the way they are exposed to different kinds of music.
03:06
Maybe to give you some idea of the state of the stream. As you can see on the left side, this is an overview over the streaming service production revenue in Germany. And as you can see, we had a massive increase in the last two or three years.
03:25
And as you can see, considering all the music sales in 2016, the subscription revenue makes about a quarter of those music sales. So we can say this has really been established in the last years. And when we look at the global market
03:42
share of music streaming services, it's no surprise that Spotify dominates the arena. But Apple Music, yeah, okay, they are about 20%. And then you have Deezer, but all the other services are not so relevant players. But these are the
04:01
most relevant services we are talking about today. Now German perspective, the talk is in English, but we are in Berlin. Germans are a bit slow when it comes to the adoption of digital services. But this is a snapshot from last year's IDT online survey. And as you can see, as you can see here,
04:27
music streaming services aren't adopted that much, but consider really larger adoption rates among younger users. So apparently maybe there's a change now coming up in the next years.
04:43
Other relevant services are music portals. Again, younger users adopt these services a bit more than older users. And what I find really interesting that 89% of young onliners are using music via YouTube, which I found strange, but I'm not so young. So maybe
05:06
this is a behavior that younger people are more prone to. And obviously also music identification services are relevant, a relevant point of access to music.
05:20
Okay, this was just a short overview of what's going on in the music streaming arena. And now we will begin with our first speaker, Ariana, explaining her perspective as a journalist and musician on music streaming services. Thank you very much. So welcome everybody first.
05:45
Nice that you're here. So as a musician in my private life and as a journalist, music journalist in my business life, I'm kind of a split personality. And during my time at the Rolling Stone, I focused on newcomers. And when you write about a newcomer, what you have to
06:04
provide to the reader is that you have to describe what the music sounds like so that the reader knows what this new artist is about. And I have always found it quite tricky and difficult to label music in a way that an algorithm would do. Because I felt like, okay, you have
06:24
different criteria, you have your knowledge and experience about music, and then you have to describe whether something sounds more like post-rock or more like shoegaze. And that brought me to the algorithm of Spotify and Discover Weekly, which you may all know, it's a service
06:44
that is provided to the heavy users of Spotify every Monday. And it's a playlist, a personalized playlist with a playtime of two hours and about 30 songs. And you can see on Twitter, we have slides here and... Oh, I can do it myself. Nice.
07:16
So these are just two tweets that can be an example, how people on social media go crazy
07:22
about Discover Weekly. They even want to marry the playlist because they find it so suitable to their taste. And I really wanted to know how Spotify is doing this, how can they succeed to such a good striking rate, strike rate. And so the algorithm that Spotify uses for
07:45
Discover Weekly is called Echo Nest. Yeah, this is the algorithm behind it. And so Echo Nest works with three elements. The first one is the taste profile. So you have to be aware that
08:01
Spotify collects every click that you have ever done on the service. So Spotify knows about every song that you have ever listened to at any time, even the Nickelback songs and the songs that you probably don't want to know people that you're listening to. So you have a
08:21
profile and this is collected by the songs that you click to and your libraries and your playlists, of course. And so you have one taste profile and Spotify compares your taste profile to all the other users' profiles. And so, of course, there are bubbles of other users who
08:43
have a similar taste like you have. And if, for example, you have a playlist with five songs and other users who have a similar taste like you have also have those five songs and playlists together with a sixth song that you don't know by now, then it's highly likely that
09:01
Discover Weekly will suggest this sixth song to you as well. The second element on which the algorithm is built on is an audio scan. So this is what interested me the most because it's music is deconstructed into minor key, major key, into speed, into tempo change, into
09:26
instruments and all this stuff. And what you get after these deconstructions is just a bunch of numbers and matrices. It takes a week for Spotify to do the next Discover Weekly because it's so many data. And this is about how sound is being deconstructed into numbers.
09:50
And the third, well, pillar or the third category on which the algorithm is built on is the text scan. So what Econes does is that it scans all texts on the internet, social media,
10:05
Facebook, Twitter. If somebody is writing about a new favorite band on a Facebook post, in a Facebook post, then the text scan can scan the words that the user is writing and using for this description. For example, they can scan the words, listen to this new band,
10:26
it's my favorite song at the moment, and then they can take the words favorite and amazing and listen to it and stuff like that and can tell what is going viral at the moment, for example. Or if a music magazine or a blog writes about a new band and describes how the music sounds
10:45
like, for example, it sounds like, I don't know, old school hip hop from the beginning of the 80s, then text scan can tell that obviously this band sounds a bit like old school music because a human being is labeled like this. And with these three elements, Spotify can provide this
11:08
amazing and very impressive Discover Weekly playlist that is personalized to your taste. This is just a quick look on the Echo Tree. It's a bit insane if you ask me, but also very
11:28
impressive because this is the genre taxonomy of Spotify, of Echo Nest. I don't know if you can read. Yeah, you can see there are 1,600 genres at the moment, I guess.
11:45
You can see how the machine put music and sound, something that is very intuitive and emotional, into labels. And there are names like CCM or B64 and
12:01
names for genres that I have never heard of. And you can also see interesting links. So what I found out through this genre tree is that deep Taiwanese pop is very similar to German folk's music. And contemporary country has a lot to do with music that is labeled as fusball.
12:26
So it's very interesting what the machine calculates. Yeah, so as amazing and sophisticated this algorithm may be, of course, what I want to put
12:42
into the discussion is whether this kind of music makes the listener lazy. Because if you have this personalized, tailored and customized playlist for your needs, then everything that you have to do is just click play and listen to what a machine
13:01
shows for you. And there are critics, for example, Ben Retle from the New York Times, who even wrote a book about it, who say even if I like half of the songs in my playlist, always feels very flat. And I feel infantinalized to a child who's been giving something
13:25
in a very passive way. And I can say the same about me. Sometimes I wonder how a machine can reveal what I like in this way. But there is something about it that still feels a bit wrong.
13:42
So yeah, maybe you have your own experiences with this playlist as well. Yeah, so these are the critics that I wrote down. I'm sure there are much more. And yeah, we'll be very happy to discuss about this with you. Yes, thanks Arianna. Next up, Andreas, you need to clicker,
14:03
I guess. Okay, Andreas has a PhD in media studies. Yeah, that's not Andreas, okay. And he's also head of Analexol, or one of the heads of Analexol, and will tell us something about how they experimented on how streaming
14:26
may change the music itself. Thank you. Yeah, we are actually a small indie label from Leipzig and the label is run by artists who are on the label. So our perspective is not so much like
14:40
a label perspective in terms of we have money and promote music. So we're musicians without any money and promote ourselves. So just to have this framing of my perspective, so when I say label then sorry guys, if any one of you makes music you don't have any much money. But what we wanted to find out was whether streaming as a distribution of music has the potential to shape
15:05
music itself. Because when we discuss streaming, from an artist perspective we mostly speak about the revenue. So because streaming gives you very little revenue out of it and this is like the most burning topic for an artist when talking about streaming. But we wanted to put it a step away
15:24
from this discussion and asking ourselves and asking our artistical network and our musicians if they want to figure out how streaming might change the music. And actually this has quite history because a lot of the things we know about music and ever since music was recorded
15:43
or distributed, this technology and also those economical circumstances shape the music. Like far on the left are those wax rolls from Edison and how the size of them limited the time a music or a track could run or the diameter of a vinyl. It limits or it
16:10
interesting albums are shaped by this length. For example things by Red Hot Chili Peppers, I just read on Wikipedia this morning for example. And also what we should not forget is
16:21
the radio. Like a proper single is about 3 minutes 30 seconds and it has a ramp where the DJ could speak on and things like this. And interestingly those things shape music as we know it. Like the length of the single, the length of the EP, the length of the album. And those those are very surprisingly stable, those conditions. Although the way we consume music
16:44
changed so much. So and a very smart person from New York, a singer-songwriter and also professor Mike Errico, he just pointed out the question so what is the what is the boundary of streaming since it's not a time boundary, right? You can upload whatever length you want.
17:02
You can upload albums with thousand songs. But the the hard boundary, the hard limit of streaming is the revenue. Like the business model and the revenue is like 0.2 or 0.8 cents per stream. And the interesting thing and that's what he pointed out, you get paid after 30 seconds playing
17:20
time. And so he just and he figured out with his number and and he asked the question is a longer than 30 seconds, right? Shouldn't we sell a single as as six songs? Having a 3 minute 30 song and then make six portions out of 30 seconds out of it. And this question is
17:42
very clever and there have been artists already who made use of this. On the left is Wolfpack. Nilo already introduced them as one act she discovered on Spotify. They made this great Sleepify project where they had an album with 12 to 10 tracks and all of them were silent and
18:02
you could run this album on repeat whole day or whole night on your computer and they earned the revenue out of it and they asked the fans to do so. So this was Sleepify and this is a project from UK punk rockers and they made one album with 130 second tracks. And we really like this punk attitude when thinking about how how can we figure out how streaming might change
18:24
the music. And then we asked our network to compose songs or tracks which are not longer or which are exactly 31 seconds like one second longer as you need. So um and yeah you can listen to the you can listen to the result on analog sol de e slash 31 seconds or 31s.
18:43
Um I also put two songs but it doesn't work here.
19:19
The PA is not so bad actually. Um let's give it a give it a second try. So this was one example
19:25
and I I took two to like two extreme examples. This was one. I'll give you a second.
20:05
So those are two examples um and you you can figure out on the Spotify players and it's 31 songs. Um so but how does that sound? What is music worth um that is 31 seconds long like the time you get paid as an artist on streaming platform like Spotify for it. And I mean you can figure
20:25
out it's not not enough to establish a musical idea in a song. That's what our artists said. It's a it's a nice thing to play with but it doesn't feel like a song to them and it is not from the structure as you could hear. Um what happens is you get things like skits or tools
20:41
or intros and it almost feels like this pre-listening stuff you got on Amazon and iTunes for example. And and that's the so that's the format makes music pretty in that sort in that terms like industrial. Um and yeah that was our big result of out of it that like the value of music doesn't stop after those 31 seconds after you get paid. It needs
21:02
obviously more time. This is something I want to bring in in our discussion. Um yeah what what effects streaming has on music and how we might change business models in order to not have negative effects on the music itself. Thank you. Yes thanks Andreas. Our last of the three speakers is Peter and Peter is one of the founders of
21:27
Resonate which is an alternative streaming music cooperative. And yeah we will talk a bit about changes in artist payment and participation of users but um Peter go for yourself.
21:42
Thank you. So the question here is with endlessly available content where's the gravity in music streaming connecting us with the artists? And this is something I discovered using the platform I built was I felt a new weight to the experience of music streaming and I'll get
22:00
to what that's about in just a moment. So first of all why another streaming service? And being Republica I get to kind of cut to the chase and say the reason is that we have a corporate governance model for a culture with the existing streaming services. So the corporate
22:20
business model detect dictates what the um what the platform does and how the content is consumed. So with Resonate we're going to do the opposite and that's to create a cooperative governance model for culture. And so what that means in the realist terms is that everyone
22:44
that joins is a shareholder as a cooperative. I'm the founder and I have one share and as a label owner he's got a share and I think you've technically got another share because you're also an artist distributing through the platform. And all of the the fans as well in addition to all
23:03
the workers that built code and such. Why this matters is that recently streaming has surpassed physical objects in terms of total sales. And so there's a lot of speculation that because it's the it's been this rebound in the music industry that's suddenly making the music
23:24
industry profitable again that what we're going to end up happening is that all of the other channels are going to start to kind of dwindle down and we're going to end up with streaming and then there's going to be consolidation in the streaming business and we're going to end up with a couple of access points to all music and it won't become about ownership anymore it'll
23:43
become about access. Don't have a lot of time to go into why I think that's a little bit disturbing but let's move on to talking about a different model for streaming and that's stream to own and that's the the model we're in beta right now and we've taken the whole
24:03
thing and turned it upside down on its head. What I found when I did my initial research on this was that the big problem with payments and why there's so much controversy around all of this is that with the monthly subscription model it's really hard to reconcile all the different listening patterns and come up with a fair system for paying artists. And so by moving away from
24:26
monthly subscription and going to a top-up model where you put down five or ten euros we can do fair payments we can do the same payment to everybody for every play and so basically what we've done is break we broke up the price of a download over nine plays.
24:46
So you put in a five or ten euro balance and then you start listening and then for each time that you listen to the same song the price starts to go up a little bit until you've reached the ninth play at that point you own the song you don't have to pay for it again from us
25:03
and you could even access the file if you want to. The really interesting thing about how the formula works is that you kind of get two distinct phases you've got a discovery phase and a fan phase in the discovery phase it's super cheap it's actually it can be as little as two to four dollars a month and you're getting to know new work and new artists
25:27
and then when you get to the more like the fan phase if you really like a particular song that's when you start paying a bit more and you reach that price of it of a download and this is where the gravity thing comes in and it's been really exciting to actually start working
25:43
with this and like you know I listen like two or three hours a day it's funny because you know they gave me the their catalog and I didn't want to listen to it I actually wanted to wait until I got it in the app because I need to test the app and I've become big fans of a couple of your artists as a result of this model and I think that what we're looking at with
26:03
this endlessly available content stream that comes from Spotify, YouTube, etc is that there's not a lot of gravity there's not a lot of connection to artists anymore because you constantly just skip from one thing to the other but by changing the whole business model around to where these little
26:21
micro transactions occur and you can look and see oh that's the third time I've heard that song and you know wait I've paid them just a little bit more than the last time I listened to it and you can feel like oh I've hit the fourth and fifth listen of the same song and how I'm like realizing I'm supporting this artist now because I'm giving them more money than they can make
26:40
on other platforms. The math for the regular streaming market is is pretty horrible you have to as a purely independent artist you have to have your song listened to 150 times to equal the cost of a download and so while streaming on on the whole has led to more
27:02
profits in the music industry for artists like his it's actually you're suffering because the per stream rates that you get with these other platforms are so small that you're the chances are you're never going to reach 150 plays on the same song by one fan in order to
27:20
reach that that price. So contact info and I look forward to the discussion. Okay thanks Peter and maybe yeah we click to the next slide um yeah these were actually our three perspectives on music streaming thanks a lot and as I said we will now have also time if you have
27:45
questions or if you have a comment or want to know much more do we have a mic here in the in the room ah perfect okay just raise your hand we have someone who assists with that yeah um thanks to you all I think these perspectives kind of come together at some point and we will
28:04
find out now where these points are but uh Arianna when I read your Rolling Stone article on how Spotify calculates a music taste I've noticed that you use two words to describe this kind of algorithmic music recommendations creepy and terrifying um what is so scary about I mean
28:25
Discover Weekly or uh Simula Systems what makes you yeah have nightmares about this um maybe sometimes I have nightmares that I like some of the songs maybe that's creepy um
28:41
um well the thing is um the thing about genre in general general maybe is that music is about emotion I don't know that people who say oh I like this song because it's a post-rock song and it's not a I don't know new gay song or whatever people like a song because they like
29:00
a song and what an algorithm tries to do is to deconstruct a song into little pieces so that into numbers and um I just find it extremely unnatural and um and even if if if the suggestions sometimes fit your taste it's a bit um it's a bit scary that this machine
29:26
is putting music into something like this because I am and also music taste changes by time and also um I think that genres are getting less and less important um people are listening
29:41
to music in different situations in different moods uh different times different forms of a day and so um it's weird to say okay here is this curated hip-hop playlist and um then you can listen to hip-hop like two hours but you you don't um there is nothing to um to discover
30:06
yourself anymore and maybe that's a bit weird for me that you have this this genre tree which is very impressive but also not the way that people perceive music I would say okay so you you feel like you're in a musical echo chamber or something like that sort of
30:26
this uh typical filter bubble effect which accounts now also for music and musical taste maybe um yeah but if you look at it from another perspective you could say okay now um these platforms they account for users tastes uh for the emotions uh and they like
30:46
want to get to know you all of this stuff yeah and wouldn't it be if you compare it with classical or with traditional radio djing djs for example the classical taste makers when you say that
31:00
these these new possibilities to discover music are like a democratization or liberation of taste making um isn't there also this point that now users are are taking into account in another direction um well I'm I'm not so oh sorry I'm not so critical um
31:30
in a way that I would say people should not listen to to discover weekly or something like that I think um people always have the choice to choose the music themselves they can go out
31:42
they can go into record stores and discover something themselves so um I wouldn't say it's a threat because you can also see this as an addition and of course um 20 years ago people went to a record store and there was the the uh the shop assistant and you also got suggestions
32:05
from a person and then you were listening to 10 different records and then maybe you liked one or two of them and maybe it's the same with this algorithm you get 30 songs and maybe you like five of them and then you're happy so I wouldn't say it's necessarily bad the only thing that I criticize is whether um whether you can approach music from from this way whether you
32:27
can say uh you can put music into playlists just because the numbers of the mattresses told them so because for example what what what genre are the Beatles or what genre is
32:40
David Bowie where where would you put this and so I think that you can't you can't put music into those kind of numbers I would say but still of course um it's a chance because you can still be your own person you can still choose yourself and you can choose wisely and
33:00
maybe uh technology is an addition it's maybe it's the same discussion that we had when television came up and people were afraid that people won't read books anymore this is not true it's just something that is an addition it won't replace the way we listened to music before but maybe you just have to be very aware yeah of how it works um Peter yeah yeah I think I
33:25
was just going to jump in on the this this notion of algorithmic curation versus human curation and the point about human curation is that it's unexpected um I feel very lucky to be pre-digital when it comes to my uh experience in history with music and to me the the people
33:47
that um really turned me on to stuff it was the things that they turned me on to was stuff I didn't know I was going to like and that that changed my perspective about what you could do with whether you know a guitar and a bass and drums or you know something else entirely
34:07
and and I think that's something that's missing from this algorithmic curation and I hope that that's something we'll be able to do and like you said to democratize curation in a way that puts that human element back in because you know it is it's too sanitary it's it's like
34:25
the Walmart of to use an American reference uh it's the Walmart of you know music consumption it's too generic um yeah I just want to jump in there too because uh actually there is one one feature within Spotify um where you can do this uh like the user-created playlists and
34:43
for us as a label those user-created playlists they become so important because it's one of the most important promotional tools nowadays which 10 years ago this would have been the music blog then you would send it to the blogger which uh which would have been the taste maker then like the new radio DJ 10 years ago was a music blogger
35:00
and now it's the playlist maker um or the playlist curator and and then it's also human so um and some of them are like really really powerful like I don't for example from hip-hop you have one playlist German German rap best of German rap or something which is monthly and if you have a track there you know it has like 20,000 plays like almost instantly because the list is
35:21
subscribed by so much um users um but on the other hand just coming back to your question is this is this um not a democratization I would say this is the problem because what are the economical and also like data circumstances where where this happens in and this is not democracy at all because like Peter mentioned in the talk before we had at lunch he's like
35:44
the business model of Spotify is not a subscription it's like selling the data but it's technically it's about selling stock um that's what Spotify's business model is it's not about making money off of music it's about their IPO um but one of their main uh revenue streams is around the
36:04
aggregation of data profiles on on listeners um which is a good point I want to come back later that you or resonate works in open source ethos I really find it interesting but we have a comment here from the audience interaction yes hello hello hello yeah Peter just mentioned
36:23
what I wanted to say hello Peter Dennis here you know um I don't feel the algorithm at all because it's always the same thing it has always been through history the same thing people tell you something and you just mentioned that I don't feel that people don't look for something they have never seen I mean it's like if we are humans we will always search for
36:42
something an algorithm will not tell us we'll always search and go for something it's always the same thing like when you listen to radio you listen your favorite radio because you want to listen to your radio that does the algorithm so I don't fear this and if you're a music lover you will go somewhere and if someone is a music lover he will send you a link and tell you this is wonderful what I fear is the data that's the problem if one day you're listening to Spotify
37:05
and you listen I don't know maybe you're listening to punk rock all of your day and then and people know your political idea also that's what I don't like about this I'd so the algorithm for me is perfect it's wonderful it's made by humans it's there to help you so uh well you
37:21
just said the data collecting and you are the you are you are giving data to people who have made money that's what I don't like about it and that's that's what I meant by the corporate governance model for culture is that the corporation sees you as an asset and and nothing more and so that's why we're we're pushing a totally different perspective on that
37:46
and it is it is very very scary I know someone who actually read the terms and conditions and left the the platform I mean who does this nobody reads the terms and conditions anything I think it takes like three to four months if you read all the terms for all the apps you use
38:03
it would take three to four months of your year and you have to do it again the next year well you have to read it and understand it one more year you have to understand it too that's just reading it yeah exactly and and so yeah this is and it's going to be really interesting what happens you know like 2018 in the in the EU um if that's going to have an impact on the way
38:22
that these these businesses operate in terms of data collection but we'll see but yes uh Peter as I said before um you uh follow another uh road I'd say would resonate regarding uh openness transparency at least this is what I read on your website um could you explain a bit more
38:41
on this what you do or with data for example um well it's still very very early days so we are collecting data I mean because we're we're looking at the way that the player works but the uh when we really go live the goal is to be able to make everything opt in
39:02
not opt out so if you want to be a fully anonymous user and not have any of your data um you know saved um you could do that and so uh where it becomes an advantage is that if you want to create a communication channel with an artist and so you want the artist to actually
39:21
know where you are because maybe you want to get a show alert you want to find out that they're coming to your town um but uh this is a part of the cooperative kind of model is is is treating everyone as an equal member and yeah this was something I was interested in as well so how do you create this connection between our artist and fans I mean this is something for you
39:47
as analog soul as a label or let's say network or platform which is interesting as well so the question of local connections for example I think you are part of listen to life yes um how
40:02
how can these global companies reflect on the relevance of local specifics or subcultural specifics for example um is there something that plays a role for you as well as a label well in in the beginning of this whole um like internet and streaming stuff like when we started we
40:21
started almost 10 years ago um it was thrilling for us to see wow somebody from New Zealand played you or then or from Canada something which was like totally totally strange for us as really really small indie artists um but then we we figured out this whole world of music blocks and then they had some of those those block connections where there was a network of
40:41
international music blogs and each month one blog created a play or they created a monthly playlist and every block out of this network um contributed one track from the country and so we started figuring out how those connections work and how yeah how to get in contact with people um but the thing with streaming is um we as artists have almost no insight there is now
41:02
Spotify stats there are some stats you can look at into your uh albums or playlists you create but um it's definitely not the kind of connection you get for example uh on SoundCloud which which is something which works uh at least in in my understanding and how all fans use it or the the networks we are in use it more like a community where you share tracks where you
41:26
comment on tracks where you comment on the music and this is something which usually doesn't happen on on those big streaming devices big streaming services and most of them don't have any message uh message function or something you can follow you can follow a profile but the
41:43
profile cannot follow back in that sense because the artist's profile cannot follow back the those those platforms became in in that in that sense a setback for us in terms of creating connection because this is something we've always been good in to create connection with the
42:01
people so we had small fan bases but with with high connection and um we we did not lose this on streaming but we don't get anything back out of it except from the revenue share we see in the end of the month where was it streamed or in which countries um so and yeah um I'm looking forward if if uh Resonate is going to get bigger and more people are going into it
42:24
if those connections like tweet like uh Peter discovering our music through the app uh and reporting us on it uh will become more I have a comment on that um well one thing that I
42:41
wanted to say is um talking about data privacy is that um it's not only about personal data but also data bubbles and we talked about this earlier um at lunch that um Spotify has the the uh information about every song that is clicked by any person at any time ever since
43:03
they like subscribed on the on the platform and so it's not only about your privacy but also about um facts like Spotify knows which songs are being streamed in Berlin on a Friday night from
43:21
10 o'clock to midnight for example and so they know what kind of music people like in Berlin Neukölln compared to Munich for example and companies definitely will be interested in this kind of data for example for advertising for commercials um so they know which kind of songs are popular in which area and I think that this is um very relevant for in terms of data privacy
43:49
but will it lead to I mean with all this data available and andres just said you don't actually have that much data on on usage past patterns so um does it influence the shape
44:02
and structure of popular music I mean it would be nice to see uh the patterns that are most popular and uh and musicians are like uh orienting the stuff on on that is there are there plans for a Spotify optimized album I mean we actually we know this kind of algorithm
44:22
in an analog way right because if you look at for example those and 90s um Ibiza mainstream techno pop songs they they all sound like they were curated from an algorithm right and from a kind of stupid algorithm because it is stupid in terms of how much maths goes actually
44:43
into um so so if we talk about taste for example or what what is popular I think we have ever since music is distributed in a mass market since like beginning of the 20th century we have those algorithms already at work in a very simplistic way I would say to say like give
45:03
it four to the floor and then here we need a rise and there we need a bridge and we need female vocals in the reference and things like this and um so but definitely um there would be people interested to have a algorithmically ideally shaped song like when do and and then
45:21
it would be interesting for me as as also like a media scientist or also a sociologist point of view what are the criteria they put in to say this is a successful song and like I know the music industry the criterion would be when do people click away so we will if we have if we would have music generated but on based on this user behavior then it would be probably
45:43
music just like like elevator music it would be music that that does not disturb anyone but it wouldn't be music that inspires anyone um just as a comment I'm thinking of Jan mentioned maybe you've heard of this um it's I just came into my mind that is a very good
46:04
example for this so this is kind of a monkey algorithm in that case so um he made monkeys decide upon the text of a popular or a pop tune and it's a bloody earworm by the way yeah it's really catchy so the monkeys did that job very well so um but when we talk about
46:25
streaming we also talk about licensing and I'm not a musician I mean I have some ukulele tracks on Spotify but no on SoundCloud oh god Spotify no and but this is a private matter so please don't google it um however um yes um streaming I think uh so when we talked about
46:46
three years ago uh Andreas for example um many musicians uh were artists were like really concerned about going on Spotify putting the music out there based mostly based on the on the business model so we all know this things like Jay-Z pulling all of all of his tracks from from
47:06
Apple Music I think and Spotify post so maybe Peter um can you tell us a bit more about why for example artists and musicians come to resonate and want to be part of the cooperative
47:21
is it more about sustainable business and and getting more revenue or is it really also a political decision or an idealistic decision oh there definitely is a political side I'm sure I mean as an electronica producer myself who went through um that experience uh in the mid 2000s
47:45
when Spotify or rather SoundCloud changed their business model and suddenly half of my catalog was gone unless I decided to pay them um for hosting it's like this is the the prime example for me personally and I think that most producers and bands and stuff have been through something
48:03
like this with a platform and that's the that's where having some kind of voice is is fundamental and I think that's a big draw as well as the possibility that maybe we'll make more revenue but I hope it's also about that um being able to open up more pathways
48:20
and communication channels you know if a band comes to us and says hey we're using this new app that allows us to send whatever to our fans could you somehow integrate it um you know as long as we've got the resources to do so and I'd like to do that you know and I think that could provide a lot more opportunity and the fact that they can ask that question
48:44
and have it considered and maybe even like say well we've got to actually take that to a vote because that's causes some other problems so let's ask the community to vote on it you know that kind of democrat democratic process you know you don't get anywhere else so I think that is a big appeal to why people are coming in yeah speaking of stream of
49:06
consciousness of streaming I mean this is something you have to explain to users and if we look at the data and especially younger users are really enjoying the pleasures of having like a small amount of money every month and paying that you have to like you lose the
49:27
idea of what it means for the artists when you stream this stuff on these big platforms I think so it's kind of educational I'd say what you try to actually achieve I think um we are just
49:42
I think we're running a bit out of time but of course you are invited to ask some questions so ah over there yes place a man in the second row hey there um Peter I was wondering how big is the share that the artists
50:01
receive at the end of the day with your platform because I think this is the main problem about Spotify it's not the 10 euros per month that's perfect amount it's just nine years go into marketing and one euro goes to the artists um well I mean actually the the
50:20
payout is much more dramatic than that Spotify is paying out 70 to like 83 percent I think is the total number of their revenue is going out it's just that the the way it's being distributed a lot of it's ending up at the major labels and there's there's no um there's no equality in terms of how it gets broken down um kudos to Deezer they're considering what's known as the
50:48
listener centric model which would make the distribution of funds based on what you're listening to not the way the the the pie is um divided up uh Spotifyartist.com has the chart
51:01
that explains how they do it and it's it's really complicated so um that's part of it with us it's it's totally one to one um in terms of you know because it's a different process you're you're topping up and so when the deduction occurs from a play it goes into that artist account instantly um I think I missed the you had another kind of the revenue that resonated
51:24
it could it we our calculations are two and a half times more um kind of on a conservative we did really conservative numbers and it was all uh estimations because we don't really know we won't we won't know for another three or four months until we were able to look at the
51:41
the data on it um so andres there's also another part of it too is that equality like you know you mentioned that uh users skipping uh a track at like halfway through um by having repeat plays and knowing how many fans you've got that are listening to
52:02
four five six seven times um and paying that much more you you really know who your fans are and you know what songs are really meaningful to your audience yeah i just wanted to add because one might think that um spotify and those services um yeah they like they calculate like every played song is put in the revenue but it's not the
52:23
case like because spotify and you could go into this you could have a look at google it's there were some interesting articles spotify actually kind was blackmailed by the three big major labels worldwide um and they had some specific contracts and they had to actually they had to agree to put like sony i would say 20 on top than the other three major labels or something
52:46
so every one of the big major labels has an own contract with spotify and in this contract it's written down how much of the revenue is going to the to the labels so um and and estimated 35 percent of the music worldwide is music made by indies like acts not on the on
53:03
one of the three companies big label publishing corps um and yeah and and that's that's something we we should not um we should not mix together actually there are not many services out there who are like those that basic and clear under structure that you click it and then the artist
53:22
gets paid that's unfortunately that's not a stand-up model not yet so okay i think there was another question um is it answered yet no then a quick one please a quick one please yeah i was i was wondering if you have any uh this one goes to paul uh i was wondering if you have any other
53:42
revenue streams um in your business model like let's say tracking or whatever it could be you said tracking yeah like tracking or like if you if you get receive any data from users and sell them to hurt no that'll never be a revenue stream for us we have no interest in that
54:01
whatsoever i don't want it and it'll as long as i'm running the company it'll never be there but there are lots of other opportunities where we could work with other startups we've had about 40 sign up to to work with us in the future um for ticketing or merchandise or vr ar blah blah blah it goes on and on the opportunities out there and yeah and in the future it may be
54:24
that streaming is just a loss leader where it ends up being totally free but there's so many other ways for an artist to make a living that we don't need to actually charge for streaming i mean that's that's an eventuality it's possible thanks okay you're happy with that thanks i mean there's plenty of time to talk about
54:45
the business model behind resonate after this session but unfortunately we're running out of time today um maybe we have a quick quick short round of statements what are the biggest challenges in music streaming ahead really quick really quick something you mentioned actually
55:03
i think we need to talk to explain to also to the artists and to the listeners better how these services work and how the politics and economics of these services are what he said i would say listeners should not get lazy um but we also have a best of both worlds so
55:25
a curated playlist um or a music journalist um can go to shows and see things that an algorithm can't do but an algorithm can calculate in millions of songs that a human being can't do so
55:42
maybe we can have both okay thanks um and maybe we could end this session with something you mentioned in your article ariana actually which fits quite well to this year's republicas motto love out loud because apparently spotty now allows you to check on tinder if the
56:02
if the music tastes of the potential subject and your music tastes fit to each other so maybe uh streaming is also something about bringing people together spreading love and stuff like that emotional experiences yeah considering my own relationship i'm not sure
56:21
well so um however i want to encourage you all um to yes keep on supporting artists uh buying CDs uh go to resonates uh go to analytics so listen to uh ariana's very cool music on suncloud i really enjoyed it and yes keep on supporting the artists and music you you like
56:44
and care about because this is also uh something about love and not only about that bad data and algorithm and stuff so um thank you all for coming here today let's give a warm round applause of applause to our speakers to andreas bishop to harris and ariana sistra
57:01
my name is sila heisse thanks and enjoy the rest of the conference