Computer Generated Music

[Tanuj Tiku]

http://www.techinsider.io/amper-media-m ... ile-2016-9

Amper media, a new start-up provides computer generated music in a few minutes right in your browser with just a few 'key words'.

As you can tell by the results, its absolutely useless at this point. As a research, I can buy into this but as a response to demand, I find it quite disturbing, the motivation.

There have been many programs in the past and recently I have seen more of these kinds of tools getting some attention. The results of all such software (AFAIK) are quite bad and pointless.

Music as we know it has evolved over hundreds of years and these are early days for computers but with things like Google deep dream, we have seen how far AI has come with regards to machine learning. Already, some programs can make algothymic music but that is totally different and very experimental in nature. Each rendition will be different from the previous one.

Humans are limited to slow biological process and it takes years to learn something. Machines on the other hand are born instantly and with a neural network, like the ones Google has been building, they can learn something much faster and a great many things, they cannot grasp at all.

Of course, there is a long way to go for machines but perhaps one day in the not so distant future, a machine could theoretically replicate a feeling through music. Until true AI, it will probably just generate newer combinations and permutations based on its learning algorithm.

I am curious to know what you guys think about this with a future perspective in mind. Today, it sounds terrible but eventually machines will catch up in many areas as they are already doing. I am not thinking of this as a threat or some kind of dooms day scenarios for musicians. The unexpected creativity and wild imagination of the human mind is how we have machines in the first place. Same thing applies to music.

But, this is a curious development. At the moment, it lacks serious development and huge funding. But, if companies like Google funded research into this in the long term, who knows where we will end up. This is also an area where AI could be used, just like how we hope today that AI will help our software understand quickly and intelligently what we want from the hundreds of parameters to get the results we desire. Not all of programming is joyful and we do waste a lot of time doing things which we know can be done faster and better by just putting live musicians in a room. That process is already undergoing major changes.

Is this the next logical step?

[Guy Rowland]

Personally (this may be foolish) I'm not too worried. We've had a number of scares a number of times. The bottom line with AI seems to be this - it's all fantastic until you're in the realms of genuine creativity, then its hopeless. Furthermore, I think fundamentally humans want it that way, not just music but books, films, paintings, theatre... I guess robots will be there as a sideshow, but humans want humanity from their creative arts. Less so when it comes to building a car or driving a truck.

I'm not saying there will be no impact in any form of music making, but I don't think its the apocalypse.

[Lawrence]

Right. The apocalypse already happened in the form of streaming and legislation that's getting friendlier to gigantic corporations.

I know we disagree on this, Guy, and I'm done with my OT-ness now. Back to robots making music!

[Raymond_Kemp]

I'm with you on that Larry, recorded music has never had less value in its history but, computer generated music will never have any value for anyone other than a computer generated listener. The real understanding of this will be by more study of our microbiome rather than out towards the perceived universe.

[Guy Rowland]

Yeees, I think the ongoing streaming debates are their own separate thing.

It'll be fascinating to see the changes to society in the coming decades. Car ownership I think will start to decline as we transition to a driverless Uber future - as it is I'm not going to encourage my kids to learn to drive unless they strongly want to. Robots / computers / machines will have a massive impact in almost every area. The BBC did a little thing last year about different professions and their liklihood of being taken over - http://www.bbc.co.uk/news/technology-34066941 . Musician's come 311th out of 366 professions (they didn't have composers). Of course that might be completely flawed methodology, but it hints at the right sort of trends I think. Creativity is the highest hanging AI fruit. If we were taxi drivers or workers in a warehouse, now would be the time to panic. By the time AI has mutated to artificial consciousness and the first virtual Mozart has a baton in its robotic hand, the rest of society will likely be totally unrecognisable from today.

[Tanuj Tiku]

Like I said, I am not thinking of a doomsday scenario. But, I do feel AI is eventually going to enter music software in a big way. We are also in a world where a lot of stuff is done with technology, not because we need it but for economical reasons or well, just because it can be done.

Yes, samples are great. But, we already know the best way to achieve that sound - hire a real orchestra and engineers in a great room. However, because of economical reasons we have moved to samples. If the goal is to create something of the highest quality then samples are already in a good place as a mock-up - we then move to record with real musicians to get that musicality and performance.

At the end of the day, it is all based on economics. Perhaps, there will be a sizable market in the future where hiring a composer is not economical. Well, it works but more money can be saved by doing it this way.

Isn't economics the biggest problem in music industry right now - devaluation of music and not wanting or having to pay for it?

[coder]

Amper seems so far to generate exceptionally boring music, which will not distract viewers from the more important visuals when used as underscore. Perfect for Hollywood pr0n, and wanna-be directors and videographers with no budget for the music.

OTOH if you want something more musically interesting, there is Nodal - http://www.csse.monash.edu.au/~cema/nodal/ :

Nodal is generative software for composing music, interactive real-time improvisation, and a musical tool for experimentation and fun. Nodal uses a new method for creating and exploring musical patterns, probably unlike anything you've used before. You can play sounds using Nodal's built-in synthesiser or any MIDI compatible hardware or software instrument.

Nodal is based around the concept of a user-defined network. The network consists of nodes (musical events) and edges (connections between events). You interactively define the network, which is then automatically traversed by any number of virtual players. Players play their instruments according to the notes specified in each node. The time taken to travel from one node to another is based on the length of the edges that connect the nodes. Nodal allows you to create complex, changing sequences using just a few simple elements. Its unique visual representation allows you to edit and interact with the music generating system as the composition plays.


Nodal is compatible with DAW software such as Ableton Live, Logic Studio, Digital Performer and Garage Band. It can transmit and receive MIDI sync. You can edit a network using the mouse and keyboard, and optionally a MIDI keyboard. Nodal recognises and sends notes, sync, continuous controller and pitch bend information.

Features

simple but powerful interface
change your composition while it is playing – great for improvisation and quickly creating new musical ideas
compatible with any MIDI synthesiser and major Digital Audio software
can run stand alone using the built-in synthesiser
live performance tools: trigger complex sequences within the network using any MIDI keyboard or MIDI input device
define your own interactive continuous controller curves (e.g. Volume, Pan, Modulation) for greater musical expression
built-in examples and tutorial information to get you started and making music quickly

That last point is worth re-iterating. This is no academic proof-of-concept hack, but a robust piece of software anyone from 5+ to 100 years old could use. It sounds perhaps a little daunting, but it really is actually fun hands-on. If you are very short of time, just install the program, and run the 14 example demos - if your interest is not piqued, then you are probably better of taking up gardening. OTOH if your interest is piqued, you may not want to tell your kids about it (After all, you don't want kiddies taking the food off your table, right?).

[wst3]

The computer geek side of me has always been fascinated with algorithmic composition tools. The two I spent the most time with are KeyKit, which was more of a general purpose scripting language, and SuperJam, a tool that did a passable job (back in the day<G>) at creating accompaniments. I also spent time playing with BIAB, but seldom managed to get results that I liked. (I still have a license somewhere, wonder if I can upgrade?). There was another tool on the Amiga from Dr. Ts - can't remember the name, and the germ for one of my favorite compositions was the result of just letting it rip on a simple motif.

In fact algorithmic composition is one of the reasons I still keep a Linux machine (actually boot disk) around. The other is DSP. The Unix community is a little more experimental.

But the reality, for me, is that while these tools are fun, sometimes even fascinating, they are huge time soaks, and I can't remember the last time I even thought about the topic. Looks like I might be losing some time again?>

[Piet De Ridder]

(...) always been fascinated with algorithmic composition tools.

Me too. And I'm not even a computer geek. The thing that appeals to me most, and also the reason why I like NI Reaktor so much, is that with this type of software a computer can be ... a computer. No charades, no posing, no camouflaging ... simply let it be what it is, and let it do stuff that only a computer can do.
Something about making music (or trying to make music) with Reaktor that I find infinitely (and genuinely) more satisfying than asking my computer to do its pathetic "I'm a symphonic orchestra"-routine again, or pretend that it is a piano or a guitar.

Having said that: still have to hear first computer generated musical idea that inspires me enough to want to take it further.

_

[Tanuj Tiku]

And now Sony is getting into it....

http://www.reuters.com/article/us-sony- ... SKBN12H1ST

[X-bassist]

As these programs are released we actually may end up agreeing with Peter Hook, who clearly states in the video that he's always written with someone else, and that collaboration with a human is more satisfying than any program can be. The whole being more than the sum of it's parts. The inventiveness and creativity of the human mind should never be underestimated. Just as producers and directors would find it laffable to have a computer direct a movie, the human element will always be important in music... sold to humans.

[Tanuj Tiku]

Agreed, 100%!

However, AI is a reality which we will face as will other areas of life. It is already happening. Of course, at first there will be the important and most lucrative areas but eventually, it will be everywhere.

In about 15-20 years, computers will be powerful enough to do enough calculations to surpass the human mind in many areas. Of course, not all as things like music are far more abstract.

But, theoretically the brain is also built of electrical connections with several modules and millions of neurons firing. It uses electrical and chemical processes. Here the only difference is that it will be synthetic and entirely electrical. Well, until quantum computers really take off. That seems unknown for now.

In about 15 years, we will start to see machines inside our bodies in some form or other. Already patients with Parkinson's have these in their brain. It requires a small surgery but eventually, the computers will become small enough to enter the body without surgery.

Just this year, the Nobel prize for chemistry was awarded to a group responsible for the design and synthesis of very small machines on a molecular scale. It is only a matter of time.

I did like Guy's POV on this that music does not need changing. But, when the internet was born, most people never imagined what would happen in 2016. Our phones and computers are already brain extenders.

[Desire Inspires]

The software will not replace humans making music. They will both exist side-by-side. Some cases, a machine will be better and in other cases a human will be better.