SoundCloud Global Limited & Co. KG is not a social enterprise but it is still (and despite everything) the platform I use by default to find new music, and it’s all because its social features:
Follow users, artists or not.
Like and comment songs
Emoji-react and comment in specific points of a song
Repost songs, with optional comment.
As a result of these ~1-click and mobile-first features, users get:
A beautiful feed with music posted or reposted by the people you follow. Press play and let it sound, skipping to next/previous whenever you want.
A gorgeous player visualizing the users, the comments and the emoji reactions in real-time.
All this is very engaging for listeners – also for artists. The experience is very social, and keeps me from listening more music on the Fediverse and the fair platforms. Also on Bandcamp and the spotigafam platforms.
This is an invitation to discuss how the fair and truly social platforms could offer such an experience to social music lovers. What we’ve got, what’s in progress, what we miss, and (if anything) what’s not gonna happen because [reasons].
There is an abandoned project that had a goal to replicate most of SoundCloud features as a fediverse platform:
SoundCloud equivalent was/is very often asked for by musicians.
But to a degree I feel likenso much of that is beginning to be offered by bandwagon.fm on one side and peertube on the other. functionality is pretty much there, just interface is different.
maybe it makes sense to sepiasearch music hashtags.-
I think that replicating (and then improving on) familiar interfaces should be high priority for federated platforms – see e.g. Principle of least astonishment - Wikipedia.
Personally, I’ve been looking for the federated open-source BeatStars Interface-wise, it’s not very far from SoundCloud, either – just a bit more marketplace-shaped.
I think replicating is a bad idea. You won’t get lots of people to leave a corporate platform by providing a clone of what they already have, just without all the other people. The network effect makes most of them stay where they are. Going in another direction and trying something new is a better strategy. I don’t think bandwagon.fm is going to create a mass exodus from SoundCloud any time soon, but a SoundCloud clone, with less users and features, is definitely not going to do it. Let’s focus on what makes bandwagon.fm (and others) good, not which corporate platforms they can replace.
I’m not advocating for replication either. Except for the characteristic Soundcloud player, the other features I mention above are generic.
I’m curious what @bandwagon and @simon etc have in mind for the longer term. So far, Bandwagon and Mirlo are building on the same paradigm that Bandcamp was built on: a catalog of artist > album > song + merch, adding concerts too. This is the paradigm of a traditional music shop, but online, with the possibility of having name-your-price / give away music, and the possibility of getting user reviews and follows – to support the promotion and sales of the artists’ new releases.
Soundcloud starts with a different paradigm. It’s more of a social network in nature, that puts at the center songs instead of short texts, pictures, videos. etc. Artists publish songs as units. “Albums” are just a playlist, an afterthought. The goal is to get your music heard, not sold, and (those who care) get followers, a fanbase. There are no sales, just optional subscriptions (that, in principle, contribute some income for artists based on plays).
While people are using Spotify basically as a smart radio with playlists at the core, Soundcloud is basically a social radio where expressing your feedback as a listener and receiving it as an artist is basically the reason to be there.
This is what I miss in the Fediverse, where the architecture is 100% social and grassroots. A social network of music makers and music listeners where publishing, listening, discovering, sharing, following and commenting is easy and fun. I’m not sure whether this is the architecture and objectives Bandwagon, Mirlo, Faircamp, Funkwhale have in mind, or whether someone else is trying to build such social network with this design from the start.
Independent platforms have less resources than major corporate actors, but also less overhead. Also, a tighter feedback loop – since they’re much more likely to listen first and foremost to the actual people involved, rather than to stockholders and LinkedIn fads.
So, anything “they” can do, I believe “we” can do better.
Except, as you say, for that damned network effect! It’s something that has been established through tremendous capital infusions during an era of unabashed technooptimism; so I agree that trying to follow the exact same playbook as Soundcloud is unlikely to lead anywhere. I’m certainly not advocating for anything like that.
All I’m saying is, if people consider Soundcloud’s player to be a good feature, then it can absolutely be made into a generic component – and then improved on, perhaps in ways that its corporate originator has no incentive to pursue.
Maybe someone even already laid the groundwork for that – even if it’s some junior frontend dev just trying to build something cool to show prospective employers. People do, after all, regularly build whole alternative clients for proprietary platforms – sometimes entirely out of UX considerations (rather than e.g. to improve privacy). In that case it would be a matter of discovering it and figuring out where it fits in the overall picture.
(And again I am reminded of the notion to add federated comments to some open source music player. I find myself severely limited by time, privacy model, and the tech stacks I’m willing to work with these days – but if someone took that idea and ran with it, it would definitely make me happy to see the result.)
So, I did some lookin’, and here’s one possible basis for a Web player with Fedi comments:
It looks amazing, it’s funded and profesionally maintained, and has a healthy community behind it on GitHub.
Can I just take a moment to say I would be madly in love with any solution which opts by default to display the music as beautiful, detailed, crisp, informativespectrograms. Now that’s something I don’t think Soundcloud would ever dare!
Not just plain old waveforms (although I absolutely love the way Faircamp styles those! ), and certainly not that “DJ view” which is made of multicolored overlaid waveforms.
Spectrograms can be quite attention-grabbing, owing to their high information density – and, if one so wishes, one could navigate by them note-for-note
And here’s one:
Plus, something you can do with spectrograms is theme them to match the album art! Already made possible by the colorMap option here (helpfully in RGBA, to blend nicely into the page background )
And here’s the example for custom annotations:
Looks a little basic, but it’s just a little demo after all. Nothing a little CSS can’t fix!
The way I see it, the only missing piece is a canonical format for specifying the timestamp in the ActivityPub object… huh.
Could we just do what YouTube does and linkify all “timestamp-shaped” strings in the Note text? That’d leave us with a PoC which could be demonstrated in favor of extending the protocol – unless, again, there’s already some type of TemporalAnnotation object I haven’t yet seen!
Note: We’re working on a slightly more thought through version of what I’m about to write, and it’s been hard for me to keep up with how quickly these threads move and track the tangents (you can fork threads by the way, and reply in different threads if you want to go off topic).
When we started building Mirlo, there wasn’t anyone else out there doing the same thing! We were involved in Resonate and Ampled, and were fairly familiar with the landscapes at the time. Then Bandcamp sold a second time and suddenly we felt a need to move quickly to capitalize on that moment, and at the same time Ampwall was announced, then we found out about Jam.coop, Tone.audio, Faircamp, and then in October Austin announced Subvert (which now has 5k members, btw). So now it’s actually a fairly crowded field, and there’s a bit of overlap! One of the reasons we personally went with a bandcamp clone–our decision preceded the sales and union busting–was that bandcamp is a profitable business (unlike Spotify and other streaming apps), so we saw it as possible to carve out a little space where we could actually build an income from and support the work on the project. Originally we actually wanted to pivot Resonate in that direction, and even passed a vote to do so, but there were interpersonal / political / burn out things that basically blocked that (underscoring to us the importance of figuring out governance structure, but also financial support).
However, we’ve been thinking a lot about whether “a global marketplace where individual musicians compete with each other for very little dollars from their fans” is actually the direction we think the music industry we’d like to see going in. I personally struggle with the sense of ownership and sunk cost I feel with the current Mirlo platform, and whether it makes sense to actually continue building it out with so many alternatives in place, especially as no one is getting paid to do this work (I’m spending a lot of personal money doing this work).
But so I think there’s some questions of what it is we do want to build! Is a more social space like soundcloud something we want? What can we tear out of the existing code bases to do new things with? How can we build on other open software (eg. Open Collective) to build musician collectives? How can those things we build be sustainable for the people working on them?
To get back to the topic–something we’ve been tinkering with the idea of–we have a person who’s been in their free time building a mobile app for Mirlo–is that a mobile first experience could be more social based, basically using a different set of databases to store interactions, but to send money to musicians based on a you choose your monthly subscription (user centric payout model).
This idea you mention sounds promising. Right now, listening to music in social platforms takes…
many clicks to find and listen to music
many clicks to like what you listen to in a way that you can find it later and listen to it again
many clicks to find and follow artists
basically impossible to get out of the way artists and music you know you don’t like
and… a lot of luck finding music you like (part of this is a matter or genres, part of this is of quality)
This mobile first app could offer…
a feed of recent music based on previous interactions (no need for complex algorithms)
like songs, and keep a playlist of your liked songs
follow artists AND other people reposting music and creating playlists
ignore artists (sorry, but this means more time and attention for new music/artists and the ones you know you like)
make it easy to comment and to federate those comments together with the music & artist links and metadata.
Yes! I don’t know about others, but I spend more money on Patreon kind of subscriptions than buying actual albums. Even when I buy something on Bandcamp I do it more with a “patreon-like” spirit than because I want to actually download the high quality files and have unlimited plays.
But what about this (hard to count before computers were invented, maybe easy to count nowadays?)
From your monthly subscription to the platform…
a % goes to the maintainers of the platform, to pay the bills
a % is distributed among the artists you tip, just for existing
a % is distributed among your # plays
a % is distributed among your # likes
As a user, you can decide the distribution of these percentages – as long as they add to 100% of course.
Users still can use it for free. Their plays and likes would count like the rest, just bringing $0 income to the artists and the platform.
(I’m half-aware of the accounting complexity this model could have, but let’s say that artists would get paid only when they accumulate $10 USD or whatever. Meanwhile, tips, plays and like would keep counting in your dashboard.)
I would pay a subscription for an app like this, bringing me music I enjoy, curated with peers, and compensated transparently with my subscription fees.
I’ve been working on a project to address this space for quite some time now, and hopefully will have something public coming months (I’m a bit of a perfectionist unfortunately).
My focus is on the electronic music / dance / rave / club scene, and that has particular dynamics. For example – most of the music that gets consumed is in the form of long-format DJ mixes or live recordings. Which does not fit in the more release-based paradigm. And it’s also often used to promote either music sold on other platforms, or to promote events.
So, basically, I’m building something that will initially work somewhat like SoundCloud (clean feed, profiles, waveform visualization, social features), but a bit more customizable per instance or actor level, and enables DJs to add timestamped track credits (something SoundCloud is missing).
This is a bit of a drive-by comment to make sure that I get back to this thread when I have more time to read through what y’all have been talking about and add more detailed thoughts
What an interesting project! I wonder whether the design will be “hardcoded” for these long-format DJ mixes or it can also be used for single songs of a few minutes as well. And of course it would be great if it would have “plugs” to the fediversal music platforms (if anything is needed beyond plain ActivityPub).
(Side comment: my feature request number 1 to Soundcloud is a way to filter out from my feed exactly those long-format mixes & radio programs. I have stopped following musicians I like because most of the times they report those long mixes that I avoid.)
That’s the beauty of ActivityPub. There’s no “normative” way to do anything. You just stuff whatever data into whatever JSON-LD you want, and pray that every other app out there has implemented a way to read it.
Honestly, there are already several different ways to make emoji reactions, so what’s one more?
Snark aside, we could certainly attach a timestamp to a Like or EmojiReact. Apps that understand it could display the additional data, and those that don’t would just display a simple Like.
ActivityPub-enabled platforms could rely on federation for this. WordPress’ ActivityPub plugin (from all places!) shows how this could look like on the platform where an album or a song is published:
An appendix like this one would look very good in your album page, right?
Related, displaying Fediverse followers on the artist page would also improve the social component of music platforms (the WordPress plugin provides this feature too). Not only looks good in your artist page, but it also helps your fans to find and follow each other.
Really interested in this! Ektoplasm was a very interesting project in this general space (although didn’t have the social aspects), and I had some interesting discussions about a decade ago with some promoters and a label on potential business models could work, so this is simlarly a drive-by comment to follow up at some point
I’m more and more inclined to think that we’d be better off educating than building. UI/UX has come to be at the expense of knowledge.
We are where we are because convenience is the competing upper hand. This has led us to a dynamic that is intrinsically dependent on the assumption that no one wants to learn and everyone is lazy.
I think collectivism is a good way to go. However i’m not sure it works if the collectivism is a global singular village, I believe the shape of mainstream social media is evidence of that. So we need many small size collectives. Just like we used to have many small shops on the streets instead of giant super malls.
Basically atomization. Can that appen without spreading the knowledge that empowers communities to build themselves and own the fruit of their labour?
The tools and protocols for social interaction are here. If we find a way to render our peers curious to try and experiment with how they communicate with their digital world, we might not have to all use them the same way.
@anon15453463 How would you apply these principles to software that would help music lovers to share, discuss and support the music and the artist they love?
My point is that I would not apply these principles on software, rather on people.
Opening sources is not enough, musicians and listeners need to be empowered to appropriate themselves the tools, so that they may wield them to their needs. We will never know everyone’s need. And I don’t think good UX is enough, because UX is abstraction of underlying functioning.
Documentation, mutual aid, transfer of knowledge is far more powerful and sustainable than any given feature. With the right knowledge all features exist already.
I don’t think you can force anyone into learning though, but I uppose to answer your question in the context of what I mean: each one teach two.
Interface-wise, it’s not very far from SoundCloud, either – just a bit more marketplace-shaped.
), and certainly not that “DJ view” which is made of multicolored overlaid waveforms.
)
