One of our project members (Rasmus Fleischer) was recently interviewed by the Swedish news journal Dagens Industri. The interview mainly came to focus on Spotify’s history and our upcoming book.
The launch of Songblocker at Transmediale was a great success and since then the company has announced that they’ve hired a new CFO – Mary Jane Smith, who’ve recently held a position at Boston Consulting Group.
Don’t miss the Songblocker ad which was presented in Berlin (credits to Christian Rossipal):
On YouTube, the video received more than 15 000 views in less than one week.
Songblocker is now also present on several social media outlets such as Twitter:
Songblocker gives you 100% ads when listening to Spotify. Try it out for yourself – just download the plugin at https://t.co/APHPA8J7De!
— Songblocker (@Songblocker) February 13, 2017
On Sunday (February 5th), one of our side projects will be launched at the opening festival of Transmediale in Berlin.
As an intervention in debates around ad-blockers, we’ve built an application called Songblocker which mutes music on Spotify and instead allows you to enjoy 100% ads. The application is part of our ongoing research on the organization of ad-markets around streamed music and the abstract for our presentation reads as follows:
“Have you ever dreamed of full access to the world of ads? During this exclusive event, Songblocker will be launched—a powerful tool engineered for satisfying the needs of advertisement aficionados. Finally an app that saturates your Spotify account with 100 percent ads: no music, no disturbance. The launch of Songblocker is followed by a presentation of an interventionist and experimental research project on Spotify.”
Read more about the project at Songblocker.com.
We are happy to announce that our research group has just signed a book contract with MIT Press. During 2017 the last manuscript is expected to be finished, and hopefully you’ll be able to pick up a fresh copy shortly thereafter.
The book – which is currently entitled “Spotify Teardown” – is planned to pull together some of our ongoing research, which for example involves the programming of multiple bots to explore, mimic, and ultimately subvert Spotify’s notions of usage and listening; ”monetary interventions” involving calculated musical tinkering that automatically play tracks for thousands of repetitions and thus tests the boundaries of online royalty systems; the documentation and tracing of Spotify’s history through constantly changing interfaces; the charting of the online music industries, and interviews/interventions with providers at the backend of Spotify.
As a whole, the text will be co-written and experimental in its form, serving as an organizing device for our ongoing research, rather than a retrospective summary of project findings. Written at the intersection of software studies, critical media industry studies, the digital humanities and social anthropology, we are very much looking forward to start the process of writing.
Youtubify is the name of a new piece of software, presented as a “Youtube music engine”. As the suffix -ify makes clear, the intention is to create a kind of Spotify clone.
Anobody who installs Youtubify on a server can create something that appears like their own music streaming service – without having any license for the music. Dozens of such clones are already up and running, Torrentfreak reports. The most public is to be found at youtubify.vebto.com which does not even require registration; immediately after loading the webpage, playback is ready to begin.
Without doubt, the record industry is already preparing for legal action to stop Youtubify. The outcome is not clear. It might be possible to target the distributor of the software, an anonymous user named Vebto using the market place Envato. But when a piece of software is out in circulation, it is unlikely to disappear (although it should be noted that Youtubify is not open source).
Another route of legal action would be to target every server where Youtubify is up and running. However, there is no guarantee that this would succeed in every jurisdiction. In particular, it is far from certain that the Russian judiciary is eager to help the US/EU record industry in the current political climate.
Youtubify is, after all, just an interface, an alternative way to access Youtube. (The next logical step would be to also create support for Soundcloud and Bandcamp.) It is not really true that everybody can now “create their own music streaming service”. On the back-end, every single clone using this software still relies on one single streaming service. However, it should be noted that Youtubiby also includes functions for statistical surveillance of music usage.
But from the user’s perspective, isn’t Youtubify still a clone of Spotify? That depends on what kind of use that we understand as central to the concept of music streaming services. This raises the question of what Spotify has to offer its users.
One possible answer is “music”. Most people in Sweden would probably still think about Spotify as a music service, in the sense of an service delivering songs and albums on demand. But over the years, especially after its US launch, Spotify has transformed itself towards offering something much broader.
Jeremy Wade Morris and Devon Powers writes in an excellent article titled “Control, curation and musical experience in streaming music services“, recently published in Creative Industries Journal:
Ultimately, we suggest that digital music services no longer sell discrete musical objects, nor do they focus exclusively on content offerings. Instead, services sell branded musical experiences, inviting consumers to see themselves and their attitudes, habits and sentiments about music reflected by the service they choose to adopt.
The control users have over the choice of music they listen to, over the platforms they listen to music on and even over the files themselves is exchanged for a branded musical experience that foregrounds instant, multi-platform accessibility. The stream, despite attempts by providers to position it as ever flowing and always on, is just as often an opportunity to separate, segment and differentiate among different levels of consumers, and different groupings of musical consumption activities.
As of today, Spotify wants to be a service that offers a seamless flow of music, following the user through different moods, activities and times of the day. This is a concept built on access to big data, and it does not seem like Youtubify can offer anything close to that. In fact, its interface more looks like a clone of early Spotify, based on the idea that the individual listener will always know what music she wants to listen to.
In other words, the proliferation of Youtubify servers opens up a truly interesting question: Are users of music streaming services simply “in it for the music”? Or are they, on the contrary, really desiring a “branded musical experience”, functioning as an algorithmic mirror of their own life?
This blog post, written by Rasmus Fleischer, is based on a post published on his Swedish-language blog.
When streamed music is exhibited on streaming platforms, it is often presented in a friendly and collective manor. Music files are frequently made to socialize with each other by being bundled up, or tied together through trails of data that detect similar artists, common histories, or auditory commonalities between sounds and artists. These alliances, or bonds, between music files and musicians do not just shape how music is framed, but also how it is enjoyed. The social networks of streamed music files are what underlie promises to enhance musical discoveries and listening experiences.
Streamed music files are frequently also made to socialize with audiences, since their movement and display is largely connected to the analysis of actual user practices–or estimations thereof. In a recent experiment Anna Johansson and Maria Eriksson are exploring how music recommendations are entangled with fantasies of listening–often based estimations of taste depending on user’s age, gender, and geography. By capturing and analyzing the music recommendations Spotify delivers to a selected number of pre-designed Spotify users, their experiment sets out to explore how the Spotify client, and it’s algorithms, are performative of user identities and taste constellations.
Partly informed by recent works in the field of software studies, their work address questions regarding the tension between (imaginary) publics and Spotify’s promise to deliver individualized or highly personalized music recommendations to everyone. It also ties into broader questions about the workings and effects of algorithmic knowledge production. An article is currently under way, and will be published in the journal Cultural Analysis during 2016.
A few weeks ago, Maria Eriksson attended the San Francisco Music Tech Summit – a yearly get together for businesses, investors and journalists working in the field of music and technology. The event – which mainly consisted of discussion seminars and product displays – took place between the 10th and the 11th of November, and gave lots of insights about where the music tech scene imagines that their future might lead.
During the summit, discussion panels for example centered on issues relating to data mining, music analysis, market expansion, royalty payments, music recommendation systems, and current legal grey-zones within the music industries. Data driven PR/branding strategies for artists and record labels was also widely discussed, and the following four things became especially noteworthy during the event:
In terms of discussing where the future of music recommendations is heading, it seems like more and more businesses and actors are getting their eyes set on biometrically driven forms of music personalization. This includes the merging of music recommendation systems with wearable technologies that for example detect what persons are present in a room, or the pulse, mood, and bodily state of individual listeners. If streaming platforms (and other music services) have been engaged with trying to establish intimate knowledge of their listeners for a long time, this indicates that they are now also occupied with trying to get under their skin–quite literally–by basing music recommendations on bodily features such as skin temperature and the pace of blood streams.
During the event, it also became apparent that lyrics have so far been a (surprisingly) overlooked element in music recommendation systems. This for example became clear when Abhay Kashyap presented his work on “Rapalytics” – an application which analyzes rap songs based on lyrics. While vast and multiple forms of data now underlie how streamed music moves and is displayed, it seems like one of the most obvious elements of music–its lyrics–have so far been put to the side. What are the logics and reasons behind music recommendation systems–as of yet–mostly appearing as illiterate devices in terms of lyrics?
When “the streaming genie is now out of the bottle, and is not likely to be put back in”, as Harris Frank, president of Interlude Music put it, new ways of finding additional sources of revenue for artists seems to occupy many peoples mind. For example, the company Booktrack suggested that artists should sell their music to soundtrack to e-books in order to help readers make their “inner imagination come alive” – thus suggesting that the field of activities that have music added to them is likely to expand in the future. As a way to safely monitor and keep track of such musical streams, the idea of introducing block chain technologies was also put forward by Stephen White, CEO of Dubset Media Holdings.
Last, after a long period of debating uneven royalty payments to artists, discussions around the controversial revenue models of streaming platforms now seem to have taken a new turn. Instead of primarily arguing for higher royalty payments, the spotlight has now instead been directed towards vaporizing streams – that is, streams that simply seem to disappear and aren’t registered as earnings. The controversy started with the digital monitoring company Audiam finding out that Spotify had not been paying the indie label Victory Records for a large amount of their streams. Arguing that the once lost – but now re-discovered streams – might only be the tip of an iceberg, accurate measurement of music streams has now become a central concern. What does it mean that streams might “fall through the cracks” of streaming platforms?
Spotify has recently claimed to be “obsessed with figuring out how to bring music into every part of your life, wherever you are, whatever you’re doing, whatever your mood.” This intense desire to soundtrack the everyday could partly be understood as an outcome of a business structure where scale is everything. Spotify’s promise to artists and other collaborators is entirely dependent upon delivering more streams, more users, and more listening. Hence, clever and timely content deliveries are nowadays frequently framed as the main method to attract new users, and thus increase revenue.
But from a larger perspective, preoccupations with timely content provision is also a major feature of digital media in general. “Real-time” updates that accommodates to “real-life” situations are both the promise and peril of digital media, or, as Geert Lovink has put it, its “new crack.” Tied to idealized notions of freshness and flexibility, the desire for instantaneous deliveries–and claims to provide it–is something which permeates the digital environments of today.
In one of our ongoing experiments, we are curious to explore how notions of time (and notions of time well spent) are embedded into services such as Spotify. We are also interested in tracing the tides and ebbs of musical deliveries, thus tapping into another rhythm than the one which descends from the music itself; that of data traffic highways.
For this reason, we’ve for example been working on extracting some of the music recommendations and greeting messages that Spotify delivers to its users (see one rough example of this below). Updated on a continuous basis and designed to carefully cater to the expected lives of its listeners, such greetings speak of the ways in which temporal imaginations are built into digital platforms.
Based on this material, and other additional data, we are currently asking ourselves how ideas of immediacy and ‘real-time’ are constructed and embedded into the Spotify platform.
How does technological features (such as Spotify’s user greetings and musical deliveries for ‘every moment’) help to organize time? What is the pace of digital music streams? And what are the political implications of the ways in which time is addressed by online platforms and its curators?