[b]This looks like a win-win for the artist and the consumer, a new way of mixing music and business: an industry game-changer.

I encourage you to play devil’s advocate and beat the living shit out of this concept … or implement it.[/b]

Writing full tracks is getting borderline trivial – and the generic public is cottoning onto this slowly but surely. That’s why purchasing MP3s is an activity viewed by many with actual contempt – the “how stupid are you to pay for something you could get for free, it’s just a file” attitude etc.

Enter the “music as a service” business model, with a twist.

Imagine, artists setting up music “systems” where you get to ‘choose your own adventure’ by tweaking some knobs and pressing some buttons as a “song” plays and loops. And for the more typical consumer, to whom interactivity is a burden but novelty is still a pleasure, every play could be automatically randomised a bit, and if they really like the exact rendition played that time, it can be bookmarked or saved.

It brings you the unpredictability of a real concert using today’s consumer tech — you can play the same “song” dozens upon dozens of times and hear something new every time. The best “songs” would offer the most variety for the money.

An interesting case study would be a “political dubstep generator” which reads in conspiracy theories from the internet and renders them in tune, set against a dubstep beat with varied grinding / transformersex riffs and glitched samples of political speeches, bound by a typical 3:30 song structure. A completely new track every play. An entire genre, distilled to a paid web service. If pop music really is formulaic, and I believe it is, then that’s a feature and not a bug, because it makes this sort of service commercially viable and technically feasible.

This is the kind of added value you can’t get from an MP3 collection. To make the complete experience “unwarezable”, stream the output as ‘flattened’ audio over the web, and provide a UI you can access from a browser/mobile app.

Now guess which app would be really great for putting at the centre of generating this musical output… You betcha.

you’re nuts.

now create the script.

I believe that people in general do not really seek unique personal experiences in music consumption. Most people can answer the question what music they like, but that answer is often highly dependent on other factors such as what their friends like, what seems popular and cool and expressive of the lifestyle values they’re adopting, etc. When the activity of collecting large amounts of mp3 files became popular among the masses some 10+ years ago (with the rise of Napster and similar file sharing services), it was largely about the feeling of not missing the opportunity of taking a piece of the cake. Getting something of an actual cake, something of perceived value.

But generally people will find little of such value in any “novelty” (as you wrote) that is not derived from, and conceived as, an already socially established value. That’s why this idea won’t work: “novelty” is reserved for the individual, but the music industry deals with social processes and will always try to build “value” in a static package-deal sort of way that can be controlled and exploited.

Yes, Renoise would be great for this purpose.

Something like this then “Huoratron’s Software Would Generate New Remix for Every Download”

And this, mentioned on the CDM blog:
Making Digital One-of-a-Kind: Inside Icarus’ Generative Album in 1000 Variations

It’s perfectly possible to do ‘unique’ remixes.

You have to narrow possibilities down a notch. Instead of giving all notes a chance (random 12) just narrow it down to the scale you play in (random 12 > select 0,2,4,5,7,9,11) and cancel some combinations out.
Ofcourse that’ll bring combination possibilities down but if you do this over entire songs, each beat and different parameters (note on/off, velocity, pitch, glissando, timbre, etc) you’ll get a LOT of different OK sounding tracks.

meh, i want whoever listens to my tune listening to the same thing i created myself.

well the idea of generic music based on 0’s & 1’s maybe is artistic, but where do the artists get in when it’s an actual product? are we going to end up just needing to push a button, an artificial intelligence checking your emotional status and boom a tune gets out. the world will still be filled with art, but where are the artists in this story?


As the artist you can decide the amount of room parts in the music have to differ. I can imagine tracks with a 4/4 beat at the start but a changing stutter effect for example so that some people may hear 3/4 stutter and some don’t. We’ll be able to layer tracks more and it keeps tracks interesting for people to listen to. I’m not for an complete overtake of the current system ofcourse. But it’s always been a nice concept in my eyes.

Service implies servants. Artists do not thrive as servants.

On mass, the audience doesn’t know best. Otherwise there would be more artists of serious quality contributing to an elevated culture.

A Definitive Version is an important ideal. If a best attempt of realisation of the ideal is not made then we are surrounded by broken ideas.

what’s cool with music, is that you can make it say anything,

it can have a precise meaning one day, and tell the opposite meaning an other day

for example,

music can be consumed by people that are focused on their ego that compares itself with other egos, those customers would like something “unique”

music can also serve the opposite goal : producing a social conformity, that helps people willing to share exactly the same identical musical experience with the others

I think we don’t have to choose between these different kinds of customers ; we just have to know that they exist, that nothing will change them, and that they always buy things, when you’ve identified their goals and explained how they will reach their goals with your product.

Not the exact same thing, but i kinda made something like this in Renoise some time ago, a randomized track that never ends. :)

here it is

eno did it - with bloom. That being said … there was a time when I was really into the concept of generative music, but not anymore. I think I came to realize that many aspects of music are programmatic, but the spark which gives it value is not. Also, I find that music that “works” is very delicate - the way it “works” comes down to every single note and sound choice. Anyway, these ideas are kind of hard to express in writing.

In what way would having some kind of randomised algorithm or auto-accompaniment make music unwarezable? Distribution is still going to be via ones and zeros and sharing digitally is going to be as easy as with any other type of file. Sure you could include some DRM or encryption, so that it can only be played by an mp3 player (soft or hard) that you have entered your personal key you used when you bought it but the same could be done for normal music files and protection can always be cracked.

youtube: Synplant easter egg


If you have MicroTonic, the easter egg is you can play a game of othello and will use the game play as generative sequence.

So are the majority of people. They like music because of the easy pattern and the catchy lyrics, specially the lyrics do the job.
But changing a fixed song could also insecure peoples emotion about the music, they don’t get what they expect and with certain songs they heard in a specific way this can turn out in a disappointment the next time they hear a different version.
So i consider it quite a dangerous concept to do with music when people are not prepared for the inexpectable.
Even if the lead vocals would remain the same, changing the background scene of the song can still be disturbing when people want to sing along.

This sounds llike Spotify with only generative music and a corporate blanket license implied as service terms. Am I understanding correctly?

So who’s gonna write the brogram?

I think this is a terrible idea. And I think it’s awesome at the same time.

Efff the masses. Let the record companies deal with marketing toward them. I believe most of us here are already difficult enough for them to market toward anyway.

I’ve had an idea similar to this. Because I like the idea of a difficult to capture, one time experience. My idea was more along the lines of internet radio that would be synced and mixed by robots.

For example all drum tracks, synths, bass lines, vocals, etcs. Would be thrown into a melting pot. Sorted by keys and tempos. Then spit back out in different combinations. However there could only be one bass line at a time, one drum track at a time. Some sounds can overlap without clashing. But much of this idea does seem problematic. However, I like problematic. It will give musicians something new to think about when creating parts of tracks.

Cool thread by the way.

I’ve been thinking of something similar but as an installation art piece, with interactive areas such as push buttons, light and camera which will change the parts depending on user interaction. Bit of a pie in the sky idea for me at the moment though…

Yes, that servant is a server in the cloud streaming Renoise’s “generative” output.

The tune/generator/XRNS itself is the definitive thing. If you don’t agree with this notion of dynamic realisation of the ideal, then I’d like to see your standing on interactive media in general, especially games. Games are audiovisual ensemble generators, driven by a controller. You want official, sanctioned, authentic, genuine, definitive, as the artist intended, it’s there. I’d intend that my track changes every goddamn time you hear it from my chosen streaming service(s). And that’s definitive :slight_smile:

The player, by its dynamic nature, can be programmed to come up with all sorts of cool arrangements and riffs on its own for each and every generator, and as others have said in earlier comments, such variations are completely bounded (or even unbounded) by the artist’s sensibilities. You, as the listener, are guaranteed an endless supply of all possible “cool bits” for a generator only if you have complete access to it, which is dictated by the website doing the streaming and charging you for the privilege of at least their CPU usage.

Thus, DRM is not required. You can record the 0’s and 1’s stream all you want. You’d need to warez the tune/generator source (e.g. an XRNS) itself, despite it never being released into the public. Judging by the relative non-proliferation of closed-source code on teh Interwebs, I’d say such risk is manageable. You’d also have to set up streaming from home, to get the full cross-device experience (e.g. even on those devices without enough CPU/space to support playing back your tune/generator collection dynamically in realtime). So even if you did manage to warez the generator and the player, good luck playing that from your mobile. You gotta stream it … well, until 8-core phones come out anyway… that gives us about 1 year of viable enterprise, does it not :wink:

Now, what’s to say you can’t wire this sucka up to a bunch of metrics coming off your mobile phone, like the accelerometer, the microphone, etc. driving some Hydras or whatnot to wind the energy of your track up and down, or sideways or into a breakdown or whatever. Right now I hear Nike is having a good ol’ time integrating pace-sensing with the choice of music being played over your iPhone/iPod. Let’s step it up a notch. You run faster, the current track gets wilder with hardly a few seconds of “lag” (if you can call it that). You change pace in the car you’re driving, the music changes to match (or indeed, to counteract, if you have safety applications in mind). Our much superior experience can only be delivered cross-device via the cloud, i.e. the complete experience is in fact unwarezable. The closest you’ll get is to roll-your-own, and that’s fine. This is for people who CBF rolling their own and want a 6-star personalised music experience.

So let’s try that again… “7-star” this time :wink: Party situation. No DJ required (OK, that’s not exactly revolutionary, but bear with me). Microphones, cameras, bluetooth, accelerometer (crowd jumping/stomping/movement) and wifi receivers read the crowd instead. Think a bunch of shitty second-hand Android phones on wi-fi running some app called musiXense or whatever Marketing decided to call it. Anyway, the generative track can now go respond in incredibly diverse ways that it just couldn’t before, because you didn’t have a realtime crowd available for audio/visual/sensory input until now. New monetizable thing! You can charge premiums for streaming tracks which respond to more inputs = more interactive = more emotional, more fun. The track now indeed responds like a living thing, an emotional form. A track which gets harder and more skitzo in only certain parts of the venue, where people are going off. Or it chills out a bit, if “calmer-than-baseline excitement” is supported for that track. Multiple streams, all synced exactly in all parts of the venue, just smoothly changing as you move around. It will feel like a sea of emotion, because it is indeed an audio rendition of a field of sensors which are used to gauge emotions…

Almost, but Spotify only streams static tracks. See above for “dynamic music” value proposition – the difference is as large as that between watching a movie and playing a videogame.

Have I convinced anyone extra, now, or did I lose you way before “sea of emotion” ;)