Eric Wahlforss is one of the guys behind SoundCloud. This is where he jots down thoughts on the web, music and strategy, among other things.
More About Eric »
Ligeti’s maximal minimalism - Visuals by Kubrick
“Minimal”
Musica Ricercata (1951-53): “The first uses almost exclusively just one pitch class, A, heard in multiple octaves. Only at the very end of the piece is a second note, D, heard.”
“Maximal”
Atmosphères (1961): - “It opens with what must be one of the largest cluster chords ever written - every note in the chromatic scale over a range of five octaves is played at once. Out of the fifty-six string players ushering in the first chord, not one plays the same note.”
“Fluxus”
Poème Symphonique (1962): “Each of the hundred metronomes is set up on the performance platform, and they are all then wound to their maximum extent and set to different speeds. Once they are all fully wound they are all started as simultaneously as possible. The performers then leave.”
AppEngine is the new HyperCard
“It feels like the web has been trying to claw its way back to the simple utility of Hypercard ever since Mosaic. GeoCities was the first massive-uptake anyone-can-build-here website haven. But it was all static html…” Rich Skrenta: AppEngine – Web HyperCard, finally
HyperCard is how I learned scripting. Much has happened since, but I will never forget how easy it was for me as an 8-year old to get started with that brilliant tool. The web, and even AppEngine, still has ways to go.
Open Wins, note on App Engine
With recent cool announcements from Google (Amazon? Nah, too expensive!), I hear a lot of people bursting out in great enthusiasm “Ah, fab, let’s move to App Engine!”. But I think it’s always good to step back and think twice… even though it is *darn cool stuff*. Here’s something to keep in mind:
“The bottom line is that if you build your application on App Engine, Force or EC2/S3, you are locked into those platforms. Moving off will require a substantial re-engineering effort.”
Although Joyent aren’t exactly neutral in this matter, they do have a point. Here’s the full post.
Good comments on recent Apple music rumors and the Bebo deal
“… However, the bigger question is: is the money in the DEVICE, or is it in the NETWORK, or both? My take is that it’s the network, and not the device. But let’s see…„
By Gerd Leonhard, here’s the full post. And here’s the original CNET article.
“Who is going to build the infrastructure the artists and the web services need? Who is going to deploy it at the micro-scale that most mp3 blogging happens? The music industry is all about demanding to get paid, but I don’t see them building the systems to make it happen easily and within the constraints of what an online business model can pay.„
By Fred Wilson, here’s the full post. Original Op-ed in NY Times [Thanks Raja]
Cloudy mother of a jukebox! Dreaming of my future music player.
With recent iTunes rumors floating around, here’s what I hope my future music player will be like:
- It will look very much like a mix between Songbird, SoundCloud, Last.fm and Spotify.
- It’s not really a piece of software, rather it’s a service. It stores playlists, music and social graph data in the cloud, with open interfaces so that my friends and their players can know about my music.
- I can drag and drop any song link on it. If the link is, say, a SoundCloud song url, the player sniffs the hAudio on that page and caches an mp3-stream locally. It loads my friends’ comments on the song as well. I can surf the web much like in Songbird and save links (=songs) in my playlists as I encounter them.
- It will have an oversized local cache, that I don’t have to think about. There is no explicit handling of files. The only thing my player handles are links of all kinds. Links to songs or collections of songs (It does not have to be explicit links to files however). I can take any of these links and pass on to my friends.
- There will be an open service (or several, ..and soon!) with unique urls to every song in the world (e.g. http://songs.ru/USPR37300012), that serve metadata and pointers to where files can be found (e.g. http://soundcloud.com/forss/soulhack.mp3). I can drag any song link to a playlist–my player identifies the song and tries to fetch it using the above-mentioned service and/or other means (it could (even) ask me to buy it). If it can’t find the actual song file it still saves a pointer and metadata for it. That way I can keep and share my playlists/bookmarks without having to care if I have the actual file or not.
- My player keeps my cache in good condition. It periodically checks for broken links, etc.
- There’s no “sync to mobile device”. Rather, the mobile device has its own interface to the cloud service and handles caching/streaming by itself. My playlists can be accessed from anywhere and are updated instantly.
- I can browse and play my library from my linux-based uPnP device as well, from whereever.
- It can do things like show a smart playlist with an activity stream style feed of what my friends have been listening to.
- With it I can build a song library over time that goes way back, much like del.icio.us does for links. My library can grow over 5, or 20 years. It’s just a collection of pointers. In the future, when better audio formats are available, my player silently refreshes its cache with FLAC files, or 5.1 24 bit surround files.
- It differs from iTunes or Spotify in the sense that it is open. Open.
- It… may never exist. But there are some signs of hope.
In a post-scarcity publishing world, the key is to own the most relevant copy
The title of this post sounds a bit cryptic, agreed. Let me try to explain what I mean. The point I’m trying to make is actually very simple. The Web is a giant copying machine. And yet, if people can avoid having to copy something, they will. The problem is that today, the music industry suffers quite a bit from illegal file sharing–a giant copy party. What is going to happen over the coming years is that this copy fest will wind down. Yes, it will! And the reason for that is that there will be services that let people listen to their music without having to copy files and manage them.
Ok, so what matters in a world where p2p is irrelevant? In this world we will instead share and discover music in a giant link-passing frenzy. This is already becoming a reality, only it’s “not evenly distributed yet”.
Here are a number of popular links to Flickermood, a song I released under the alias Forss on Berlin-based Sonar Kollektiv:
- sonarkollektiv.com/tracks/DE-P96-03-00042
- last.fm/music/Forss/_/Flickermood
- forss.imeem.com/music/q9s1RHX1/forss_flickermood
- mp3web.org/dwnl.asp?id=87092
- open.spotify.com/track/6nFNSuVrBNqZ2Qe2TwZZ4G
…oh, here’s a link where you can buy the song, complete with drm and in worse quality. http://phobos.apple.com/WebObjects/MZStore.woa/wa/viewAlbum?i=7018176&id=7018192&s=143456. Feel free to go there and, ..eh nevermind.
And then we have soundcloud.com/forss/flickermood. Let’s go back to the cryptic title of this post. Since Flickermood seems to be available all over the place, the key, being a music service or a label or an artist in this world of link-passing, is to own the most relevant copy. With most relevant I mean the most happening copy, the most accessible copy, the most usable copy, the coolest copy, the earliest copy, the most exclusive copy, the copy with the best sound quality, the most permanent copy, the most social copy, the most remixed copy, the most authentic copy, the most interoperable copy, etc.
Although the above links are all cool I’ll focus on the last one to further highlight what I mean. It’s a link to the track on SoundCloud, a service I’m currently working on, where you can listen to the full song in a pretty cool player with discussions happening inside the track through “timed” comments. I’m there too, discussing samples with friends and fans. What’s also interesting about this link is this:
- type
soundcloud.com/forss/flickermood.mp3or.aiffto get an mp3 or an aiff, etc. - type
soundcloud.com/forss.m3u,.xspf,.rssor.atomto get a playlist with all my releases soundcloud.com/forss/flickermood.rssto get a feed of comments- sniff the audio with an haudio compatible app
- go to the url with your iPhone to play it
- embed the track on MySpace or just about any other place, in a player that is better than any other out there.
http://soundcloud.com/api/tracks/flickermood.xmlandhttp://soundcloud.com/api/tracks/flickermood/stream.mp3to do just about anything with the track.- and there is in fact much more… (some of these features are still in development, in case you’re a lucky beta tester)
Good song permalinks is the shit. All this really means is that the track is so accessible, it’s impossible to top. The problem today is that the vast majority of “relevant” copies of songs are in places where labels and artists and other commercial players have little or no control over them. 1% of the listeners may be in a place where labels/artists/platforms are, they may pay, etc, but the other 99% are somewhere else–on p2p nets, on russian pirate sites, on trashy yasn sites, heck, even in their own music players. The key to survival on the emerging media web is to make these copies irrelevant by being drastically more relevant. Downloads won’t survive long in a post-scarcity publishing world–it’s making yet another irrelevant copy of an irrelevant, un-sexy copy to begin with. Just contrast iTunes with SoundCloud. Right now, I just see a lot of lost ad dollars.
Cool URI:s don’t change–but they bring change! Nightly note on music distribution.
The future of digital distribution was not about DRM protected AAC-files in your locked down iTunes or Windows media library. We already knew that. However, it’s important to realise that the future of digital music is not going to be about Mp3:s either. Not Flac. Nor is it going to be about some superior and/or open format after that. Not at all.
No. The future of digital is a link.
It’s a simple, permanent link to a song that you can grab a hold of, listen to, pass around, comment on, blog, tweet or Jaiku about, listen to in your yet-to-be-invented music player (or hey, maybe it’s Songbird?), bookmark in your social playlist tool, create a context around, remix, mashup, mix live on a wireless device that doesn’t yet exist (or is it an iPhone 3G?), embed on your insert-social-network-here profile, shout to somebody at a party, license in two clicks, play back from your home stereo or car stereo or ..heck, iPod. But it will take a few years.
Tim Berners-Lee was more right than he could ever have hoped to be. Links rule, they stay cool, and they will change the game in how music is discovered and listened to over the coming years…
(And yes, I broke my promise–but the next post will be about SoundCloud!)
Please let me *not* own my music. On Spotify, SoundCloud.
According to Steve Jobs people still want to own their music. Lately I’ve been asking both friends and strangers if they really want to own music, and I have to say that I’m not sure about if Mr. Jobs is right, at least not for much longer. In fact I think he knows that he is wrong, but he won’t admit that for a while…
It seems as if we’re in some sort of transition period regarding ownership of music today. My impression is that people, at least in countries where mobile broadband has really started to take off, e.g. the Nordics, have started to realise that what’s important no longer is owning music, but owning access to it. And I think we are nearing a tipping point in this matter.
Many friends of mine have been managing their own growing digital collections of music for quite some time now. A few years ago, the amount of data and files the average music “enthusiast” had to manage passed the point where a database model for organizing files started to make more sense than simply dropping files in folders on the filesystem. iTunes emerged as one of the best music library managers out there, with its great handling of searching, playlist tools, and more. Most people I know with large collections of music moved to iTunes, or similar systems.
To date, these people still “own” their music, at least if you define ownership as including stolen goods. After all, about 99% of my friends’ huge harddrives are filled with illegally downloaded music…
So, how many people have digital collections of music these days? Quite a few I’m sure, but my mother, who is a professional musician, certainly does not. She got an iPod not too long ago, but when I showed her how to convert her CD:s into digital files she realised immediately that it would take her far too much time and effort to convert her gigantic CD collection into an iPod-friendly format. It didn’t help when I told her she had to be careful managing the massive amounts of data this would result in, or else her collection could vaporize! Laptop drives do crash, music collections with hundreds of hours of work and years of collecting material behind them do vanish into thin air every now and then. This already happened to a good number of friends, and to myself. Have I learned anything? Do I keep a backup these days? No. Shame on me.
So, why do almost every music lover I know have their own mp3-libraries still? And, more interestingly, why do the non-techy, non-music nerd people I know increasingly listen to music through services such as YouTube and Last.fm with the amount of music sitting on their harddisks actually declining?
I think I know why. Managing your own music collection in a digital world really only means securing access. Hence it’s not really about owning, but owning the access to a certain set of files. It’s your own little realm in the giant p2p net.
What’s becoming increasingly clear is that the managing of your music collection can be outsourced. In other words, you don’t have to keep your own collection of music these days, somebody else can do it for you. It’s actually a rather old idea, but it’s finally becoming a reality. And the people with other things on their minds than just music are “leapfrogging”!
It takes quite a bit of time to manage your own music library, and it’s risky too. That’s why this only makes sense if you have enough time on our hands devoted to collecting music. If you don’t, it is much more rational to outsource.
What if you could keep your iTunes-like interface and still have access to the playlists you so tediously put together over the years? The answer is: you can.
How, you ask? Spotify is part of the answer. It is the slickest application for managing your music (yes, it’s still your music) I’ve seen so far. And yet there’s no music on your harddrive, all of it sits on Spotify’s servers. And it’s legal. A search for Michael Jackson yields his entire catalog. Pick and choose the songs you need for your playlist. Hit play, and the song starts playing immediately. Believe it or not, it feels faster than iTunes.
Get an early beta account if you can. You’ll be blown away.
What one realises when the file managment/ownership issues are out of the way is that the only really important things are your playlists. What songs do you listen to and in what combination and when? And you also begin to see that there are lots of possibilities to evolve music listening now that your music collection lives on the web, side by side with all your friends’ collections. That brings me to what we’re doing at SoundCloud, the company I’m starting, but this post is already too long! So I will have to promise you to talk more about that in the next post.
What’s he building in there?
“What’s he building in there, what’s he building in there?” is the question Tom Waits keeps asking in his famous song with the same name (last.fm, Spotify). The question readers of this blog may ask from time to time is the same, namely “what exactly is Eric building down in Berlin”? Well, as we’re closing in on the (semi-)public launch of the Cloud I hope to find the time to blog a bit here about what exactly it is we’re doing in Berlin. I also hope to put down some of the thoughts that have been occupying my mind in recent times. Stay tuned!