Streaming Machinery has a new address: www.gcstein.com
Music. And Spreadsheets.
Bono famously didn’t like that iTunes looked like a spreadsheet. He told that himself to Steve Jobs in 2009:
So 5 years ago I began a conversation with Steve Jobs at my house in France and I said to Steve: “How is it that for a person who cares about the way things look and feel more than anyone else in the world that iTunes looks like a spreadsheet?” He was not happy. That was before they managed to even get a full screen photograph up. So he made a promise to me that we would work on this together, and with the team at Apple we’ve been doing it for years, and it’s not ready yet for Songs of Innocence, it will be ready for Songs of Experience. And it’s very exciting.
(Of course the release of Songs of Innocence ended up being the opposite of exciting, and when Songs of Experience was finally finished, after several delays, it had a much discreeter launch three years later.)
Earlier in 2014 Spotify took some bold steps to make their app look less a productivity app and more like “going to a movie”:
Spotify says it spent six months working on the redesign, which is rolling out to its users from today. The new theme is heavy on black backgrounds, in an attempt to place more focus on album and artist artwork.
“It’s like when you go to the cinema, and they dim the lights, and the movie steps forward and takes over the room,” Spotify’s director of product Michelle Kadir told The Guardian. “When you log on to Spotify now, everything else is secondary except the music, which is popping.”
Yeah, spreadsheets are boring and lame but maybe by now digital music services are mature enough to recognize and embrace the opportunities available if they could embrace their “spreadsheet heritage” a little more full-heartedly. After all, sorry the pun but “playlists” are half “play” and half “lists”.
Excel and its ilk have been refining their capabilities to aggregate, analyze, manipulate data for decades now. Song libraries with more than 10,000 songs, a fast expanding universe of over half a hundred million of songs and a few billions playlists created, maintained and shared by users are all-together a rich enough “dataset” that could use more features. And maybe those features in turn could bring real utility for listeners.
Starting with the basics, Spotify could allow users to customize columns of a song list, like iTunes has allowed for two decades. The difference being that Spotify has much much more metadata than what we could add to those ID3 fields back then. The acquirers of Echo Nest could provide some really interesting columns!
The columns could serve as filters. So it would be easy to search songs with multiple criteria and within ranges of values.
A single list does not has to be “monolithic”. It should be possible to distribute the same content in multiple “layers”. Excel supports multi-worksheets since 1993! With those, it could be much easier to organize and analyze some stuff. For example, a “megalist “ could be automatically organized in different smaller lists by genre, decade, label, “mood” etc. So you can look at stuff separately without losing the whole.
Basic Excel tools like grouping and collapsing data could be useful.
Filter songs from a list: likened, disliked or neither. Added or not added. Listened or not listened. Already in particular a playlist or not. What users would come up with Spotify PivotTables??
Between “songs”, “albums”, “artists”, “playlists” and “friends” there are many dimensions of interlinked data that users could explore their way into.
How useful these features would be for power curators to create their ultimate playlists?
“Genre” as an organizing tool in Spotify is a mess. A catch-all term for all kind of stuff and an organizing crutch.
Browsing for genre currently you will find: Playlists for Students. Music + Talk episodes. “Songwriters”. Obama’s Higher Ground. And where are “enhanced albums”? Spotify Clips?
Isn’t about time to adopt a more robust, productivity-app inspired way to organize and make sense of content types? If “folders” are too bland, maybe call them “shelves” or “racks”??
The home page now shows rows of playlists organized thematically. There “Queen” playlists, “Good Dreams”, “Workout”. There is no way to search, browse, save, organize these “playlists groups”.
Spotify does have a lot of cutting-edge analytics. But direct access to it is usually outside the app.
The company has also been adding useful new features increasingly faster. Like genre and mood filters (only in English speaking countries) and detailed, collapsable/expandable listening histories, even “taste match scores”. But the core philosophy of “sparing” the user from complex choices is maintained.
Productivity apps have long focused on bringing new features and catering to both personal and professional users, from the most clueless beginners to the advanced specialists. Their makers are comfortable with notions like “90% of users use 10% of the features and 10% of users use 90% of features”, and evolving those features so they become both more powerful and/or easier to use.
It would really be an interesting experiment seeing a music streaming embracing the “complexity” and offering open-ended tools that users could try for themselves and find useful applications.
Maybe adopting some inspiration and references from a boring productivity app could be the boldest innovation for a truly ground breaking music service?
Playlists as Fast Food
Complaining about playlist is old news by now. Cherie Hu was articulating “Our New “post-playlist” Reality” almost three years ago.
Yet, by design and by necessity, the playlists that get the most reach and engagement on Spotify treat their artists as disposable.
That might sound bleak, but just think about it: Spotify’s playlists need to serve Spotify as a product first and foremost before serving the artists they feature. As a subscription service, Spotify needs to keep users “sticky” and engaged on their platform, which in part means not turning users away with badly curated playlists.
What this means for curation depends on the playlist. For more genre-focused channels, the content needs to be updated regularly and kept as “fresh” as possible; for more mood-driven channels, the content doesn’t need to change as frequently, so long as it continues to provide functional value to the end user, regardless of which artists show up. In both cases, the artist is disposable in service of the product, just like features are disposable in the ongoing development and improvement of software.
Hence, “post-playlist” can be characterized as a growing attitude of disillusionment among up-and-coming artists and labels that playlists are not as meaningful to or aligned with their business as the hype had promised.
To my taste, as a music fan, there is something of a fast-food aspect to how playlists are usually presented on Spotify:
A big bold menu with simple choices.
All choices ready for instant consumption. Little extra information, context or anything that might distract the customer.
Little way to really make a custom-made list, mix “flavors” or even plan a multi course “meal” with a sense of beginning and of ending.
(It is somewhat ironic that restaurants and other business have much better options to either choose instant music (like “AI-generated stations from sound tags”) or program a full schedule (“Drag-and-drop scheduling”) from Spotify’s partner Soundtrack Your Brand.)
There is no real way to “savor” a playlist thru days or weeks. Every time you choose a list it starts in song #1 or at some random point in shuffle mode. Every listening session is in practice as disposable as a less than warm Big Mac.
Another way in which the playlists are perishable: New songs come in, old songs move to the bottom and then are thrown away. You can listen and “save” them only for a while. There is no way to go back one playlist as it was at any point in the past. There is only “now” at Spotify’s “counter”.
Certainly this all means a lot of convenience for users. Most of the time they are in a hurry and need something quick to start listening.
But this cannot be the only way. There must be more options to serve music. Some rough possible ideas:
Richer ways to explore the alternatives.
Structured ways to browse the content.
Subtler ways to present recommendations.
Tools for the listener prepare, organize, select listening sessions.
Practical ways to switch between listening “modes” (for examples, a robust queue of new music that the user can play intermittently, only when in the mood. Or it could be easy to try a new playlist and later return to the previous one at the exact point it was switched etc.)
Ways to shape the menu to their taste and circumstances.
Ample access to trusted curators with good and compatible taste.ll
Access to more context and information.
Dedicated apps for specific uses and genres…
Even as metaphor fast food should be used in moderation!
Spotify and Microsoft
A welcome development:
Billboard will launch a newsletter in October highlighting the unfiltered thoughts of the CEOs, decision-makers and power players at the intersection of technology and music, written by Billboard’s director of technology coverage, Micah Singleton.
Great topic for the inaugural column:
Will Spotify Stay Independent — and Should It?
Spotify is growing slower than it has in previous years, its profitability is negligible, and the company’s main competitors are loss-leaders for their respective corporations (Apple and Amazon), which can pump millions into them at any time.
…
“A year ago? I could see Microsoft buying Spotify,” one CEO says. “Now? That’s very hard with M&A. It’s becoming harder and harder to buy it.” The newly-emboldened Federal Trade Commission — led by chair Lina Khan, a longtime proponent of breaking up the big tech conglomerates — would likely raise hell if a tech giant tried to buy Spotify, but multiple executives brought up Netflix (whose co-CEO Ted Sarandos sits on Spotify’s board of directors) as a potential partner for an “untraditional merger based on streaming economics,” as one investor put it.
“The industry needs them not to be bought by a tech company,” another CEO says of Spotify. “We need competition, and we need a big player that is not Apple or Amazon.”
Two interesting things about seeing Microsoft being mentioned in this context:
1) It is a very neat overlap: Apple, Amazon and Alphabet are the three tech giants along Microsoft presenting a market cap of more than US$ 1 trillion. Apple Music, Amazon Music and Alphabet’s Youtube Music are the three major Spotify competitors.
2) Microsoft has a very long story of trying (and failing) to lead digital music. By now, Windows Media Player is 30 years old! This article by Cherie Hu is a great overview of the story. She counted “at least six different music services that Microsoft acquired, launched or powered, and subsequently spun off or shut down, between 2000 and 2018”. An acquisition of Spotify would be an amazing coda to this saga.
P.S.: Just for curiosity’s sake and to point to a great essay, here is Matthew Ball analysis of what Netflix could do on audio…
Compare all of the above to what it would take for Netflix to launch an on demand audio service. The world doesn’t need “another Spotify”, but the core stack is relatively easy to establish as Netflix already has on-demand audio functionality (i.e. tied to its video) and a robust recommendations engine (obviously more work would be needed). Furthermore, audio doesn’t require complex social features or multiplayer live ops. And with a few billion in minimum guarantees, Netflix could acquire non-exclusive rights to nearly all music ever created. $300MM a year could then provide the company with an expansive slate of originals, too. The business model is also straightforward: sell for $10 per month and pay out $7.5 in content costs. Audio is an important competitor in the “leisure wars”, and I don’t mean to suggest Netflix could easily conquer it. However, gaming is far more threatening in reach, implications, and entry.
Radio Chaos
It is always an odd experience to try Apple Music’s Radio service.
How it went this time: I go to the Radio main page and start browsing. There are several options, they seem all very nice. In the section “Every Show, on Demand” the first option is “80’s Radio with Huey Lewis”. I click on it.
The following page has some art with a Huey Lewis photo, a text description, the most recent 8 episodes and then a section “You Might Also Like” with another 4 shows. Finally, there is a “Listen Live” section recommending the live station Apple Music Hits.
The description: “Huey Lewis and his band The News dominated radio in the 1980s, sending 12 singles into the top 10, including “The Power of Love”, “The Heart of Rock and Roll” and “Hip to Be Square”. Here, on his own show, ’80s Radio With Huey Lewis, the legendary musician shares music, stories and one-of-a-kind insight into one of the most important decades in rock ‘n’ roll.”
Ok then.
I choose the latest episode, 36. The cropped subtitle: “Huey highlights hits from the ‘80s that feature the…” Intrigued about who or what is being featured, I click on title/subtitle to find out more.
There is a problem. Typically, in Apple Music, clicking on the title or almost any area of the icon of an album or playlist brings you to a descriptive page of its contents. There are two exceptions. The bottom left corner of the icon show a PLAY NOW button when the mouse is hovering (and can be clicked). The bottom right corner of the icon, if clicked, shows a OPTIONS menu with a list of additional actions (add to library, share etc.)
Basically, clicking on the title of a playlist or album, or clicking on almost any point of their icons, besides the bottom left and the bottom right, results in loading another page.
Radio stations behave differently. Their icons and titles look just the same as albums and playlists, but there is no another page to be loaded. The right corner OPTIONS click works the same, but any other click on the icon or title means PLAY NOW and starts playing the station.
This a long way to explain that my click on the title does not bring me to a page describing the episode as I expected (of which I hadn’t even able to read the entire subtitle!). Instead it starts playing the episode right away, what it is to be a 60 minutes sequence of songs featuring “the saxophone”.
This is annoying because I already was in the middle of a very nice listening session and had no intention of interrupting it to listen a few seconds of Huey’s voice quickly followed by “I Want a New Drug”. I was listening something calmer, maybe the middle of Kate Bush’s “Cloudbusting”.
Unfortunately, there is no “undo” function to bring me back to that exact song at that exact point. Nor to the start of the song. Not even to that particular playlist (which was a smart playlist of all songs in my library with play count equal to zero, sorted by length and starting from the shortest).
The last play action I had done on the app was right inside the main view of my “Unplayed” smart list. There is not an option to return to this contextually relevant ‘page’.
Even worse: while I was listen “Cloudbusting” from “Unplayed”, the “playing next” function was working fine: clicking on an upper right button on the app would bring a floating sidebar on the right showing a list of all upcoming songs, all from the same playlist, in the proper order. (The sidebar can either show playing next or “history”, with the previously played songs)
Unfortunately, when I started accidentally playing the radio episode, the “playing next” sidebar stopped working. There is no way to see what’s coming next, neither in the multi-song episode, nor after its completion. And, by the way, the rest of upcoming “Unplayed” songs were swiftly and irretrievably erased from “playing next” too. Really drastic.
By the way, when any radio episode ended, the App would started playing it again, from the beginning. Even if there was a following episode available.
The typical controls when playing songs “Previous”, “Next”, Repeat” and “Shuffle” are not available in Radio mode. Only play/pause. The only way to move back and forward is by clicking on the progress bar above the song title.
When the episode plays a song, it is clearly identified on the app, and it is possible to “add to library”, “like” etc. But apparently the only way to find out the list of songs featured in the episode is by listening to it in its entirety. No access to a song list or any notes about the episode or related links.
Another problem I found is pausing the episode and trying to resume it at the same point. That was actually possible if I switched episodes within the same page, but it stopped working if I navigated to another page and then resumed the episode. Restarting the app would also erase the “save point”. It also did not sync across devices.
I have more questions:
When did Episode 36 first aired? There is no way to know.
Did the previous episodes used to arrive in regularly scheduled intervals? Also no way to know.
Are future episodes expected to arrive? In a particular schedule? Eventually? No information as well.
Also, if ever this show gets a new episode, is not possible to “follow” it and get a notification when it arrives.
All of this is a bit puzzling.
In Apple’s rhetoric, its Radio service has been a key factor of differentiation for Apple Music since its launch in 2015. That was reiterated as recently as August 2020, when its main station changed its name from “Beats 1” to “Apple Music 1” and two new live stations were launched, “Country” and “Hits”.
In that opportunity, Apple called the former Beats 1 “one of the most-listened-to radio stations in the world”. From the release:
For the past five years, if ever there was a meaningful moment in music culture, Beats 1 was there bringing human curation to the forefront and drawing in listeners with exclusive shows from some of the most innovative, respected, and beloved people in music,” said Oliver Schusser, vice president of Apple Music, Beats, and International Content. “Now, Apple Music radio provides an unparalleled global platform for artists across all genres to talk about, create, and share music with their fans, and this is just the beginning. We will continue to invest in live radio and create opportunities for listeners around the world to connect with the music they love.”
All this shows, guest hosts, artist interviews, original content and so on likely represent a meaningful expenditure.
Radio gets a very prominent placement in the Apple Music app. There are five buttons on the main bottom bar in iPhone. Besides “Listen Now” and “Browse” to left, and “Library” and “Search” to right, “Radio” is right in the middle.
All of this raises a few questions:
Why the basic user experience of using Radio can be so frustrating? At such a basic level?
Or by their internal criteria the software is performing adequately?
Are these kind of issues a concern for Apple? What prevents them from acting?
Or real world demand for the shows is so low that fixing them is far from a priority?
Will the issues ever be addressed? How long may this take?
Is the overall approach by Apple on how to develop Apple Music software features ever going to change?
The issues with Radio functionality provide a stark contrast with Spotify. Apple’s rival offers no live radio nor a roster of global DJs, but is comparable podcast functionality (including its growing and improving Song + Talk format) address all these kind of issues (sync, follow, queue, play next episode etc.) quite effectively. Playing an Apple Music Station makes me miss using Spotify!
(By the way, speaking of podcasts, the recent, momentarily tragic update to Apple Podcasts on iOS 14.5 may have raised more than a few more questions on its own.)
When Songs Meet Podcasts
Music streaming and podcasts used to be two distinct things. One had songs, the other had people talking. Radio of course has successfully combined both, but this was in another century. Now music streaming and podcasting are increasingly converging and the developments could eventually be very interesting.
As Apple Music acquisitions of Primephonic made the news earlier this week, this Music.ally bit on the classical streamer from September 2020 caught my attention:
“Primephonic is one of the streaming services focusing on classical music, with a comprehensive collection of albums and playlists. However, neither of those are its most popular format with listeners. It’s podcasts.
The company has been producing its own shows blending interviews with classical music artists, composers and conductors with their music.
“These podcasts are our most popular content. They are more popular than playlists, and more popular than normal albums,” said Primephonic’s CEO Thomas Steffens, in an online panel session organised by his company this week.
Steffens suggested that it’s been a helpful way to connect the interviewees with their audiences, at a time when classical music concerts – like the rest of the live music industry – have been shut down in many countries due to Covid-19.
“It doesn’t substitute for a live performance, but it does close the gap between remote, home listening and the live performance,” said Steffens.”
Certainly, classical music is a peculiar niche and the general shutdown of concerts may have had a disproportionate effect to the relative popularity of ways to listen to classical music, but the fact that “podcasts blending interviews with music” were more popular than playlists and albums is a bit of a surprise.
Apple Music and Spotify both offer distinct, exclusive formats mixing music with talk, but I believe their results have also been quite ‘mixed’ so far.
Apple Music has had its Radio service since launch in 2015, featuring both live and prerecorded shows. It is more of a direct replication of the conventional radio experience, with great content but a quite frustrating user experience. Perhaps the upcoming integration of Primephonic may result in interesting podcast services trickling down (up?) to the main Apple Music app.
Spotify on the other hand is developing its “Music + Talk” podcast and this is an interesting story go into detail. It begins when Music + Talk was announced on October 14th 2020, still with another name (“Shows with Music”).
“Everyone loves both a great playlist and a highly engaging conversation. Today, we’re beginning to test a new listening experience that brings together music and spoken-word content in an easy and elegant package, allowing full songs and talk commentary to live together wrapped up in one show.”
How it works
Think of your favorite drive-time radio show, that music journalist whose insights help you appreciate a band’s leap forward, a DJ whose perspective makes that next track hit perfectly.
Now imagine that you’re able to enjoy that perfect blend of music and commentary whenever and wherever you want, interactively and on demand. With shows that use this new format, listeners can interact with the music within the episodes, in the same way they interact with all other song tracks on Spotify (for example liking, saving, and reading more information about a track) without having to leave the episode page or search for it manually.
For any episode of a show with music, hit “Explore Episode” on the episode page, or tap the play bar at the bottom of the screen to pull up the episode track list. From there, you can skip around to different segments and songs and save songs for future listening.
Shows using this format are exclusive to Spotify because they rely on Spotify’s music catalog licenses and compensate musicians and songwriters just like any regular stream of a music track on Spotify. Spotify Premium listeners will hear full tracks as part of these shows, and those with the Free tier will hear 30-second music previews.”
Spotify also encouraged music artists to start their own podcasts:
“Spotify for Artists wants to help you make that deeper connection with our newest content format, an audio experience that brings together talk commentary with music. Now, you can record talk segments and integrate them with tracks from Spotify’s catalog of 65 million licensed songs to make your own shows…”
Several topics were suggested:
“Play and discuss the music that inspires you… Tell the stories behind your songs… Spotlight your collabs and collaborators… Let fans know what music you have in rotation right now… Engage with shows that feature your music… Feature shows on your artist profile… Share your most memorable backstage… Show your extracurriculars….”
Direct interactions between musicians and users are typically very constrained in Spotify. The service has historically focused on the interaction of users with artists catalogues and the universe of playlists, both functioning apart from the podcast content. Musician-hosted Talk + Music podcasts could certainly be an innovative bridge between the silos.
Music.ally provided some context on the feature:
“This isn’t the first example of a hybrid podcast/music format. In February 2019, Pandora launched its ‘stories’ feature, inviting artists, influencers and celebrities to create playlists blending their song picks with spoken-word intros and explanations. Classical streaming service Primephonic has recently been producing its own podcasts mixing interviews with musicians, composers and conductors with the music itself, too.
Further back, in 2014, Spotify worked with artist Billy Bragg on a series of ‘talking playlists’ with Bragg’s introductions to some of his favourite tracks interspersed with the songs themselves.
More recently, Spotify algo-personalised playlists like ‘Your Daily Drive’ and ‘Daily Wellness’ have mixed podcast clips with music, although these are clearly still playlists, so different from what’s being announced today.”
And some commentary on the implications:
“Another reason today’s news is significant: it could create a proper network of on-platform music influencers on Spotify. At a stroke, it’s created a platform for anyone (well, anyone using Anchor) to play radio DJ, hosting their own shows with music…
YouTube, Instagram and TikTok all have their communities of native creators – people who found an audience first on that platform, with content honed for its culture and features. Could Spotify’s new podcast format nurture an equivalent community for the streaming service – a native tribe of music podcasters?”
Still, one limitation to the service then was geographic. Only Anchor users in US, UK, Canada, Australia, New Zealand and Ireland were permitted to create M+T shows and episodes. (For listening though the episodes were available globally).
The release highlighted seven “Shows with Music” at launch, and a new hub in Browse:
All of our Spotify Original Shows can be found in the “Shows with Music” hub in Browse, or in a programmed shelf on your Home tab. Ready to experience the future of audio? Get started with one of these:
Rock This With Allison Hagendorf
Conspiracy Theories: Music Edition
How many of these shows have been successful? To my certainly limited knowledge, only Ringer’s “60 Songs That Explain the 90s” gained some traction in the general world of music Twitter, sites and newsletters.
Other than that, I can’t recall any more news about Music + Talk . That is until two weeks ago, August 18th, when there were some big news. First, some international expansion: “Our catalog of Music + Talk Spotify Originals will also get a boost across a number of markets as new shows from Argentina, Brazil, Colombia, Chile, India, Japan, and the Philippines join our library today.”
There was also a quite odd way to express to that the product had become more ‘usable’:
Since we rolled out Music + Talk last year, we’ve learned from the needs and behaviors of listeners, as we always do with new products. Savvy streamers will notice that we’ve implemented small tweaks to improve the listening experience and reflect those needs: clearer visual distinction between the music and talk segments of an episode and music previews on episode pages.
Anchor blog provides more details on the changes:
“The feature also has a new look that makes it stand out more to listeners—and with listeners able to easily identify and discover Music + Talk shows on Spotify, the format is optimized for creator success. An updated color palette, visual identity, and new shapes (plus the official name!) are all meant to increase awareness for the format across audiences, as well as give creators a clear, tangible way to talk about their show.
And whether you’re a listener or creator—or in many cases, both!—you’ll notice some improvements to the overall experience. First and foremost, creators can now connect their Spotify playlists on mobile and desktop web for easy creation and access to their favorite songs; previously, this was available only on desktop web. Listeners will see clearer distinctions between the different segments of a Music + Talk show, as well as a preview of the artists featured in the episodes on the show page. We’ve also added a helpful tutorial video to walk creators through the Music + Talk process and answer any questions that may come up.
Finally, we’re excited to launch a Spotify-branded show called “Music + Talk: Unlocked”—which will showcase five independent, innovative uses of Music + Talk, and illustrate all the creation and consumption possibilities. Hopefully, it’ll serve as all-around inspiration, too!”
The scale and scope of the issues addressed in this first revision seem substantial: making it easier for listeners to search and find those shows. Easier to distinguish shows with music from regular ones. An “official name!”. A visual identity! Easier ways to create those shows, even on mobile. (Creating a podcast demands much more work than a playlist, obviously). “Clearer distinctions between the different segments of a Music + Talk show”. “A preview of the artists featured in the episodes on the show page”. A tutorial video. A Spotify-branded podcast showing possible uses!
One can certainly say that creating a new audio format that mixes two other existing audio formats, provides great authoring tools for a large number of creators and an exciting user interface for a even larger number of listeners to find, discover, experiment and return to these shows is no small challenge!
I am really curious about what further changes, improvements and refinements will still be made to the format in the future. I bet there will a lot. More than that, I hope that many great podcasts will make use of this format eventually.
Thinking about all of this, I look for a starting point of what a great music + talk show of the future could look like.
A very recent, very good article comes to mind to use as an starting point. It basically continues the path of “60 Songs that Explained the 90” but is a bit less sprawling. And instead of being a series of 60 distinct episodes about every kind of genre, it would be one show, and a bit more personal.
The article is “The Best Songs By The Who, Ranked” by Steven Hyden at UPROXX. It features “only” 50 songs, but this is far from a listicle. At 7,500 words, this is more an essay about a monumental band by an excellent writer deeply connected to the band. The ranking scheme is a creative way to structure the writing and the fact that it provides a great playlist is just a bonus.
There is a lot of good stuff in there: great songs. A meaningful criteria for sequencing. Interesting content professionally written but shaped by a very personal, emotional connection to the music, the artists, the history and the culture around them. What more can you ask from your podcast?
What could be the length of the show chapters? A huge single session alternating the text and the songs? A series of 45-minute episodes? 5 blocks of 10 songs?
Now let’s throw another idea. Maybe evolve it into “Music + Talk + Text”?? All integrated inside the Spotify app.
You could go through the show in different modes. In the “lean in” mode, you are likely in front a PC or holding an iPad, and so text and why not images too gets prominence.
The “lean back” mode, when you are driving, walking, cooking etc and keeping the device in your pocket or using a smart speaker, has no visual component and the show sticks to to the pure audio format.
There could be some mixing of modes. You could read the entire article while listening to your preferred lo-fi beats list, and check-marking some songs to listen on gym session
There could be two versions of the “talk” part: full sections and abridged versions, to be served depending on the context as well. If you are in the mood for music you get/choose the short talk. In the mood for stories, get/choose the long version of the talk.
Going further, what if the “lean in” mode also had videoclips??? Either featuring whole song performances, or historical footage or even talking heads.
That’s quite a tall challenge, but the same show could be a podcast, a playlist, an audiobook, an article and even a documentary.
No need to read the text on the browser, see performances on YouTube, watch the doc on Netflix and listen to the podcast on Spotify.
Of course this mix of songs, text, talk and video looks like those multimedia CD-ROMs of the early 90s, but the key difference is that the music remains the anchor. Our The Who multimedia stravaganza is happening right on top of the user’s music library and along her listening history. The music is provided with an abundance of context, but all this auxiliary context is shaped towards the user attention, not against it. The experience can be truly immersive instead of all-over-the place as currently is.
And this super “Song +” format could serve for so many (most?) music topics. For example, another great, massive article by Steven Hyden that I recall promptly: the greatness of Queen, as explained by their discography, album by album in chronological order, as a detailed riposte to the controversial biopic. Imagine navigating all the albums by Queen guided by this perspective, focusing on either highlighted songs from each album or in their entirety.
I would definitively read/watch/listen “All 213 Beatles Songs, Ranked From Worst to Best”, by Bill Wyman on Song +.
Books could become Song + shows too. The whole 33 1/3 Series. A highly interactive “1001 Albums You must Hear Before You Die” (you may have read this before). Bob Stanley’s great Yeah Yeah Yeah: The Story of Modern Pop.
This is all very far-fetched, but doesn’t also seem an exciting possibility?
About Curation
Spotify users have created over 4 billion playlists. Most of them are not that great. But there sure are a few great ones. Finding them can be hard. I found one of them and want to write about this experience.
A key point is that this particular playlist is awesome very specifically for one particular listener – me in this case. Many others listeners will certainly not share my enthusiasm for that particular mix of pop, rock, “indie” etc.
Indeed, I am one of a total of less than 500 Spotify users who have “liked” that list on the platform. The playlist’s creator maintains other lists, and several are more popular.
Finding a great list can be harder than finding a needle in the haystack. Every user is looking for a different needle. What might be a needle for Alice will look like haystack for Bob and Carol and Daniel.
Recommending music for friends and relatives can be hard. Preferences vary wildly. What is great for Elias may sound awful for Fred and indifferent for Georgina. Hubert cares a lot about rhythm, Ingo is more about melody and Jacob focus on lyrics. And so on.
So what should be an awesome list? A lot of music that is really enjoyable, obviously. A majority of very good songs. A good amount of great songs. A significant amount of amazing songs. A minority of songs that are just meh. A very small number of songs that are annoying. Also important, most of the very good/great/awesome stuff should be new to the listener.
Isn’t that an obvious description for a great list? But this is exactly what I encountered. A playlist that has over 6,000 songs. The most recent song was added today, the first was added a decade ago. There is a lot of stuff in there that is very familiar, but not a majority. Many familiar bands, but also surprises. It contains actually thousands of good/great/awesome songs that I did not know. Really few annoying songs.
It is really a great find, a huge source of great music that is exactly the kind of music that I would like to listen and also recommend to other people.
And more importantly, I remain eager to listen to the next song of the list because I am inclined to trust the decision processes, tastes and sensibilities that drove the inclusion of every one of them. I found this list via twitter, by chance.
In my opinion, the challenge for Spotify is to multiply this kind of experience. Encourage curators to express their distinctive tastes and sensibilities. Help users find those that relate to these particular perspectives. Create the tools to support and improve all this activity.
“Curation” is a word that has been so abused online in the last decade but – done right and truthfully – may be the best way to serve the best music.
P.S. Spotify has released “Blend” today, which seems a very intriguing feature “for two users to merge their musical tastes into one shared playlist made just for them, making it even easier for users to connect, discover, and bond over the music they love with one another”. But this seems a right step at the end of the problem: the first challenge is finding the right other users. Spotify mentions blending tastes with “a friend or loved one”. This remind the old phrase “Twitter makes me like people I’ve never met and Facebook makes me hate people I know in real life”. Blend as described feels more like Facebook than Twitter.
But this other detail is very intriguing: “now, the experience includes… taste match scores to see your listening preferences compared to your friends’”. Maybe those ‘taste match scores’ will eventually serve as a foundation for the ultimate search tool?
The HiFi Saga
This story is a bit odd, but someone should write it down for one reason or another. It begins when Spotify announced “Spotify HiFi” on February 22, 2021:
“High-quality music streaming is consistently one of the most requested new features by our users.”, they wrote – and I agree. In my opinion, Spotify’s highest setting (Ogg Vorbis 320 kbs) sounds much worse than Apple Music typical AAC 256 kbps.
“Spotify HiFi will begin rolling out in select markets later this year, and we will have more details to share soon.”
“Premium subscribers in select markets will be able to upgrade their sound quality to Spotify HiFi and listen to their favorite songs the way artists intended.” Evidently, HiFi is meant to add a few dollars to the subscription price.
I guess it makes sense. The monthly price of US$ 10 was inexpensive when Spotify launched in U.S. in 2011 and it’s even cheaper now, adjusting for inflation. Offering music in a “CD-quality, lossless audio format” would be a natural option to experiment with more pricing tiers.
To promote the future product, there even was a video with Billie Eilish and FINNEAS promoting the idea as they “speak about the importance of high-quality audio options for creators and fans. “
Spotify of course is the leading music streaming service in the world, with a 32% market share.
What did Apple Music, the tech giant and number two music streamer in the world? Basically everything that could mess with Spotify plan!
(Of course, the “feud” between the companies has been so long and well documented that Spotify maintains an official timeline of its complaints and the remedial actions it has been pursuing.)
Apple announcement came on May 17, 2021. “Apple Music announces Spatial Audio with Dolby Atmos; will bring Lossless Audio to entire catalog”.
“The next generation of sound on Apple Music is coming to subscribers June 2021 at no additional cost”.
It is really remarkable how Apple reaction was a point-by-point “rebuttal” of Spotify’s:
- Instead of coming “later this year” and any details “coming soon”, Apple said June.
- Instead of “being able to upgrade”, Apple would provide it “at no additional cost”.
- Instead of “select markets”, Apple basically just went global.
- Instead of “CD quality lossless” (aka 44.1 kHz 16 bits), Apple did the same PLUS high definition lossless (up to 192 kHz 24 bits) PLUS Spatial Audio (Dolby Atmos).
And Apple apparent rebuttal of HiFi went even further. Apple launched lossless and Spatial Audio on June 7th and on the next day Billboard published and interview with Eddy Cue, Apple’s Senior VP of internet software and services, where he not only hyped Spatial Audio as a “game changer”, the future of audio but also expressly stated that lossless [as championed by Spotify] is NOT the future.
Q: The music industry, for a few years now, has been talking about lossless. It has said it will be the next-gen technology that’s going to change everything. And you seem to think Spatial Audio is going to be that. Was that initial thinking around lossless incorrect?
A: There’s no question it’s not going to be lossless. Because the reality of lossless is: if you take a 100 people and you take a stereo song in lossless and you take a song that’s been in Apple Music that’s compressed, I don’t know if it’s 99 or 98 can’t tell the difference.
For the difference of lossless, our ears aren’t that good. Yeah, there are a set of people who have these incredible ears, and that’s one piece of it. There’s the other piece of it, which is do you have the level of equipment that can really tell the difference? It requires very, very high-quality stereo equipment. What you find is, for somebody who’s a true, for example classical connoisseur, they may be able to tell the difference in lossless. I can’t tell personally — I do the blind tests all the time with the team — I can’t tell. That’s a problem. That’s not going to work because that’s a marketing play, not a true customer play. Dolby Atmos, Spatial Audio? You can tell. I can tell, everyone can tell. That’s going to make all the difference in the world.
Now, we’re supporting lossless and we think there’s a set of customers. It’s a small set of customers, but they want it and we’ll certainly give it to them, and they’ll have it as part of this. The good news is they’ll have lossless and they’ll have Dolby Atmos and Spatial. It really does work very well for that [set of customers], but it’s not going to be lossless [leading the way].”
Of course, Apple’s AAC 256 Kbs compressed format has long been a high-quality audio format, and the debate whether most/any listeners can really distinguish between lossy and lossless audio has been endless, but to hear an Apple senior executive dismiss lossless and high-definition lossless as something almost irrelevant that is demanded by a “small set of consumers” is quite striking.
(In my subjective opinion, lossless sound much better than compressed on any decent entry-level headphone, high-definition lossless may be usually overkill indeed, and spatial audio on AirPods and such is basically a gimmick.)
Anyway, Apple approach of providing lossless for no additional cost had another major impact in the competitive landscape. Amazon, the third largest music streamer, already offered lossless and HD lossless for Amazon Music Unlimited subscribers since September 2019, in a few markets. Amazon priced lossless quite competitively, for only 5$ additional dollars, half the premium charged by lossless pioneers Tidal and Qobuz. Upon Apple’s announcement of lossless at no extra cost, Amazon announced it would to do the same immediately, bringing down the price of Music Unlimited HD down to US$10.
(Curiously, Amazon has shown no rush to extend the international reach of HD. The service remains available only in US, UK, Germany, Canada, France, Italy and Spain. No announcements of any expansion have been made.)
Anyway, one can guess that Spotify’s plan to offer HiFi as an upgrade has been seriously challenged.
Meanwhile, Spotify has been conspicuously silent about those developments. As I write, more than six months have passed since the HiFi announcement, and no further details have ever released.
The only official words about Spotify HiFi come from @SpotifyCares, and they are always the same when inquired about the subject:
“Hey there! All information about Spotify’s plans for HiFi can be found on our blog, For the Record: https://bit.ly/3joOJd0. We don’t have any additional details to share at this time, but give us a shout for anything else.”
(Some details about HiFi have been ‘leaked‘).
One last detail about this odd brawl between music streaming services is that Spotify explicitly emphasized it would use integrated HiFi with Spotify Connect. “Ubiquity is at the core of everything we do at Spotify, and we’re working with some of the world’s biggest speaker manufacturers to make Spotify HiFi accessible to as many fans as possible through Spotify Connect. “
Apple on the other hand has made no much effort so far to bring lossless to the users’ audio setups. On the FAQ about how to listen to lossless on your iPhone or iPad [or Mac] it is simply stated: “To listen to songs at sample rates higher than 48 kHz, you need an external digital-to-analog converter.”
It is really astonishing to see Apple communicating this way, instead of using the launch of lossless to release some innovative DAC of their own, or maybe some kind of AirPlay upgrade. Instead, Apple is clearly positioning this feature for that “small set of users” who may have some idea and care about “external digital-to-analog converters”.
John Darko explains well why Apple Music’s hi-res streaming is a “bit of a mess” (the video at the end is surprisingly funny, considering the specificity of the topic).
(Meanwhile, I am curious how long the hype around Spatial Audio for music on headphones will still last and whether Apple will eventually regret/feel embarrassed about its role on it…)
Anyway, there still one third of year left to Spotify release its HiFi product and bring a new chapter this to this odd Cold War.
P.S. As I was finishing this article comes the news that Apple has bought classical music streamer Primephonic for an undisclosed price.
“With the addition of Primephonic, Apple Music subscribers will get a significantly improved classical music experience beginning with Primephonic playlists and exclusive audio content. In the coming months, Apple Music Classical fans will get a dedicated experience with the best features of Primephonic, including better browsing and search capabilities by composer and by repertoire, detailed displays of classical music metadata, plus new features and benefits.
“We love and have a deep respect for classical music, and Primephonic has become a fan favorite for classical enthusiasts,” said Oliver Schusser, Apple’s vice president of Apple Music and Beats. “Together, we’re bringing great new classical features to Apple Music, and in the near future, we’ll deliver a dedicated classical experience that will truly be the best in the world.”
(This is really amazing. Finally we are going to see some specialized apps instead of the current one-size-fits-all approach?!)
P.P.S. The Apple press release on the acquisitions quotes Oliver Schusser, Apple’s vice president of Apple Music and Beats, not his direct boss, Eddy Cue. The lossless press release was also 100% Schusser and and zero Cue. So why on Earth was Eddy Cue talking with Billboard two weeks later about ‘lossless not being the future’? Why does Schusser not even mentioned during the interview? Apple is usually always so disciplined in its dealings with the press, this Billboard interview is really an outlier.
Constant Revolutions: The ‘Plasticity’ of Music
The impact of technological changes on how music is created and consumed is a well covered subject.
Music history is full of dramatic changes: electric microphones, stereo LPs, transistor radios, multitrack studios, FM stations, arena-filling sound amplification, the walkman, digital synthesizers, samplers, CDs, MTV, SSL consoles, SoundScan, the iPod, Spotify, bluetooth earphones, voice-activated smart speakers, AirPods etc.
In important ways, they all changed everything: the type and variety of music being made, the type of musician, the age/socioeconomic profile of listeners, the amount and frequency of music consumed, the places where it was consumed.
One particular detail that greatly amplifies the power of technological changes to alter how music is created and consumed has to with the plasticity of music itself.
‘Plasticity’ is the quality of ‘being easily shaped or molded’ and music listening has lots of it. People enjoy music in many, many different ways: How much they spent listening. How often. In how many different contexts. Alone or communally. The emotional responde to a particular genre can be particularly intense. The memories attached to a particular song can be overwhelming. Opinions about what is good or bad music are incredibly subjective. Opinions about what is good sounding music as well. People have different preferences in even in music volume! And so on.
That means that music demand has tremendous flexibility. And that’s why technological innovations can have a very disproportionate and unpredictable impact.
In other words, one can almost say that the music landscape has lived in a constant flow of revolutions. Everything is going fine until a new element or gadget comes along and changes the way music is made/marketed/listened. And it is always unexpected. We never knew we needed Walkmans or iPods until we did. And then in hindsight they all look obvious and “inevitable”: of course the miniaturization of electronic components, together with digital compression algorithms and powerful song library management desktop software (offering also a fast and seamless synchronization feature) would naturally result in a stylish rounded metallic device holding a thousands songs that fits in your pocket!
But the obviousness only comes with hindsight. The actual revolutionary innovation had to bring something new. And “new” can take many forms in this music realm. Anything that explores, illuminates, tweaks or solves one or more of variable dimensions of music interaction can be a “killer app”. The new idea can change how often we listen or how long. How portable. How instantaneously. How cheaply. How easy it is to search a song. Or remember it. How easy it makes to hide our taste, or share it. How it increases the range of where to listen or with whom etc. etc.
And there is so much room for surprises because our connection with music is very deep, broad and variable. We have a strong emotional connection with music. In fact, music is more than entertainment, culture, distraction or escapism: it is an input that our minds absorb and use. Music is brain food. The plasticity of music is sourced on the complexity and neuroplasticity of the brain itself.
Like the brain, always rewiring itself as it grows and heals, our relationship with music is always rewiring itself. That’s why changes in technology, hardware, software or interface can have such profound consequences on how we consume music.
I can speculate – what is going to be the next revolution in music: maybe those bone conduction headphones? Or future iterations of those speaker-featuring glasses? 360 audio as Sony dearly hopes? Perhaps AirPods Pro Max Series 5 will work without a watch or phone nearby, and the transparency mode will be so good that you can wear them the whole day? (And Siri evolves in a functional digital companion?) Maybe the audio part will the killer app for AR glasses?
Or going into software, the future of music will be intrinsically linked with TikTok and Metaverses? Will AI evolve so far that it can accurately divine our taste? When you go to a party, will your Spotify account register all the songs that you listened (and take note of which made your heartbeat/blood pressure /etc. go up??) Maybe someday music streaming will be fully integrated with your social graph? (2030’s iTunes Ping will be awesome??)
Another interesting point is the suspicion that we may have crossed an important threshold in this technological evolution of the music-brain interface: music streaming platforms, with full catalogues served to a global audience, may look in the future like a primitive central nervous system, the ancestor of something much more impressive.
Because, unlike other art forms or fields of knowledge, music streaming services are more than databases that index and describe whatever exists in the real world: they are becoming de facto main music layer we interact with. And all future improvements can be built upon this digital base.
Nowadays, it is quite feasible to describe music streaming services as low-margin services with limited differentiation. Spotify still struggles to ensure long-term profitability and most of its competitors are tech giants whose music services represent a small fraction of their total revenues. There are several ongoing discussions about user-centric streams models, and how creators are paid, what role exactly each of several ‘stakeholders’ should have etc.
But that’s fine. The evolution of this early digital musical brain we have now is going to be full of surprises and turns. What is more certain is that technology has always powered music plasticity, and the current state of digitalization might very well be an inflection point from which the pace and range of innovation will be increasingly exciting.
And of course, once those new fantastic developments eventually arrive, we will later look back and think that they were all obvious and inevitable.
April 12 edit: Changed the structure of the first paragraphs for clarity.
Waiting for HiFi
I have been really enjoying the improvements to user interface/experience of Spotify apps. I am also optimistic about the accelerating pace of improvements:
“The future of Desktop – As mentioned above, this change to the Desktop UI gives us the ability to move faster in bringing you new improvements, features and functionality — so you can expect to see continued improvements to the client in the weeks and months to come.”
The problem is that the sound quality remains underwhelming. Even at the highest resolution (320 Kbs) the sound seems to me markedly inferior to what I get from Apple Music 256 Kbs AAC files.
Qobuz lossless files sound even better, but the apps are much less practical for daily use and less conductive to discovery. Tidal is fine but sits in a weird spot: sounds maybe a little less great than Qobuz, but the app is less useful than Apple/Spotify. Amazon HD is not available in my country.
So now all I have left to do is wait for Spotify HiFi. The problems are that it is coming only “later this year” (and we are still in April) and in “select markets” (meanwhile Spotify’s recent track-record of expanding new features internationally has been less than stellar).
They promised that they “will have more details to share soon” but this was 44 days ago! (What they understand by “soon”?)
Finally, we are left to wonder if Spotify lossless implementation will sound great or less-than great, such as Deezer’s?
Until Spotify HiFi is actually launched and evaluated we have the right to remain skeptical. Apple Music released a superior sounding product nearly six years ago and during all these period Spotify did little to close the gap. (Actually there was some improvement in loudness levels in 2017, but this was obviously not enough to sound competitive against 256 AAC and lossless FLAC).
(Until a few years ago the highest quality setting at Spotify desktop was called “extreme” instead of “very high” as it is now. That was funny.)
But of course, it would make very little sense to Spotify release a HiFi product and be careless about implementation, even more so since they are focusing on “CD level” quality and wisely not entering the distracting battles of 192-96 kHz/24 bit/MQA/360 features…
There is also the small question of how Apple will respond to this initiative. It has a long tradition of not only supporting lossless and hi-resolution music but also encouraging labels to provide high quality original files (Apple Digital Masters). On the other hand Apple insists that its 256 Kbs AAC files “are virtually indistinguishable from the original 24 bit studio masters”, so the launch of different audio tier would be a drastic change of messaging.
Let’s wait…
Streaming Machinery 